EEVblog Electronics Community Forum
Electronics => Metrology => Topic started by: cellularmitosis on April 28, 2018, 04:49:25 am
-
Ok, enough delays! Time to start round 2 of the cal club. :scared:
This round will use a 7V LTZ1000 board (PX-ref v2.4).
Good things about this round / improvements over last round:
- The ref is an LTZ1000, woohoo! :-+
- This particular LTZ chip shows lower-than-typical 1/f noise (see date code 9811 in this page (https://github.com/cellularmitosis/logs/blob/master/20180324-ltz1000-1f-noise/README.md))
- Reverse polarity protection (schottky on positive input)
- Output short protection (LTC2057 buffer)
- Standardized on a common set of connectors
- Low EMI sensitivity (ferrite common-mode chokes on supply and output)
- Steel TEKO case (provides some amount of magnetic shielding)
- Star-pattern shipping (should detect shipping-induced hysteresis)
- Light weight package (about 9 oz.) == cheap shipping (about $4)
Things which aren't ideal about this round:
- Due to a mishap I lost the ~3 months of ageing on the LTZ which I had intended to use for this round :-BROKE
- ...which means I also lost Doug's (voltagestandard.com) measurement using his calibrated 3458A :palm:
- This LTZ has only been ageing for a week (but has only moved 1 count on a Keithley 2015 in that time)
- Tempco of the ref hasn't been measured / adjusted via the 400k resistor
But you know what, it's never going to be perfect -- there will always be more tweaks to be made and work to be done. Time to get the show on the road!
Shipping:
This time, we'll use a star-pattern: the ref will get shipped back to me at each leg of the round. Hopefully we can use this technique to detect any hysteresis / drift picked up during shipping.
Shipping is going to be pretty cheap (about $4), so here's how I'll approach it: I'll include a return-shipping label for convenience, and if you want to drop me a few bucks in paypal, that's great. If not, no big deal -- the shipping is going to be a relatively small amount of money anyway, so I'm not worried about it.
Certainly, if you have calibrated gear, I don't think you should worry about shipping -- you're doing the rest of us a big favor just by participating :-+
Participants:
I'm not sure I have a complete list of everyone who wanted to participate in this round, so please chime in if you don't see your name!
- Muxr
- technogeeky
- VintageNut
- vindoline
- Conrad Hoffman
- dr.diesel
- CalMachine
- CatalinaWOW
- kj7e
- bitseeker
- RandallMcRee
- nikonoid
- martinr33
- orin
- Vacuuminded
- Vgkid
- MaxFrister
- SirAlucard
- flittle
- hwj-d
- rhb
- ArthurDent
- edavid
- GEOelectronics
- TWMIV
- Svgeesus
...and I think I might be sending the ref to a few surprise guests internationally as a "thank you" for being excellent contributors to the forum :-+
I think I'll choose the shipping order using a random sort, which is equally fair / unfair for everyone ::)
-
Some additional detail about the ref:
Supply voltage:
Valid supply voltage range is from about 9.1V (the LTZ falls out of temperature regulation below this point) to 15V (max specified supply voltage for the LT1013). Let's standardize on 12V for this round.
Supply connection:
The ref has a 2-pin 0.1" female header socket for the supply connection, and will ship with two "pigtails": one with banana plugs and another with croc-clips.
The pigtail connector has been colored silver on the positive lead. The CAT5 wire is solid-color for negative and white/stripped for positive. The banana plug heat-shrink is yellow for positive and black for negative. The croc clips are red for positive and green for negative.
Output connection:
The output pigtail connection is made using a paired set of Mil-Max machine pin headers with solder-cup termination. These are gold-plated beryllium copper. With a wrap of tissue, etc, these should track thermally and present minimal thermal EMF.
https://www.digikey.com/product-detail/en/mill-max-manufacturing-corp/380-10-164-00-001000/ED10164-64-ND/474594 (https://www.digikey.com/product-detail/en/mill-max-manufacturing-corp/380-10-164-00-001000/ED10164-64-ND/474594)
https://www.digikey.com/product-detail/en/mill-max-manufacturing-corp/329-43-164-41-540000/ED90338-ND/1869520 (https://www.digikey.com/product-detail/en/mill-max-manufacturing-corp/329-43-164-41-540000/ED90338-ND/1869520)
Two pigtails will be supplied for the output connection: one with a Pomona 4892 dual banana plug (gold-plated beryllium copper spring, brass body), and one with bare copper wires.
https://www.digikey.com/product-detail/en/pomona-electronics/4892-0/501-1553-ND/737021 (https://www.digikey.com/product-detail/en/pomona-electronics/4892-0/501-1553-ND/737021)
-
Thanks for putting this together and kicking off a new round. Looking forward to it. What all do you want us to capture/upload/post about the ref when we receive it?
-
Hello,
By the way: with single supply the LT1013 will go up to 30V so it is not the limitation.
It would be also a good idea to put a voltage regulator directly to the cirquit so that everyone has the same voltage (since PSRR is not infinite).
Rhats why I always use a LT1763 based regulator in my cirquits either on board or also off board with the little single sided PCB that I have already shown.
With best regards
Andreas
-
As much as possible, we should standardize our testing method. ie polling frequency, instrument setup, log file format etc.
That would make it much easier to compare/analyze results.
-
Cool! Looking forward to it.
-
Great idea.Count me in.
Located in Missouri, sort of i the middle of USA.
Have access to a factory calibrated LTZ1000 anchored Keysight DMM.
George Dowell
-
I have created FTP section and account for data store, if anyone willing to use this way to interchange files.
FTP server: ftp.xdevs.com
Login: usac_run2
Password: same as login, case-sensitive
FTP mode: passive
No size limits.
I also suggest members to list what equipment to be used, and measurement conditions. :)
Maybe group data by the "type", such as:
Type 1a: Direct measurement, uncalibrated equipment (unknown uncertainty, over calibration 1year cycle), e.g. ebay Keithley 2000, etc.
Type 1b: Direct measurement, calibrated equipment (in 1 year spec or better), e.g. HP 3458A, calibrated by Keysight in August 2017, etc. If uncertainty of calibration point can be published - even the better :)
Type 2: Differential measurement vs calibrated DC-standard (e.g. Fluke 732/Datron 491x) with null-meter connection. e.g. UCC2 ref vs Fluke 732A, null-meter HP 3458A on 100mV range.
cellularmitosis got himself into rabbit hole, but if he can maintain and update excel charts/datasheet on each spoke of the star result set, that would be really useful.
I felt like last club was somewhat black box to the outsiders, as there was no end result and no summary reports published? (could be that I missed them either, apologies if that's true). :)
-
I am not sure if this is the best place for this but for those that had bad luck on the UGSimple I think I have found a good solution. I was messing with the GUI and I noticed that if I clicked find in between each read I could get successful sequential reads. So I looked at their c++ sample and decided to try sending the same command over and over between each read. It worked so I have it logging to stdout one line for each reading.
C:\Users\flittle\Downloads>uglogger 1 3000 -1 "F1R3A1H1"
uglogger - v0.1
by flittle
Testing write & read combination before logging....Write test good.
Read test good.
Begin logging to standard out every 3000ms Ctrl-C to stop.
7.50702
7.50703
7.50703
7.50703
7.50703
7.50704
^C
C:\Users\flittle\Downloads>uglogger 1 3000 -1 "F1R3A1H1" >data.txt
uglogger - v0.1
by flittle
Testing write & read combination before logging....Write test good.
Read test good.
Begin logging to standard out every 3000ms Ctrl-C to stop.
^C
If anyone is interested I can post the code and/or link to an exe.
if you look at their sample c++ i just put this in a loop to get it to read sequentially.
proc = Gwrite(address, (char *)commands.str().c_str());
istringstream r(Gread(address));//read from device
double dr = 0.00;
r >> dr;
cout << dr << endl;
Sleep(readDelay);
please excuse my vc++ I am used to arduino these days.
-
Cellular: Just a friendly head's up - and not to be a wet blanket - but that LTZ you have is a mere infant and hasn't been under power near long enough to stabilize. I'd suggest to wait a while to let it settle down before starting shipping - otherwise the first people to receive the unit probably won't be on the same sheet of music as the people who get it later on. The fact that it's drifting about the same rate as your meter may or may not be very meaningful. Trying to hurry along here isn't going to help the data at all.
Be careful of a '2057 used by itself as an output buffer...keep the current very low (<<2mA), say driving a meter input circuit only. Otherwise its demodulater won't work well and you'll get errors. For instance: If someone were to drive an ADC Ref input that would need another amp in between. Or they should be careful trying to do a direct comparison with another Vref, etc. AZ Choppers usually need another companion amp in their feedback loop if you're using it as a buffer - and that should be low noise.
Carry On!
-
What all do you want us to capture/upload/post about the ref when we receive it?
Nothing in particular -- I figure each individual will probably be interested in different things anyway, so we'll get a variety of measurements and perspectives.
By the way: with single supply the LT1013 will go up to 30V so it is not the limitation.
It would be also a good idea to put a voltage regulator directly to the cirquit so that everyone has the same voltage (since PSRR is not infinite).
Thats why I always use a LT1763 based regulator in my cirquits either on board or also off board with the little single sided PCB that I have already shown.
Ah, that's a good point: +/-15 == 30 :)
I could include an off-board regulator as a "pigtail" -- that's a good idea.
As much as possible, we should standardize our testing method. ie polling frequency, instrument setup, log file format etc.
Good idea!
Have access to a factory calibrated LTZ1000 anchored Keysight DMM.
PS
Presently can competently test for
DCV microiVolts to kiloVolts
R.F. ULF to uW
Resistance to milli to TOhms
Capacitance and inductance
Pulse width and repetition rate.
:-+ :-+ :-+
I have created FTP section and account for data store, if anyone willing to use this way to interchange files.
cellularmitosis got himself into rabbit hole, but if he can maintain and update excel charts/datasheet on each spoke of the star result set, that would be really useful.
I felt like last club was somewhat black box to the outsiders, as there was no end result and no summary reports published? (could be that I missed them either, apologies if that's true). :)
Thanks! And that's good feedback -- there wasn't ever really an overall analysis (I guess I lost interest because the same ref couldn't continued to be passed around)
If anyone is interested I can post the code and/or link to an exe.
if you look at their sample c++ i just put this in a loop to get it to read sequentially.
That's fantastic! I know the UGSimple is currently the cheapest option for GPIB, and I'm sure a lot of metrology forum members would be interested in pursuing that further (myself included -- I also have a UGSimple unit). I would love to see a "UGSimple hacks" thread!
Cellular: Just a friendly head's up - and not to be a wet blanket - but that LTZ you have is a mere infant and hasn't been under power near long enough to stabilize. I'd suggest to wait a while to let it settle down before starting shipping - otherwise the first people to receive the unit probably won't be on the same sheet of music as the people who get it later on. The fact that it's drifting about the same rate as your meter may or may not be very meaningful. Trying to hurry along here isn't going to help the data at all.
Be careful of a '2057 used by itself as an output buffer...keep the current very low (<<2mA), say driving a meter input circuit only. Otherwise its demodulater won't work well and you'll get errors. For instance: If someone were to drive an ADC Ref input that would need another amp in between. Or they should be careful trying to do a direct comparison with another Vref, etc. AZ Choppers usually need another companion amp in their feedback loop if you're using it as a buffer - and that should be low noise.
Yeah, that's a totally valid criticism of this setup, and something I should probably highlight more for those unfamiliar with Vref ageing : at one week in, the ref is still in the steepest part of its ageing curve.
I intentionally went with an older chip, hoping that the new ageing curve I've kicked off by soldering it won't be as steep as the ageing curve of a brand new chip. Maybe!
I debating between keeping everyone waiting for another couple of months, but then I thought "$4 in shipping is pretty cheap entertainment", so I thought I'd start up round 2 now and see if anyone was interested.
-
Hi Cellular,
very nice to see its going on now. Thank you so much.
I would throw in two simular devices, thats now under power, since I show it in the px-thread from the 8. of March, but not (temp-) calibrated except with my 34461a.
As an idea, I had implemented a lm35caz TO-92 each, that has +-0.5 to 1°C accuracy, connected to supply voltage. So we can measure the internal temp directly as voltage (multiplied with 100) with an extra wire.
Also I have laying around some pizerow's. That's the little things with wlan and bluetooth on board (to be switched off during the measurement), I also would throw in two of them. They are thin as two stamps, so it doesn't matter to send it in the package. If someone could write a script (have no time to do that for now) for interfacing, measuring, uploading to TiN, we have also a standardized protocol method, if that makes sense, as a first thought!
Maybe one can additionally circular in the USA, the other comes back to EU, and vice versa? I would send them to cellular at the first. Or one to Andreas? One to Cellular?
Whats your thoughts?
Thanks to all.
HW
-
I would certainly appreciate guidance on the type of data to be collected. I am only marginally a volt-nut and my personal interest is primarily in seeing roughly where my instruments lie. But I am very willing to help in ways that I can. Obviously with a population of long out of calibration instruments, without even access to the last calibration data I will not have a meaningful contribution to absolute accuracy, and any contribution I can make on drift will be difficult to tease out and only available after this loop has been around multiple times.
I can get what I believe to be meaningful noise data (including spectral cuts of various types), tempco data over a limited range of temperature. I also have a couple of HP mux switches which I can set up to make successive closely timed measurements of multiple sources using the same instrument.
What type of data would be most useful to the group? I can guide my efforts in that direction.
-
Looking forward to it CM! Thank you for organizing this again.
-
I would certainly appreciate guidance on the type of data to be collected. I am only marginally a volt-nut and my personal interest is primarily in seeing roughly where my instruments lie. But I am very willing to help in ways that I can. Obviously with a population of long out of calibration instruments, without even access to the last calibration data I will not have a meaningful contribution to absolute accuracy, and any contribution I can make on drift will be difficult to tease out and only available after this loop has been around multiple times.
I can get what I believe to be meaningful noise data (including spectral cuts of various types), tempco data over a limited range of temperature. I also have a couple of HP mux switches which I can set up to make successive closely timed measurements of multiple sources using the same instrument.
What type of data would be most useful to the group? I can guide my efforts in that direction.
I feel probably similarly to Catalina here. I'm at best a millivolt-nut (or volt-millinut perhaps?), but would love to get a sense of where my instruments lie as well as possibly contribute something back to the group (beyond just the obvious few bucks for shipping offsets).
If there's a semi-standard process to adhere to, all the better, as that gives me something to direct my efforts as well as to learn about why things are done that particular way.
-
I think I should point out that summer and winter are poor times to be shipping a voltage reference as they will likely be subjected to extreme temperatures in transit. We're already at the point where the back of a truck could hit 135 F sitting in a freight yard for a few hours while coming and going from Austin.
We did a weekend move in July in Dallas from one location to another location about 15 miles away. Not a single one of the workstations would boot on Monday morning until I went around and reseated all the connectors.
-
I've got my ltz1000's and opamp's from Digikey Thief-River-Falls in wintertime, they had nearly minus 20°C a that date coming with plain with unheated cargo space, and now, they are all defective, as we can see ... :palm: :popcorn:
-
For the nature of this project, long term burn-in, shipping temp and a unified standard of measurement may be out of scope. The fact we have an LTZ1000 to send around is on order of magnitude better than the SVR-T from round one. The limiting factor for many here is going to be their own gear, not the DUT. Shipping and external environmentals cant be controlled. A standard of measurement is hard to define since each has unique gear and logging ability. I think the spirit of this is just to provide the best data you can and maybe learn where you can improve along the way. Looking forward to contributing as I can.
-
I certainly want to participate. But my brother in-law makes fine wines and they do not ship wine to Texas in the summer. Jason had a mishap with the LTZ1000 and I thought it worth pointing out that waiting until the fall might be a good idea. It would give the reference time to reequilibrate and would avoid the risk of subjecting it to additional thermal stress.
I intend to contribute some passive references. i have a bunch of NOS WW II mil-spec stuff that should be very stable by now.
My eBay 34401A reported 5.00000 VDC +- 1 digit on a DMMCHeck Plus recently calibrated to 1 ppm on an in cal 3458A. The 34401A appears to have had its last cal in 1999.
There are strong fundamental physical grounds for thinking that aging drift is asymptotic to zero after a sufficiently long period. Proving this would require a lot of data spanning 20 years, so I'm not likely to be able to prove it except indirectly by comparing 34401A and 3456A measurements of a precision reference and plotting measured values against last known cal.
-
For the nature of this project, long term burn-in, shipping temp and a unified standard of measurement may be out of scope. The fact we have an LTZ1000 to send around is on order of magnitude better than the SVR-T from round one. The limiting factor for many here is going to be their own gear, not the DUT. Shipping and external environmentals cant be controlled. A standard of measurement is hard to define since each has unique gear and logging ability. I think the spirit of this is just to provide the best data you can and maybe learn where you can improve along the way. Looking forward to contributing as I can.
:-+
But my brother in-law makes fine wines and they do not ship wine to Texas in the summer. Jason had a mishap with the LTZ1000 and I thought it worth pointing out that waiting until the fall might be a good idea. It would give the reference time to reequilibrate and would avoid the risk of subjecting it to additional thermal stress.
That's a valid point, and is certainly something I would plan around if I were paying $$$ for a "real" calibration.
Again, this round isn't "perfect", but its time we got the ball rolling again. There will always be future rounds where we can dial this in further!
-
AFAIK, calibration lab techs don't get the summer off due to lack of business. It's also doubtful that sales and shipments of precision meters cease during mid winter and mid summer. The storage temperature range for the LTZ1000 is -65 to 150 C, so I doubt you could damage one. Didn't see a spec for metal foil resistors, but I'd be surprised if it was an issue. The big question is hysterisis of everything as you cycle the temperature and the only way to find out is to try it. Temperature and vibration in shipping is one reason I dislike trimpots. Too many change by tapping them, so they're certainly subject to changes during shipping.
-
Erm, hi, guys. I just thought something like this might help with shipping your LTZ. This info is for DHL, but most large courier companies will offer similar facilities.
http://www.dhl.co.uk/content/dam/downloads/g0/express/services/industry/dhl_express_temperature_sensitive_packaging_brochure_en.pdf (http://www.dhl.co.uk/content/dam/downloads/g0/express/services/industry/dhl_express_temperature_sensitive_packaging_brochure_en.pdf)
-
Fractional and even Ohm resistors-
Once, in a package from Russia, with some items totally not related to this group discussion, the seller included 2 boxes of 1/2% 200 ohm resistors with a note of thanks for the purchase. Each box contained several cardboard carriers with 25 resistors lined up in idividual precise slots.
At that time I checked them with the bench Simpson 260 and sure enough they were 200 Ohms. Checked several and they were all 200 Ohms. On the Simpson VOM all tested were remarkably the same, so next they were checked on the bench Simpson 463, a 3 1/2 digit .1% (I think) digital meter. There they were again, 200.0.
That's when I decided to put them away for a future project, aside from the usual resistor stash (I almost said TRASH!).
Reading this thread today prompted a second look at them, this time with updated equipment
Each sample tested read between 200.2 and 200.7 Ohms. A random sample of 10 tied in parallel read 20.0529. Interesting.
Next step was to figure out what combinations 10 or 20 of those could created if hooked together with clip leads, in various clusters of series/ parallel groups.
Simple calculated that from 1 to 10 in parallel would yield 10 different resistances, same # for series. So 20 different combinations divided over 2 groups of 10 resistors.
Next I tackled the possible combinations that utilized one set of series plus one set of parallel possibilities, in series. This yielded 100 combinations
I didn't even begin the possible combinations in parallel.
There are some very interesting numbers on that list, which include 22.22222, 33.33333, 66.66666 266.66666 along with many the normal whole numbers starting at 10.
This is where my research stalled (read: mind=boggled) and decided to ask the computer experts here if they can do a spreadsheet that would calculate all the possibilities of 20 ea. 200 Ohm resistors.
I can envision two smallish boards with two rows of 10 terminal each, one row on each side and the resistors soldered across the terminals ladder style or maybe criss-crossed so they would all be in series already.
Clip leads to do the dirty work, and a spread sheet for "programming" information.
If this or some similar would be of interest to the traveling calibration folks, I will donate 100 resistors to the cause, enough for 5 pairs.
George Dowell
-
"This is where my research stalled (read: mind=boggled) and decided to ask the computer experts here if they can do a spreadsheet that would calculate all the possibilities of 20 ea. 200 Ohm resistors."
This sounds like it could be a big deal... with 20 resistors, there are 20 factorial ways of ordering them - although half of them are aliases. It is a neat question, especially if all the resistors have different values.
Then, you can take readings from 18 points in the line. So you could multiply by 18, (or 18^2 if you measure two points relative to each other, rather than one against ground).
That calculates out to 20!/2 *18 = 2.18 E19... this could take a while.
And then there are all the series-parallel combinations that you can think up.
-
AFAIK, calibration lab techs don't get the summer off due to lack of business. It's also doubtful that sales and shipments of precision meters cease during mid winter and mid summer. The storage temperature range for the LTZ1000 is -65 to 150 C, so I doubt you could damage one. Didn't see a spec for metal foil resistors, but I'd be surprised if it was an issue. The big question is hysterisis of everything as you cycle the temperature and the only way to find out is to try it. Temperature and vibration in shipping is one reason I dislike trimpots. Too many change by tapping them, so they're certainly subject to changes during shipping.
Of course, mine were indeed not damaged. ;)
-
Precision Resistor Standard with many combinations:
I think I have this worked out. It will take 20 identical resistors in series, at the junction of each pair- a center-off DPDT toggle switch, use both contacts in parallel. 21 switches total. (Clip leads can substitute for actual switches)
Two buss bars, 2 binding posts.
Each junction can be neutral/Buss1/Buss2.
Any number of resistors in series could be placed in parallel with any other number of resistors in series.
The range would be from R/20 to 20R.
It's up to the math whizs to calculate the programming chart.
I'll call it the Diddle Box.
George Dowell
have a look at this:-
https://www.eevblog.com/forum/metrology/poor-man_s-resistive-ring-10k-standard/msg1447922/#msg1447922 (https://www.eevblog.com/forum/metrology/poor-man_s-resistive-ring-10k-standard/msg1447922/#msg1447922)
and especially this:-
https://www.eevblog.com/forum/metrology/poor-man_s-resistive-ring-10k-standard/msg1456821/#msg1456821 (https://www.eevblog.com/forum/metrology/poor-man_s-resistive-ring-10k-standard/msg1456821/#msg1456821)
best regards.
-zia
-
Thanks Zia
George Dowell
-
I would love to participate in this as I have two 4356a's that differ 4 counts from each other.
I have a gpib interface, temp sensors, and I believe everything else needed to measure and trend the data.
I am going to be doing an internship until the beginning of August, so I wouldn't be able to participate until then.
I may also be able to check it against a calibrated meter in one of my school's labs, but I wouldnt be able to get more than a few hours of data from it.
Sent from my Moto G (4) using Tapatalk
-
AFAIK, calibration lab techs don't get the summer off due to lack of business. It's also doubtful that sales and shipments of precision meters cease during mid winter and mid summer. The storage temperature range for the LTZ1000 is -65 to 150 C, so I doubt you could damage one. Didn't see a spec for metal foil resistors, but I'd be surprised if it was an issue. The big question is hysterisis of everything as you cycle the temperature and the only way to find out is to try it. Temperature and vibration in shipping is one reason I dislike trimpots. Too many change by tapping them, so they're certainly subject to changes during shipping.
Well no, that's not a question of damage, but of hysteresis of the chip, in the completed circuit.
Have a look into the specifications of the Datron 4910 and 7001, both are limited in their storage temperatures.
The 4910 from -40 to +50°C (Great!), but the 7001 has limits of -18..+45°C for transit, and 0 ..45°C for storage, plus the clear hint inside the manual, that excessive (low) temperatures will cause hysteresis, being removed by the unique conditioning procedure.
Early this year, I also shipped two LTZ1000 modules during cold weather to friendly volt-nuts, several degrees below zero °C, and both references showed a persistent shift of output, several ppms, although they showed absolutely no hysteresis during their characterization, between +15° and +40°C, at 50°C oven temperature.
Therefore, the selection of seasonal shipment is very important, as is the star-wise shipment, to get a baseline, back at the initial lab.
Although both of my references had shifted, I could to some limits decide from these drift data, which voltage value of the returned sample was to be expected... So I concluded in the end, that my Volt agreed to acberns calibrated Volt within parts of one ppm.
Frank
-
Interesting, I'm guessing the Pickering-patent anti-hysteresis conditioning in the 7001 only applies to the LTZ chip itself (by manipulating the heater set-point), and no equivalent anti-hysteresis conditioning is performed to the precision resistors?
Dr. Frank, were you able to attribute different levels of hysteresis to the 7V vs 10V output of your ref?
-
Interesting, I'm guessing the Pickering-patent anti-hysteresis conditioning in the 7001 only applies to the LTZ chip itself (by manipulating the heater set-point), and no equivalent anti-hysteresis conditioning is performed to the precision resistors?
Dr. Frank, were you able to attribute different levels of hysteresis to the 7V vs 10V output of your ref?
Correctly, that's the scope of the patent. Still, I doubt that that procedure was effectively working inside the 7001, as in another document, Pickering mentions a low oven temperature of maybe 45°C only, which would prohibit a symmetrical +/- 40°C temperature swing.
I'm looking forward to TiNs reverse engineering..
I did only monitor the 7.15V output.
But I also encountered the well known hysteresis on the 10k VHP202Z resistor reference, which was also in these packages..
'Well known' means, that Vishay re-measured two of them at -40 and +175°C, and sent them back with a hefty persistent shift.. which also could be removed by thermal cycling.
The direction of the shift correlated in both cases with the direction of the temperature excursion.
So the specific resistors for the 7/10V amplifier in your circuit should also be characterized for a broader temperature range.
-
For a reference which travels by common carrier, it seems to me that several features are desirable:
the ability to thermally cycle the entire reference by means of a Peltier device through an annealing cycle as well as maintain a measurement temp independent of ambient.
temperature logging during transit and use
phase change media container with 5-10 C and 35-40 C melting points if we can find affordable media
lots of insulation in the shipping container (e.g. 6+" of styrofoam)
I can't see any reason that a reference can't travel and provide results close to the limits set by 1/f noise with good lab technique. Not likely to be easy to achieve, but physics is not random. It just gets to set the rules to suit itself.
-
Ok.
Now that we know in detail what would be most desirable, along with all the negative circumstances that might cause the project to fail for transport reasons, let's start with what would be practically feasible.
The task:
to transport a 50mm x 55mm x 25mm metal box and accessories over a period of about 1 week as temperature-neutral as possible.
That shouldn't be an insoluble task, also for pedantic Voltnuts! :P
And please, we are not dealing with requirements as they are placed on a 732a/b, nor is the operation of a peltir element in a "in itself contained" transport box not particularly, excuse me, "intelligent". ;)
How about, for example, a matching, slightly more solid but as small as possible styrofoam box, as used in the medical and biological fields?
-
For a reference which travels by common carrier, it seems to me that several features are desirable:
the ability to thermally cycle the entire reference by means of a Peltier device through an annealing cycle as well as maintain a measurement temp independent of ambient.
temperature logging during transit and use
phase change media container with 5-10 C and 35-40 C melting points if we can find affordable media
lots of insulation in the shipping container (e.g. 6+" of styrofoam)
I can't see any reason that a reference can't travel and provide results close to the limits set by 1/f noise with good lab technique. Not likely to be easy to achieve, but physics is not random. It just gets to set the rules to suit itself.
hello,
use these:-
http://www.tempil.com/temperature-indicators/temperature-indicating-sticks/tempilstik/ (http://www.tempil.com/temperature-indicators/temperature-indicating-sticks/tempilstik/)
best regards.
-zia
-
The point of the Peltier devices is not to maintain a constant temperature during transport. It is to provide facilities for annealing if it is required. As they appear to be low cost, it seems to me sensible. I'll know more when the 10 I ordered arrive. I can't see any justification for shipping hot. I think there are less difficult solutions.
Using Peltier devices for annealing appears easy and cheap. Finding phase change materials is a lot harder. That's in the "would be nice' category.
I can see no reason not to attempt to develop shipping procedures suitable for a 732a/b. As a scientist, I have been quite underwhelmed by what I've read on the subject. It is perhaps merely poor writing, but I have not been impressed by what I have read so far.
What I have suggested are the same procedures as used for shipping food or medical materials. I have an acquaintance who represents a company which makes temperature controlled shipping containers for medical use. Subjecting a voltage reference to dry ice in a styrofoam chest as done with steaks is not a good idea if you want the voltage at the receiving end to be the same as the voltage at the shipping end.
There are three requirements:
Maintain as consistent a temperature in transit as possible
Track what the transit temperatures are
Have a means of annealing the device if required
I happen to think that the "amateurs" n this group are capable of doing first rate professional level science with a modest amount of cooperation. In my view the biggest obstacle is people who "fail to recognize their limitations" to paraphrase Harry Callahan in "Magnum Force".
-
I happen to think that the "amateurs" n this group are capable of doing first rate professional level science with a modest amount of cooperation. In my view the biggest obstacle is people who "fail to recognize their limitations" to paraphrase Harry Callahan in "Magnum Force".
This is a nice attempt to shift a technical problem to the side of an "ad hominem". I do not like such categorizations because they don't contribute to the solution. It is sufficient if we try to solve the technical priority, please.
-
I can't see any reason that a reference can't travel and provide results close to the limits set by 1/f noise with good lab technique. Not likely to be easy to achieve, but physics is not random. It just gets to set the rules to suit itself.
Hi,
Unfortunately, there can still be serious limits to how practically predictable it is - think weather, earthquakes, etc.
Just because semiconductors are small does not mean that they're simple. I can't remember the details now, but there was a paper a while back comparing microfractures in a crystal with faults in the earth's crust.
-
:popcorn: USA Cal club reference ready to do some initial checking now.
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_fxref_proto_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_fxref_proto.jpg)
-
:-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-+ :-DMM
-
What a beauty ... :-+ :-+ :-+
-
:popcorn: USA Cal club reference ready to do some initial checking now.
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_fxref_proto_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_fxref_proto.jpg)
Wow! That looks really fancy to my volt-noob eyes. Looking forward to working with it.
-
I would be very happy to participate in this round, if that is okay. My meter is not up in volt-nut territory, it is a 6.5 digit Keysight 34465A so this LTZ would be considerably better than the LM399 in the meter. On the other hand, the meter is a recent purchase (purchased new, four months old) and it was purchased with both calibration and uncertainties.
-
I would be very happy to participate in this round, if that is okay. My meter is not up in volt-nut territory, it is a 6.5 digit Keysight 34465A so this LTZ would be considerably better than the LM399 in the meter. On the other hand, the meter is a recent purchase (purchased new, four months old) and it was purchased with both calibration and uncertainties.
PS I live in Missouri.
George Dowell
-
I would be very happy to participate in this round, if that is okay. My meter is not up in volt-nut territory, it is a 6.5 digit Keysight 34465A so this LTZ would be considerably better than the LM399 in the meter. On the other hand, the meter is a recent purchase (purchased new, four months old) and it was purchased with both calibration and uncertainties.
That sounds great!
----
I threw together this little guy this evening, to include with the ref. It also spits out CSV data over the USB (serial) connection (for datalogging).
Now I just need to print up a shipping label and get this train a movin!
-
Initial tempco looks alright. Good enough not to spoil linear response with compensation resistor voodoo. My output 10V stage on previous boards also have opposite tempco so overall we might break even.
(https://xdevs.com/usac_run2/_xdevs_fx_ref/usacfx_ref_test1.png)
I'd like to make note that this reference needs bipolar 12V supply!
-
I threw together this little guy this evening, to include with the ref. It also spits out CSV data over the USB (serial) connection (for datalogging).
Nice. We can be on the same page as far as temperature is concerned. I know my Control Company NIST traceable thermometers are out of cal.
So I dropped off the Fluke 8845A for calibration with data at Fluke Command Central in Everett this morning. It was quite entertaining wading through the quagmire of their web site to get the RMA... No, it's not broken, I just want a calibration. The signs on the Fluke Service Center building are no better... SW corner of the building... OK... Locked door with threatening signs about requiring escort beyond this point, but how to get their attention?
With a little luck it will be back and still be in cal when it's my turn ;) It's the 34461A's turn next year.
-
I'd like to make note that this reference needs bipolar 12V supply!
TiN, what kind of stability does the supply need to have to power it?
-
Decent quality linear bench PSU will be fine. For critical measurements (e.g. noise) I use two VRLA 12V batteries.
-
12v Bipolar power supply... Check!
On another but related note, I'm putting together a post detailing the recent journey my KX based LTZ1000 based 10v reference made to ENI Labs (thanks CalMachine!), where it was compared to their house master references as well as CalMachine's home references over the last few weeks. The reference was shipped powered on both directions. Here is a very brief summery, more details coming in its own thread.
As measured by my Keithley DMM7510 prior shipping;
10.000000
As measured at ENI Labs;
9.9999637 - 3458 option 002, cal'ed at Fluke Std Labs 5 moths ago
Dif;
0.0000363 (3.6 ppm low)
Ryan adjusted the buffer trim upward by 37.7uV to the reference now measured;
10.0000014 (0.14ppm high)
I should now “theoretically” measure;
10.000038
My real measurement now is;
10.000039
I love it when the math checks out.
Also, this was adjusted per his home Fluke 732B, which was last known to measure 10.0000007 or 0.07ppm high compared the ENI Labs master 732B (which was away at Fluke Park Standards lab the last few weeks). Ryan is going to verify his 732B against the Shops 732B again to re-verify this number. So I'm confident I now have a very, very good realization of 10v to compare from. :-+
-
Decent quality linear bench PSU will be fine. For critical measurements (e.g. noise) I use two VRLA 12V batteries.
how about going the battery route as the standard power supply protocol.
it is cheap enough and readily available.
and keeps the mains-line-daemons away ;)
best regards.
-zia
-
Decent quality linear bench PSU will be fine. For critical measurements (e.g. noise) I use two VRLA 12V batteries.
OK, just checking. Thanks for the confirmation.
-
We should probably include this, otherwise we add one additional variable, and the possibility of hitch-up damage.
-
That sounds great!
Thanks!
I threw together this little guy this evening, to include with the ref. It also spits out CSV data over the USB (serial) connection (for datalogging).
Nice. I have a 4-wire PT100 temperature sensor, so I can also provide temp readings from that (but no humidity sensor alas).
-
Nice. I have a 4-wire PT100 temperature sensor, so I can also provide temp readings from that (but no humidity sensor alas).
That would be great -- I still haven't gotten around to performing a boiling-point / ice-bath calibration (on anything!) yet!
-
Or the triple point of water method. Amazing this about the most accurate way to achieve a known temp.
https://us.flukecal.com/literature/articles-and-education/temperature-calibration/video/triple-point-water-realization-techn
-
Or the triple point of water method. Amazing this about the most accurate way to achieve a known temp.
https://us.flukecal.com/literature/articles-and-education/temperature-calibration/video/triple-point-water-realization-techn
so... as it happens, there is a glass blowing shop which is going to open about a month from now, which is literally in the same office park as where I work.
Been planning on learning how to make glass-to-metal feedthrough seals, but now I guess I'll have to add triple point cells to that list... >:D
-
Or the triple point of water method. Amazing this about the most accurate way to achieve a known temp.
https://us.flukecal.com/literature/articles-and-education/temperature-calibration/video/triple-point-water-realization-techn (https://us.flukecal.com/literature/articles-and-education/temperature-calibration/video/triple-point-water-realization-techn)
even the ice-bath (double point cell) method (very easily do-able) is very good.
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0ahUKEwih-rO_uejaAhVBcRQKHVoaBPsQFggpMAE&url=http%3A%2F%2Fwww.burnsengineering.com%2Flocal%2Fuploads%2Ffiles%2FApproximating_the_TPW_Presentation.pdf&usg=AOvVaw2iIgLhXVRfoLhXa0c6GoYR (https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&cad=rja&uact=8&ved=0ahUKEwih-rO_uejaAhVBcRQKHVoaBPsQFggpMAE&url=http%3A%2F%2Fwww.burnsengineering.com%2Flocal%2Fuploads%2Ffiles%2FApproximating_the_TPW_Presentation.pdf&usg=AOvVaw2iIgLhXVRfoLhXa0c6GoYR)
best regards.
-zia
-
Been planning on learning how to make glass-to-metal feedthrough seals, but now I guess I'll have to add triple point cells to that list... >:D
just break apart a 2N3055 (TO-3) style package, ie; take the cap off, kill the bond wires,
now you have two glass feedthrough's available *including* pretty easy mounting (seal with shellac).
(flea-bay / taobao can be a cheap source of these)
best regards.
-zia
-
I think the original article had more illustrations, but it didn't seem too difficult to build- https://www.scientificamerican.com/article/tackling-the-triple-point/ (https://www.scientificamerican.com/article/tackling-the-triple-point/)
-
That's George S. from the Fusor group!
George Dowell
-
Ok, today's lucky winner is Vgkid! I've got the ref packaged up and it will ship to him tomorrow.
I threw together a little LM7812 board, so that we can also have a shared supply voltage to use.
To get us started, I hooked the ref up to a Keithley 2015 and logged it overnight, and then the next day, at different temperatures in my apartment. It looks like the ref + DMM have a significant tempco. (Relative humidity isn't in the charts, but has been in the low 70% recently).
https://github.com/cellularmitosis/logs/tree/master/20180503-k2015-cal-club-ref
-
Looks like the combined tempco of DMM + ref is about 0.6 ppm/C. :o
-
Looks like the combined tempco of DMM + ref is about 0.6 ppm/C. :o
CM, thanks for getting thing going for round 2! I have one question though. How are you able to get sub micro-volt resolution for your data with a 6 1/2 digit meter? Is there some smoothing or averaging? If so, I think we should provide that info with our graphic data so others can try and reproduce with it. Thanks.
-
Nope, that’s the secret of the Keithley 2000 / 2015 and HP 34401a: they give you way more resolution over GPIB / RS-232. Not more accuracy, of course ;D
-
Looks like the combined tempco of DMM + ref is about 0.6 ppm/C. :o
90% meter, 10% DUT.
-
I just wanted to chime in to say a few things:
1. Thanks for doing this again! This is an awesome outcome of a great community. There is a lot of individual effort involved too, which I'm sure we all appreciate.
2. Is the TiN-made reference going to be part of USA Cal Club 2? Or 3? Or is he just teasing us?
3. I wanted to remind cellularmitosis to make sure and check that people's addresses are the same before shipping. If they have changed, the device might get blackholed :(
-
Oh, good point about double-checking addresses!
I’m not sure what TiN’s timeline is (I’m not sure if he’s waiting on more resistors), but it is possible round 2 and round 3 might overlap! My eventual goal is for the cal club to have a “library” of references, which can be checked out at any time.
-
Teasing? Tempco? What tempco. :-DD
I think my tweaking efforts should end on -0.018 ppm/K, based on Live SVG data from today run. (https://xdevs.com/usac_fx1_test3/) at that reaching 3458A resolution/noise already even with 32C span. Next step is 10V output stange tuning.
I don't need any more parts for USA Cal club FX ("wish to say that for rest of them, eh"), so it's just testing time and then initial 200 hours monitoring to see that reference settled.
Then calibration to my lab DC Volt (that is, measurement only). Then another 800 hours to make sure everybody happy and stable, final TC characterization with everything assembled, noise measurement and then it's ready. So if all goes well, that is 50 days from now. I don't think rushing anything with LTZ1000A-based reference make any sense at 500$ reference. :popcorn:
Also how about 10KOhm standard for USA Cal Club? There is L&N 4030 (https://xdevs.com/ln4030-10k_t1/) measured 10000.215 Ohm (+/-8ppm max abs) that just sits, doing nothing. This is giveaway resistor, I bought it specially for USAC needs, it was just $50.
-
That’s great news TiN!
Interesting to hear about the L&N resistor — I saw some of those on eBay. I think those are manganin? I have heard they don’t like being handled a lot, but the stability can be very good. I wonder if it would better as a reference which stays at home, which a traveling transfer standard can be compared to? Do you have any idea of the tempco or stability yet?
-
Well, ManateeMafia measured it on his calibrated 3458A (click on the link) over a week, so that much you have for stability. "Oscillations" is due to his aircon setup in warm sunny house :).
Since me and MM use higher grade standards to transfer between our labs, and "stay-at-home standards are SR104's", I don't really care if this L&N unit dies from USA Cal club misuse or shipping damage. After all you can use it as testament if 4030 can survive frequent shipping with regular post service ;D. All you need to do is to send address where to ship. All I want in return is some data plots from participants :-DMM
-
phase change media container with 5-10 C and 35-40 C melting points if we can find affordable media
a pretty cheap phase change material can be made by microfine silca (AeroSil type, SiO2) and paraffin wax
with melting / freezing point at 55/45 degC.
ref:
Fabrication and Properties of Microencapsulated Paraffin@SiO 2
Phase Change Composite for Thermal Energy Storage
Benxia Li,* Tongxuan Liu, Luyang Hu, Yanfen Wang, and Lina Gao
can be an *excellent* adjunct for transport of vrefs.
best regards.
-zia
-
phase change media container with 5-10 C and 35-40 C melting points if we can find affordable media
a pretty cheap phase change material can be made by microfine silca (AeroSil type, SiO2) and paraffin wax
with melting / freezing point at 55/45 degC.
ref:
Fabrication and Properties of Microencapsulated Paraffin@SiO 2
Phase Change Composite for Thermal Energy Storage
Benxia Li,* Tongxuan Liu, Luyang Hu, Yanfen Wang, and Lina Gao
can be an *excellent* adjunct for transport of vrefs.
best regards.
-zia
That's a good suggestion, but 131 F is too high in my view (but I also have no data). However, one might be able to create a phase change medium using a mixture of paraffin and mineral spirits. I use that as a wood finish and it's remarkably temperature sensitive. In my case it was an annoyance as the paraffin was coming out of solution. Time for another round of phase behavior of petroleum. I've got a copy of the NIST STRAP program, but it's probably 15+ years out of date and on 5 1/4" floppies.
I'd not considered mixture of petroleum products when I posted that. My focus was on bulk modulus when I dealt with the subject, but now that you bring it up, I suspect it would be relatively easy to come up with phase change media tuned to the problem of shipping precision references. A blow molded box top and bottom with two different petroleum mixtures.
There is, of course, the Swedish concern which rents temperature controlled containers for shipping medical supplies to places like the African interior. But rather pricey for a hobby pursuit. And they are really in the shippable refrigerator business as opposed to maintaining 23-25 C.
-
Not endorsing use of a phase change material, but Glauber's salts in the decahydrate form has a melting point of 32 C (91 F). Nice temperature for protecting the reference, but being well below the ambient temperature in much of the US in summer time you are forced to really confront heat flow, exposure time and total heat capacity of the phase change medium. Which would be true for any phase change material solution. We could probably get time-temperature profiles from somewhere, maybe just by mailing a logger around. But we would still end up with a statistical solution. Do we want to protect against a 3 sigma heat event, or 5 or 8 sigma?
-
You'd have to have everything in a well insulated box for phase change media to do any good. 32 C is too low to be useful in the summer, but it would work for protecting shipments in the spring and fall. We do get days when the temperature in the back of a truck would get too high even in the early spring and late fall. But that would only be during the day, so you'd just need to be able to hold the temperature for 8-10 hours.
It might actually be a good idea to have some small, cheap to ship temperature loggers to send ahead prior to shipping a high quality reference.
-
You'd have to have everything in a well insulated box for phase change media to do any good. 32 C is too low to be useful in the summer, but it would work for protecting shipments in the spring and fall. We do get days when the temperature in the back of a truck would get too high even in the early spring and late fall. But that would only be during the day, so you'd just need to be able to hold the temperature for 8-10 hours.
It might actually be a good idea to have some small, cheap to ship temperature loggers to send ahead prior to shipping a high quality reference.
Yeah, shipping a temperature logger is a great idea. An attiny with a watch battery should be pretty light weight.
-
Hello,
the question is: what harms more cold or hot?
I would go pragmatic in this case:
- put the LTZ into the freezer + in the oven
and check the deviation against previous values
Then you would need only a temperature logger to see if these limits are maintained during transport.
If the cold phase is the problem the question is wether a phase change material
or a battery for heating adds less weight to the transport.
with best regards
Andreas
-
None of the questions can be answered meaningfully without a temperature profile. The integrated heat flow above the desired high temperature determines the energy which must be absorbed in the combination of thermal mass and phase change. The integrated heat flow below the minimum desired temperature is similar, but there is an option to use a battery to story energy instead of phase change material. The following clip from an article suggests the nature of the problem.
The effect of temperature exposure during shipment on a commercially available demineralized bone matrix putty
Authors
Authors and affiliations
Mark SchallenbergerHelena LovickJalane LockeTodd MeyerGregory JudaEmail author
1.
2.
Open AccessArticle
First Online: 25 August 2016
714
Downloads
Abstract
During August and September of 2013, temperature data loggers were shipped to and from an AATB accredited and FDA registered allograft tissue processing facility in Belgrade, MT (Bacterin International, Inc.) to five warm climate cities (Dallas, TX, El Paso, TX, New Orleans, LA, Phoenix, AZ, and Tampa, FL). Shipping data acquired from 72 independent shipments were analyzed to generate an assessment of temperature exposure, shipment times, and shipping event durations experienced during routine distribution. Overall the packages experienced an average temperature of 26.2 ± 2.3 °C which mirrored the average external ambient temperature of 25.8 ± 3.0 °C. However, temperature spikes above 40 °C were frequently observed. The data from the model shipments were extrapolated to provide a worst-case high temperature spike of 52.9 °C for 12 h and 14 min. Multiple lots of a commercially available demineralized bone matrix (DBM) putty (OsteoSelect® DBM Putty) were subjected to continuous heating at 50 °C, to multiple worst-case temperature spikes, and to multiple freeze–thaw cycles to assess the effects of these temperature extremes on the handling and osteoinductivity of the allograft tissue. Five weeks of continuous exposure to 50 °C and 12 simulated worst-case one-way shipments did not adversely affect the handling characteristics or the in vivo osteoinductivity of the product.
The attached graph from the same article gives at least a first guess at what to expect.
-
Another source of bounding information for this problem is Mil-Hdbk-310 (Widely available on line). It is a compendium of worst case environmental data world wide, and definitions of diurnal temperature cycles for a variety of locations and probability of occurrences. May or may not relate to conditions observed during the shipping cycle, but at least provides some basis for evaluating proposed solutions. At any rate it is one stop shopping for a lot of geeky weather data, with references to chase down even more detail.
-
the question is: what harms more cold or hot?
I would go pragmatic in this case:
- put the LTZ into the freezer + in the oven
and check the deviation against previous values
I don't like uncontrolled freezing, so I set my TECbox (have two LTZ-based refs right now) to +10.000C and programmed slow excursion to +20C and back to +10C.
Will see how it goes :). Right now reference value is +9.9999848V at +10.000 (+/-5mC) after overnight soak.
-
Well a little white box arrived today , it is currently hooked up to my 10v precision supply. My precision 50v supply is currently occupied(it has been for over a year... supplying 12v to an ocxo) I can easily swap to that though.
Looks like it survived shipping alright.
-
I ran some test after leaving the 2 HP meters , and the lambda power supply powered up for a day. The Vpsu is set to 10v(9.999(8/9)X) There is no gpib , so I relied on the meters internal stat functions(the 34401A is disappointing in that regard) the ac has been off , so 12C DeltaT. using the dual Pomona banana connector( they are a tight fit on the 3456) , with TiN's safety banana plug interface(cat5 wire + Q-Tips) .
Onto the measurements:
34401A:
Min- 7.045648 , Max - 7.045690 , most readings were 7.0456(6/7)
3456A:
Min - 7.04562 , Max - 7.04566 , Mean - 7.04564 ,
Variance - 0.062449E^-9 , std dev - 7.9025uV , count 63.264K readings.
Next I will power down the 34401A , turn on the AC(hopefully more stable temperatures will follow , and hook the ltz up to the pls50 set to 12v.
-
Appears your 3456A is slightly more stable vs the 34401A in your environment.
-
Ran it at 12v , with the doors open to the hallway
Min- 7.04561 , max-7.04563 , mean- 7.04562 , var-0.0251 , 5uV std dev.
So a bit more stable.
Other than switching psu's it hasn't changed any. Even going from 10v>>12v yielded no voltage shifts.
-
Looking into using the ratio mode with my 3456A , I fired up my mv106, I set the knobs to read the same as the ltz output , and the ratio was rather far off(1.003xxx). To get it reading as close to one took getting my mv106a far from its equal set point :?
-
You could also set the MV106 equal and then measure the difference between them. Connect the negative outputs together and connect both positives to the meter.
-
the question is: what harms more cold or hot?
I would go pragmatic in this case:
- put the LTZ into the freezer + in the oven
and check the deviation against previous values
Hi,
It might also be about how often it sees a particular temperature, rather than just whether it's high or low. Testing for that (unless there is already data on it available somewhere) would likely be a long term thing.
Deliberately subjecting them to a wider range of temperatures is not ideal, but it may still be a good option if you don't want to have to always send them in a box the size of a fridge.
There's still the problem of what to do to minimize issues like hysteresis when you fire them up again, but assuming that it's just the LTZ that you need to work on, adjusting its temperature should not make the whole box much bigger than it would need to be otherwise.
-
I've ordered this Mini Templogger:
https://shop.insidegadgets.com/product/mini-temp-logger/
(https://shop.insidegadgets.com/wp-content/uploads/2018/05/IMG_4407.jpg)
But it's from Australia, so it might take some time.
In the moment I'm in contact with ap from ab-precision to calibrate some of my ltz1000 (within the possible). After that, I could send it to the USA-Cal-Club for going around and free disposal there as a second item, if cellular or an other member want to manage that too.
But for now, it's only an idea, for a second item. An other one of these could circle around in Europe.
Please tell me, what you think about it.
-
Ran it at 12v , with the doors open to the hallway
Min- 7.04561 , max-7.04563 , mean- 7.04562 , var-0.0251 , 5uV std dev.
So a bit more stable.
Other than switching psu's it hasn't changed any. Even going from 10v>>12v yielded no voltage shifts.
Hey Vgkid, any update?
-
Ran it at 12v , with the doors open to the hallway
Min- 7.04561 , max-7.04563 , mean- 7.04562 , var-0.0251 , 5uV std dev.
So a bit more stable.
Other than switching psu's it hasn't changed any. Even going from 10v>>12v yielded no voltage shifts.
Hey Vgkid, any update?
Not yet , haven't had any free time at home. :( :scared:
I will be running it against the mv106 using the 3456a as a null meter (actually a few mv of offset, unless anyone has a different idea) in a few days. Then I will ship it out.
I have yet to power it down for more than 10 minutes.
-
Just powered up the mv106, so I can run the null test shortly.
Did another stability test similiar to the last, except the results are better :D .
Min:7.04563, max:7.04564 , avg:7.04563 , 59968 readings , 0.007233e^-9 var , 2.69uV std deviation.
-
Ran it against the mv-106 as a null, with a constant offset. Lets just say that this was extremely sensitive to anything... Makes me really want to repurpose my L&N linear amp.
min(-100.0142)max(-99.99937)avg(-100.0064)var(0.016487)stddev(4.06uV, still not bad)
will run another test before shipping out either Tuesday, or Weds.
As an aside anyone want to buy some vhp4(1r , unknown tolerance) resistors
-
I will ship out tomorrow, last weeks tropical storm delayed it a week.
Last call on the resistors.
-
I dropped it off at the post office a few minutes ago.
-
@ CellularMitosis , has the ltz1000 box arrived yet , if not , it shouldn't be too much longer.
In other news , my box-o-resistors came today.
-
Interesting. I didn't know resistors came in that packaging. At first, I thought they were transistors. Do people mount them to heat sinks or something?
-
Interesting. I didn't know resistors came in that packaging. At first, I thought they were transistors. Do people mount them to heat sinks or something?
Correct , they are precision 10W (max heatsunk) or 3w in air resistors. They are 1 Ohm/.1% vhp-4 resistors
-
Ah, yes. Makes sense. I'm just used to seeing the wire wound power resistors with fins.
-
@ CellularMitosis , has the ltz1000 box arrived yet , if not , it shouldn't be too much longer.
In other news , my box-o-resistors came today.
Not yet, hopefully tomorrow! :D
-
Heyo!
The ref arrived last night and I had it running today. Surprisingly, it looks like it has drifted up by a few microvolts since last month:
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1516537/#msg1516537 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1516537/#msg1516537)
I know I said I would follow a random participant order, but I'm going to violate that rule and send it off to kj7e next, to take advantage of his recent calibration ;D :-DMM
I need to get a few more of these into rotation! We've got a lot of participants this time :)
-
I know I said I would follow a random participant order, but I'm going to violate that rule and send it off to kj7e next, to take advantage of his recent calibration ;D :-DMM
Well, that sounds like a good exception to the rule.
-
I know I said I would follow a random participant order, but I'm going to violate that rule and send it off to kj7e next, to take advantage of his recent calibration ;D :-DMM
Well, that sounds like a good exception to the rule.
And FWIW, take me out of the rotation for the first two weeks of September.
For obvious reasons, it might be a good idea to let cellularmitosis know of your vacation plans... (privately, even though I just violated that idea).
-
I'm fine with the idea of randomizing the order of the list. It might be a good idea though to pick the order of the list now so that participants have a rough idea of when their turn is. This can help us to be prepared and more efficient during our turn.
-
Even if you don't pick or publish the entire list now, it would be good to have a month or two of warning, both for preparation on our end, and to give time for feedback to you if we aren't going to be available. Obviously all of our plans change from time to time, but most of us do have some major plans of some sort.
-
Even if you don't pick or publish the entire list now, it would be good to have a month or two of warning, both for preparation on our end, and to give time for feedback to you if we aren't going to be available. Obviously all of our plans change from time to time, but most of us do have some major plans of some sort.
That's a good point, but tough considering an unknown turn-around time for each tester.
Perhaps just the list, so one would know they are next? I'm for example gone about 40% of the warm months, so at least some flexibility needs to be considered.
-
Good ideas / concerns around scheduling -- I'll get the list made and published.
-
But then you will not have surprise guests, :-DD.
And I'm still waiting on resistors. :=\
-
This is a good point :). There might have to be a few more rule violations
-
Include some entries marked "RESERVED" or "? ? ? ? ?" to enable surprises, sudden schedule changes, and future rule violations.
-
Speaking of participant order, I will be away on foreign travel from the last week of June until the first week of September. Still able to read and post, but no access to my equipment while traveling.
-
A continually updated list would be a good idea because things can change quickly and unexpectedly. Ten days ago I was putting the final touches on my first Dr Frank board, which was looking good, when I decided to clean all the remaining flux off the board before permanently mounting it in the enclosure. After adding a transistor buffer for the 10V output, making a few changes, and selecting the trim resistors, it needed cleaning. Part way through the cleaning I heard a click from my power supply and suddenly realized I hadn’t disconnected the power before mucking with the board and the click I heard was the supply going into current limit. I checked and the 7 volt output from the LTZ1000A was now 12 volts. DAMN!
My second board is looking good and I hope to have the replacement LTZ1000A for the first board in any day now. I’ll be more careful with this one and both boards should be ready by the time my turn comes up.
-
A continually updated list would be a good idea because things can change quickly and unexpectedly. Ten days ago I was putting the final touches on my first Dr Frank board, which was looking good, when I decided to clean all the remaining flux off the board before permanently mounting it in the enclosure. After adding a transistor buffer for the 10V output, making a few changes, and selecting the trim resistors, it needed cleaning. Part way through the cleaning I heard a click from my power supply and suddenly realized I hadn’t disconnected the power before mucking with the board and the click I heard was the supply going into current limit. I checked and the 7 volt output from the LTZ1000A was now 12 volts. DAMN!
My second board is looking good and I hope to have the replacement LTZ1000A for the first board in any day now. I’ll be more careful with this one and both boards should be ready by the time my turn comes up.
Ouch!! I've done something similar awhile back. I desoldered the output and power wires after a test, while the power was still on lol. Thankfully the LTZ seemed to have went unscathed.
-
I think, it's probably too expensive to send something to germany back and forth (heard something like 60-80$). So, my participation can't be not more like quiet rhetorical, after spending one of my refs to cellular USA Cal Club (I hope, I still get it on the line, next time :phew:). So, please take me out of list. Good luck to the club :-+
-
I think, it's probably too expensive to send something to germany back and forth (heard something like 60-80$).
It mainly depends on the weight. And the service that you use.
from here to US it is 3,70 Eur below 500g (if you send it as letter) + 2,50 Eur for tracking (Einschreiben)
or 7 Eur below 1000g. (+ the tracking).
Of course the value of the content should be below the customs limits otherwise there are customs fees.
From US to D I payed between 7.10$ and 23.50$ for resistors from Edwin depending on USPS service.
So it is somewhat more but not 60$.
with best regards
Andreas
-
The other night I threw the ref into a small styrofoam ice chest along with the temperature logger and a tupperware container full of hot water. A temperature controller would be better, but this was something I could do quickly in the time I had available. The ref was hooked up to my Keithley 2015 (the same I've used in previous measurements for this ref).
Interestingly, it looks like the ref by itself has a negative temperature coefficient (in this temperature range). Previously, when I had adjusted the aircon (and thus measured the combined tempco of the ref and my meter), I had measured a positive tempco: https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1516561/#msg1516561 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1516561/#msg1516561)
The ref is packaged up and I'm about to run by the post office to send it off to kj7e (I hope they are still open on Saturday past noon!).
-
Looking forward to it. I have a short list of tests I want to run. 0.1-10Hz noise, 1 hour and 24 hour stability and comparison to my references at a steady 23C, TC sweeps from 10C to 40C and back, another stability test at 23C after the TC sweep, loading the output at 10G, 10M, 1M and 100K effects, maybe add in some EMI tests with shielded and non-shielded test leads.
-
I think, it's probably too expensive to send something to germany back and forth (heard something like 60-80$).
It mainly depends on the weight. And the service that you use.
from here to US it is 3,70 Eur below 500g (if you send it as letter) + 2,50 Eur for tracking (Einschreiben)
or 7 Eur below 1000g. (+ the tracking).
Of course the value of the content should be below the customs limits otherwise there are customs fees.
From US to D I payed between 7.10$ and 23.50$ for resistors from Edwin depending on USPS service.
So it is somewhat more but not 60$.
with best regards
Andreas
That sounds better.
My info was from ups. They told me, the cheapest from germany to USA, value of goods 10€, is 'UPS Expedited', plus fuel surcharge 15€, plus private delivery, incl., something about 90€. (Send it as letter with this outer dimension like a packed Hammond 1590B not possible) :wtf:
-
A international letter can also be in a card box (not only a jiffy bag)
(maximum size l + w + h = 90 cm)
https://www.deutschepost.de/de/produkte.html (https://www.deutschepost.de/de/produkte.html)
-
A international letter can also be in a card box (not only a jiffy bag)
(maximum size l + w + h = 90 cm)
Will ask Deutsche Post for that.
But, that also has to be done by customs. As i understood, handling flats for that are the expensive part. I don't want to cellular, that he suddenly has to pay customs and other costs.
-
When I shipped resistors to Andreas it wasn’t that expensive, I don’t recall the exact amount but it was deifninitely under $20. At the rate this club is going, this is a pretty cheap hobby, so don’t worry about it Hwj-d :)
-
Looking forward to it. I have a short list of tests I want to run. 0.1-10Hz noise, 1 hour and 24 hour stability and comparison to my references at a steady 23C, TC sweeps from 10C to 40C and back, another stability test at 23C after the TC sweep, loading the output at 10G, 10M, 1M and 100K effects, maybe add in some EMI tests with shielded and non-shielded test leads.
Sounds like a great set of benchmark tests for the ref!
-
When I shipped resistors to Andreas it wasn’t that expensive, I don’t recall the exact amount but it was deifninitely under $20. At the rate this club is going, this is a pretty cheap hobby, so don’t worry about it Hwj-d :)
I've been careful with this, after arrow, at this time as the ltz1000 had elsewhere delivery problems, send me one with dhl in a 'huge' packet. Dhl demanded customs duty and handling flat at delivery time, that more as doubles the price. I didn't accept the part with that delivery costs. After that i saw in online delivery protocoll, that dhl "parked" that package for more than one month in their hand-over stock, before they give it back to arrow. Luckily i got my money back from paypal.
Edit: But with ~20$ i'm in again. Thank you cellular. :scared:
-
Try Mouser next time. It has free shipping over 50USD. I'm in Croatia, so pretty much same as Germany...
They ship some from EU, and some from USA, but no difference for you. They don't have LTZ1000 on stock now but they will at the end of July...
Regards,
Sinisa
-
Try Mouser next time. It has free shipping over 50USD. I'm in Croatia, so pretty much same as Germany...
They ship some from EU, and some from USA, but no difference for you. They don't have LTZ1000 on stock now but they will at the end of July...
Regards,
Sinisa
Hi Sinisa,
yes, Digikey the same. They need 3 days to deliver from Thief River Falls to me. Big german distributors haven't completet the package in this time... :-//
-
Package arrived this afternoon. First test was a power on warmup plot. The reference had been sitting in the lab at 23C for about an hour, then powered up to see how long it would take before it settled down. First screen shot is 1 hour, so it took about 30 min for things to warm up in the tin enclosure. Second screen shot is a 5 hour plot, my wife turned off the room A/C around the 13K sec mark, you can see how the plot stabilized some. Tomorrow I will have to see if this is the reference or my meter showing the thermal sensitive.
Off the bat, I am measuring ~5ppm lower than cellularmitosis.
-
Looks like I don't need to worry about my multimeters being too drifty.
-
Excited!!! ;D
-
Newark is having a sale on some precision resistors that may be of interest to club members (or anyone, really)...
on their site search for "bargains" and filter (very bad user experience, slow site) or try
"bargains hpz" or "bargains alpha electronics hc series" to bring up just the good ones. I also bought some cheapish ltc5400 networks although those are not so much of a bargain.
Have fun,
Randall
-
They only seem to have the 9k + 1k LT5400 when I looked just now.
With the Linear site and direct-order shop shut down, basing my big DAC-centric project around 10k LT5400 is looking like a less good decision ...
-
Sorry for the delay,
Measured on a recently calibrated DMM7510: https://drive.google.com/file/d/13cKhjYruHux91m3_qTZ2RLhpINJhBPT-/view (https://drive.google.com/file/d/13cKhjYruHux91m3_qTZ2RLhpINJhBPT-/view)
It appears either my DMM7510 has drifted up or my 10v reference down by about 0.4ppm since last month.
Before and after temp sweeps, settled within 1uV, final measurement at 23.0C = 7.045,641,29
The Cal Club LTZ1000A TC is -0.08ppm/Deg C
Measured noise is very low, ~820nV P-P and 141nV RMS (divide the scope displayed value by 1000, so 1mV = 1uV)
-
:-+
Looks like the RMS 1/f noise measurement lines up with my own (though my p-p values seem to differ): https://github.com/cellularmitosis/logs/blob/master/20180324-ltz1000-1f-noise/README.md#setup-8-ltz1000-11
-
Quick logistics note: martinr33 reached out to me when the ref was about to embark to kj7e. He's got some pretty high-spec gear and believes he is sub-ppm at the moment. It turns out he's a stone's throw away from kj7e over in Cali, so he asked if the ref could make a stop at his lab on its way back to me. I think this makes sense -- minimize the travel/time between what are possibly our two best data points so far.
So the ref should be in martinr33's hands in a day or two! :-DMM
-
Sounds good to me. Looking forward to seeing how the data compares.
-
My measurements are done. My goal was to compare with a known 732a.
My conclusion is 7.04564563V as of July 1, 2018
Compare to KJ7E 7.04564129
Difference of about 0.6 ppm - (corrected from 6ppm) thanks to CalMachine's obervation. The traveling standard does settle out very quickly. Almost surprisingly so.
Here are the details
3458a reading of 732A 10V standard 10.0000565 V
Actual standard voltage verified 1/7/18 10.000050 V
Scale factor 0.99999935
Measured voltage 7.04565021 V
Scaled Voltage 7.04564563 V
Standard Deviation 150 nV
Temperature 23C
My 732a was confirmed today July 1 2018 against two other 732as and a 732b. My 732a is not moving from last time we checked it. The 732b is checked regularly against a bank of very stable 732as maintained by a friend of mine. (actually, none of the 732as had moved, but the B had shifted a little). I did not do any lead reversal to manage offsets, but I don't expect that to make much difference. I ran autocal before the readings.
I'll have the box in the mail back to home base tomorrow. Thanks to KJ7E and cellularmitosis!
(I just took one last look at the setup. The temp is now 27C, so 4 degrees up. The unit is reading 7.0456506, so up 0.6ppm - but with no autocal. Point is, not much movement compared to the difference in readings).
-
My measurements are done. My goal was to compare with a known 732a.
My conclusion is 7.04564563V as of July 1, 2018
Compare to KJ7E 7.04564129
Difference of about 6 ppm - much more than I was expecting. The traveling standard does settle out very quickly, thouugh.
That is 0.6 ppm.
-
^ What he says. 4.32uV at 7.0v is ~0.6ppm. Not too shabby at all.
-
Well, that was a quick fix...
And if you figure your meter drifted by 0.4ppm, that leaves the gap at 0.2ppm. Now, that's with a brass connection in there, and I did not null out my 3458a - but I think that the gap would drop another uV if I did that.
I'll correct the original post to reflect this conversation.
-
Hi Martin,
There should be a set of lead included which terminate as bare copper wire. Couuld you also take a measurement with the copper wires clamped under the binding posts of the 3458A? I'd e curious to see the difference.
-
Also differential measurement vs known 732A will be much better Sdev/resolution than direct DCV measurement.
-
OK, so it's warm out there. I am taking these readings by eye, but the setup is very stable - +/- a couple of counts once it settles. No surprise given the 150nV standard deviation.
With the bare copper - the last three digits switch from 480 to 490 if I reverse the connections at the meter (that's 48.0 to 49.0 microvolts)
With the nickel-over-brass banana jack - same result, once it settles. Doesn't take but a minute or two, it is warm out there (27 degrees). Off the bat, maybe 475 - 495 rather than 470 - 480.
I also did a quick zero test on the meter with a short, no surprise 1uV (10 counts) of offset (which matches the difference from reversing the cables, which is why you should always take two readings, one forwards and one reverse voltage.
Conclusion: the non-thermal connectors aren't making a significant difference if you keep air currents away, and you let them stabilze.
My garage is thermally variable, electrically noisy - but it is not drafty!
(I live in sight of some big TV towers - even my garage door opener will only work from about 25 feet from the opener, which means 10 feet from the door... so everything I do tends to be twisted pair).
Now, if you really want to see voltages move, try a nanovoltmeter.
-
"Also differential measurement vs known 732A will be much better Sdev/resolution than direct DCV measurement."
Yes,I should have done it that way!
However, the SD is only 150nV, and I would still be on the 10V range because of the 3V difference. The whole setup seems surprisingly quiet.
The K2002 could do this on the 3V range. But I have not figured out how to do an easy recal on that unit.
(and as a complete aside, the K2002 is a lot noisier than the 3458a, but seems more temperature stable - and the noise is random so it can be averaged out)
-
Thanks for the follow-up!
With the nickel-over-brass banana jack - same result, once it settles. Doesn't take but a minute or two, it is warm out there (27 degrees). Off the bat, maybe 475 - 495 rather than 470 - 480.
The dual banana jack is the Pomona 4892. The body which the wires screw into is brass, but the banana plug spring is copper. From the datasheet, the 4898 is nickel plated, while the 4892 is gold plated, but it is a bit ambiguous whether that means the 4892 is both nickel plated and then gold plated, or if the gold is directly plated onto the brass and copper.
-
Nice results, Martin, especially with the warm temps recently.
-
Gold needs a diffusion barrier to be plated over copper. Nickel is common, silver also works.
The Viborg bananas are gold over silver over copper, an give good results.
For these banana jacks, the issue is the big chunk of brass that makes up the body. You can get a good thermal differential across the two terminals that takes a while to settle.
-
The ref arrived this weekend, but I just hooked it up (got distracted over in the ovens thread ::) ).
Initial reading is 7,045,697uV to 7,045,700uV at 21.3C. Looks like it is trending slightly downward as it warms up.
Can't wait to throw this thing into a 25C oven tomorrow!
I had wanted to coincide this post with an announcement of a new PX-ref board, but unfortunately there may be a bug in it :( I found a stock of old OHMTEK CERDIP chips which have enough resistors in them to serve as R2 through R5 of an LTZ circuit, they are hermetically sealed, low tempco, just under $10 from taobao. I designed a board around these and I was hoping to give them out with lowimpedance's LTZ1000 chips (the user would just need to supply a 120R resistor), but the first one I soldered up latched the LTZ heater control fully on :palm: so this board may not be ready for prime time yet. I'll keep you all posted!
-
I see noone cared enough to upload data to single spot, so my regrets about current 'CalClub' format and lack of transparency stays :)
-
I see noone cared enough to upload data to single spot, so my regrets about current 'CalClub' format and lack of transparency stays :)
I created a directory on ftp.xdevs.com;
/CalClub/USA/Round_2/PX_Vref/kj7e
And will upload my data later today. Data uploaded.
-
OK, time to stop fooling around with ovens get this ref on its way again!
Here's a random ordering of the rest of the participants in round 2:
nikonoid
Svgeesus
dr.diesel
hwj-d
Vacuuminded
flittle
bitseeker
orin
GEOelectronics
rhb
RandallMcRee
ArthurDent
edavid
CalMachine
Conrad Hoffman
VintageNut
vindoline
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
If you don't see your name in the list, I'm sorry I forgot you! Just send me a PM.
Probably the biggest issue right now is how long it will take to get through this list with only one active ref. There are some things in the works to address that (TiN's ref is in progress, I'm working on a number of additional refs, and soon ~6 others will get lowimpedance's LTZ chips and may be able to participate), but all of those will take time to stabilize as well, so things may be slow for a bit longer.
As they say, the best time to plant a tree was 20 years ago, the second best time is right now!
-
Wow, I didn't realize that we had grown so much. Yay, cal club.
Even if it does take some time, better that than no club ref. :-+
-
Recently acquired a new instrument in order to feel the love from the cal club--an Advantest R6581T (8.5 digits!).
This was mislabeled on ebay as an advantest r6518t...but obviously from the pics was the 6581. Dim display...I think it was still a steal.
Just trying to learn how to drive it. It has easy calibration steps, although I'm pretty sure I goofed somewhere. If anyone wants one there is still one out there on ebay.
Randy
-
Nice catch - these are good units. I did replace the battery on mine. But it was still at 3.7V after 20 years. Japanese low power SRAMs from that era are remarkable.
Next step is to clean it up, and bed it in. And then see if you can measure its drift.
and then after that comes a 732a 10V standard. Then you need a 10k standard. I built a 10k box after kj7e's pattern, it is quite good.
-
Wow, that's quite a catch. Way to go, Randall! :clap:
Interesting that the cal sticker has ten years between last and next cal.
-
Yeah, seems normal for the Advantest? You can see the same sticker in TiN's writeup here:
https://xdevs.com/fix/r6581t/ (https://xdevs.com/fix/r6581t/) It's specs aren't as impressive as a 3458 so I don't think that ten years means that it has any "special powers".
There was a bit of anxiety upon first powering it on. It comes on without the display enabled and the trigger off. So had to figure that out. (Let me come clean--I just pushed buttons randomly and it started working. I cannot lie.) Plus, of course set to rear not front panel...but it was pretty easy in the end to get the usual: 100 PLC, 8.5 digits, auto off, etc. Unlike my other meters, GPIB can toggle between the front and rear panel, so I can put that to good use for monitoring two sources.
I have a Fluke 731B for a 10volt calibration but now I suddenly need a 10K standard resistor. Or have someone measure mine, perhaps?
-
Be sure to read thru 6581 repair thread (https://www.eevblog.com/forum/metrology/advantest-r6581-8-5-digit-dmm-mini-teardownrepair/). Many came to conclusion that 6581(T) is quite a disappointment and declared specs are, well, too optimistic. Hopefully you paid under 1k for it :)
-
Be sure to read thru 6581 repair thread (https://www.eevblog.com/forum/metrology/advantest-r6581-8-5-digit-dmm-mini-teardownrepair/). Many came to conclusion that 6581(T) is quite a disappointment and declared specs are, well, too optimistic. Hopefully you paid under 1k for it :)
Uhh no, over $1500. But, I did read through the thread and I'm not sure where the disappointment comes from? The 3458 goes for +$4000 and the Keithley 2002 is in the same stratosphere. So in the $2000 ballpark what would else be possible? Is it considered to be a 7.5 digit meter?
-
Glad to see this is still going on.. nice going cellularmitosis!
I haven't done too much on my metrology lab in a while. Just been ageing few of the reference I built since the last time. Time to fire up the instruments and do some new measurements! My K 2015 bit the dust and will need repairs, but my 3458a and my 8846A are still going strong.
-
Welcome back Muxr. Metrology like wine, only better (and more expensive judging from ebay) with time :popcorn:
And I'm still waiting on my February Vishay resistors order....
-
Welcome back Muxr. Metrology like wine, only better (and more expensive judging from ebay) with time :popcorn:
Thanks TiN!
was just admiring the work you did on your F5720A calibrator, nuts! :)
And I'm still waiting on my February Vishay resistors order....
Wow still? I finally got another LTZ1000A like 6 months later or something, will need to build another LTZ1000 reference.
-
I got the new ref last week. Here are first impressions:
To start with I powered it with Keithley 2450 @ 12.00V and measured with Keithley 2002. 2002 was calibrated by CalMachine 9 months ago and read 0.1ppm high on 10V range at the time of calibration. It does not get much better than this. (Thanks CM).
I used 2450 to precisely track power consumed by reference. On the start it took 25mA and then consumption slowly drifted down to 22mA, where it reached thermal equilibrium. After getting stable it would fluctuate very slightly with room temperature.
Unfortunately I could not get enclosed temperature monitor to work. There is some sort of issue with Arduino drivers on my computer. My lab temperature is very stable though varying less than 0.5C around 21C.
Average reading from K2002 was 7.0456737. This is about 4ppm high comparing to other results reported. I assume my Keithley 2002 could have drifted that much in 9 months.
Noise levels coming from K2002 were quite high. At times reaching 0.8ppm or about 5.5uV peak to peak. At this time we should consider noise coming from 2450 and also large amount of equipment concentrated in a same space.
For the second log, I used 2 of 6V SLA battaries to power reference, used 3458a for logging (TiN logging parameters: https://xdevs.com/article/hp3458a_gpib/ (https://xdevs.com/article/hp3458a_gpib/)) and relocated reference and 3458a away from other equipment. This time noise improved a lot to about 0.08uV peak to peak over a course of few minutes. I would say that this is a remarkably low.
I am considering using Datron 4910 (buffered out) with SR1010 (1kOhm per step) as a voltage divider to create very stable 7.0000V and then use that in series opposition with the PX reference to measure noise directly against Datron. More results and upload of files is to follow.
Cellular, this is an excellent reference, especially considering the size and weight.
-
When do we get the intercomparison report with all the EN-values and stuff? 8)
-
When I measured the unit, I saw only about 150nV of noise on my 3458a. I used 100NPLC and the math average function over a few minutes. I have noticed that the Keithley 2002 can be very noisy (700nV SD), but it seems a bit less drifty with temperature than the 3458a. It also averages out nicely if you put the filtering on.
-
When I measured the unit, I saw only about 150nV of noise on my 3458a. I used 100NPLC and the math average function over a few minutes. I have noticed that the Keithley 2002 can be very noisy (700nV SD), but it seems a bit less drifty with temperature than the 3458a. It also averages out nicely if you put the filtering on.
Keithley 2002 usually auto calibrates to temperature on every reading (it can be disabled). In contrast 3458a requires explicit call to ACAL DCV to adjust to temperature change.
-
Keithley 2002 usually auto calibrates to temperature on every reading (it can be disabled).
I'm not sure that is right. It does autozero, but it's not related to temperature. At least I see clearly visible tempco with temperature excursions, which would be corrected otherwise if it does temp correction.
-
Keithley 2002 usually auto calibrates to temperature on every reading (it can be disabled).
I'm not sure that is right. It does autozero, but it's not related to temperature. At least I see clearly visible tempco with temperature excursions, which would be corrected otherwise if it does temp correction.
I am not sure how effective is this autozero, but this is what I found in the manual:
AUTOZERO: In order to maintain stability and accuracy
over time and temperature, the Model 2002 intermittently
measures internal voltages corresponding to offsets and
gains of amplifiers. This process is known as autozeroing.
There are three types of autozero: synchronous, normal, and
off. The characteristics of each are described below.
• Synchronous (the default mode) is the most accurate,
but slowest mode. In this mode, each trigger causes
three A/D conversions: one for input signal, one for an
internal zero, and one for an internal gain. This mode
also yields a constant reading rate
It might be a cool test to log Keithley 2002 over range of temperatures with Synchronous autozero and with autozero Off.
-
I am sure TiN wouldn't mind using my 2002 with a reference and test the different settings. However, my A/C might change the room temperature too quickly.
-
The reference is back to Cellular. I am uploading files to TiNs FTP Server.
I logged the PX reference with Keithley 2002, then yet to be calibrated 3458a (noise assessment) and finally freshly calibrated 3458a (Thank you, Cal Machine).
Keithely 2002
Calibrated 11/7/2017
Calibration point 19V: 19.0000026
Log Date: 8/5/2017
PX Reference powered by Keithely 2450 at 12V
Average logged value: 7.0456737
Corrected for calibration: 7.0456747
HP 3458a
Calibrated on 08/06/2018
10V value was calibrated as: 9.9999992
Logged on 8/13/2018
PX Voltage reference is powered by 12V SLA battery
Positive logging average: 7.045661201
Negative logging average: -7.04566173
PX Reference value (corrected for calibration): 7.04566147
3458a just arrived from calibration and exhibited a bit of drift/erratic behavior. It seems to settle ever since, but the log that was taken right after it arrived can be off by as much as 0.5ppm.
It looks like my Keithley 2002 drifted 1.9ppm in 9 months.
-
Ref back in hand! I'll log some data and get it shipped back out.
At this point I have two other LTZ's which have about 3 months of ageing on them, so I can start to provide relative measurements between them, which should lessen the influence of my meter.
-
Keithley 2002 usually auto calibrates to temperature on every reading (it can be disabled).
I'm not sure that is right. It does autozero, but it's not related to temperature. At least I see clearly visible tempco with temperature excursions, which would be corrected otherwise if it does temp correction.
I am not sure how effective is this autozero, but this is what I found in the manual:
...
It might be a cool test to log Keithley 2002 over range of temperatures with Synchronous autozero and with autozero Off.
Perhaps I'd need to be more specific. K2002 AZ autocorrects for gain and offsets drift (synchronous corrects on every sample set, while normal only on range/function change). So if drift/offset error caused by thermals change, those will be corrected as byproduct, however this function does not actually take internal temperature measurement (K2002 does have a sensor though, available in TEMP mode, showing difference to TCAL) to correct the tempco of the parts used in signal path. So TCR of shunts/current source references for resistance is not removed/compensated for, it's still there. So in best case meter readings have tempco of the internal working standards involved for specific function/range. Hope this is bit more clear.
To stay on topic, CalClub's 10V reference going thru testing this week and tentative shipping in 2 weeks. Sorry that it took so long. :phew:
-
To stay on topic, CalClub's 10V reference going thru testing this week and tentative shipping in 2 weeks. Sorry that it took so long. :phew:
Awesome!!! :-+
-
Awesome! I'm actually quite glad I'm near the end, for one I'll be moving soon like in the next month, so that'll be a pita. Secondly I'm looking forward to seeing everyone else's findings so that I can compare and check that my measly stuff are at least within ballpark. As I don't really have the fancy DMMs and equipment yet to do thorough testing. Looking forward to seeing everyone's results and keeping a sharp eye open for cheap equipment on the bay of evil.
-
Yeah, moving's never fun. Hope the new digs are, though.
-
Is there any news on the club? This thread has gone very quiet lately. I think we're waiting for LTZ1000 ref from TiN to arrive?
-
Howdy volt nuts, I'm sorry I fell off the face of the earth. Work got crazy busy for a while, and then I got sick for several weeks, which I'm just now coming off of (immunosuppressants are no fun :( ). I fell into a bit of a mental hole, and hobby time got completely put on the shelf.
Time to pick back up and get the (7V) ref in circulation again!
-
No worries. Real life has a way of disrupting things. Glad you're feeling better!
-
CM, I'm glad to see you back and feeling better. Be well.
-
Got the ref boxed up and mailed to Svgeesus today! :-DMM
I'm sorry this took so long. I seem to be on the tail end of some sort of depressive episode. It was as if all of the things which usually brought me joy had no appeal at all. I feel like I've neglected this community during that time and I feel bad about that :--. Finally getting off my butt and getting (even a small thing) accomplished feels like a breath of fresh air!
In other news, I found some mold growing near one of my AC vents (the humidity in this apartment is typically 65%+). It appears the affected area was pretty small (just a few registers -- the coils and plenum look ok). Because I've already dealt with this problem in another apartment, I decided to invest in a dehumidifier. I must admit that being a volt-nut made this purchasing decision ver easy ;D Year-round regulated humidity means epoxy-packaged resistors and IC's will have less seasonal drift :-+
-
"Inch by inch, life's a cinch. Yard by yard, life's hard." — John Bytheway
Life has its ups and downs. Just work on those inches and you'll get there. Good to see you back on the forum, CM!
-
Got the ref boxed up and mailed to Svgeesus today! :-DMM
I'm sorry this took so long.
I'm also sorry. I got this last night, when I arrived back at my house. (I had planned to be back 8 days earlier, but life intervened).
Unpacked, all looks good, now re-reading this thread before powering it up.
Note that the reference is better than the LM399 in my (in cal) Keysight 34465A, so I am really testing my meter not the reference.
-
I uploaded this to the xdevs ftp server, directory svgeesus:
Operator: Chris Lilley (svgeesus)
Date: 10 Nov 2018
DUT: US Cal Club 2 PX reference
Type 1b: Direct measurement, calibrated equipment (in 1 year spec or better)
DMM: Keysight 34465A, instrument ID 869657
Cal date: 28 Nov 2017
Cal type: Accredited 17025 with uncertainties
Cal temp: 23.0C
Cal humidity: 23.0% RH
Cal procedure: 34460-90901 Truevolt DMM service manual, July 2016 edition 3
Cal inst: Fluke 5725A amplifier, ID 06-0090, NIST 12678290, Cal date 03 Aug 2017
Cal inst: Fluke 5700A/EP03 calibrator, ID 21-0140, NIST 11901768, Cal date 01 Dec 2016
Cal inst: Agilent/HP 33250A function generator, ID 407410002, NIST 12901089, Cal date 13 Oct 2017
Cal front panel zero offset, short, 10V range: as left 0.00000, min -0.00002V max 0.00002V, uncertainty 5.8uV
Cal front panel +10V, 10V range: as left 9.99999V, min 9.99966V max 10.00034V, uncertainty 35uV
Measurement details: front panel, 100PLC, HiZ
PX2 mean: 7.0456366V
PX2 lowest: 7.04563V
PX2 highest: 7.04564V
PX2 samples: 562
PX2 sd: 1uV
PX2 corrected: 7.045643V
ambient: 22.0C 33% RH (arduino, free air) 22.5 (k-type, above ref)
power: 11.94688V from supplied voltage regulator, fed with 15.0V from Tenma 72-8335A linear bench PSU
Equipment:
- Keysight 34465A dmm, instrument ID 869657
- Tenma 72-8335A linear bench PSU
- DM6801A temperature meter, K-type probe
- Cal club Arduino temp/humidity sensor
- Ian Johnston PDVS2 voltage source
Vreg was fed with 15.0V from Tenma 72-8335A linear bench PSU.
Output measured on Keysight 34465A, front jacks, 100PLC
Time series plot from cold shows that the regulator reached a stable voltage fairly rapidly.
Once warmed up (100PLC, HiZ, front jacks, T=21.7C, RH=32%:
mean 11.94688
lowest 11.94665
highest 11.94721
sd 115uV
ambient 21.7C
Next measured the reference, from room temperature, powered by the vreg, 100PLC, HiZ, front jacks.
I'm now considering using my Ian Johnston PDVS2 set to 7.0456V in opposition to the PX ref and measuring the difference. Note that the PX ref is more stable and less noisy than the LM399 in both the PDVS2 and the 34465A.
Once it was stable, measured the stability for 30 minutes.
-
Used my Ian Johnston PDVS2 set to 7.0456V, in opposition to the PX ref and measured the difference. Note that the PX ref is more stable and less noisy than the LM399 in both the PDVS2 and the 34465A. So really, this is more a stability measurement on the PDVS2.
Result: mean -69.5μV min -75.3 max -61.1 sd 2.4μV, n=1493.
Package posted back on Weds 14th, expected delivery Mon 19th.
-
I don't recall the exact order, but please skip over me till spring.
-
:=\ ... ... what??
(edit: that's me)
Hi Jason,
pm send. :)
Many greetings,
and relaxing holidays to all. :-+
-
:=\ ... ... what??
(edit: that's me)
Haha, sorry my friend ;) I had asked for hwj-d's mailing address, but he was "asleep" for a little while. Then, after not hearing from him, on literally the same morning that I decided to skip over him and mail the ref to Vacuumminded instead, he replied with his mailing address :-DD Don't worry, you're next in line! ;D
I took a bit of data before mailing the ref out to Vacuumminded. It was a bit noisy, so I added a Savitsky-Golay filter (thanks again to Mr. Hoffman for the tip):
https://github.com/cellularmitosis/logs/tree/master/20181219-k2015-cal-club (https://github.com/cellularmitosis/logs/tree/master/20181219-k2015-cal-club)
Edit: I had discovered that by the end of my data, my lithium ion battery had dropped below the minimum input voltage of the 15V regulator, so the supply was no longer in regulation. That might explain the dip at the end.
Edit2: the x scale is just shy of 24 hours (wed 11pm to thu 10pm).
Edit3: And here's the order of the queue: https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1680698/#msg1680698 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1680698/#msg1680698)
-
Much to my surprise, I received the package TODAY! :-+ Will be spending some time to go back over what has been done this round, and get the logging squared away. I did quickly shuffle things around and fire up the ref itself, and everything appears to be fine so far :) Letting the meters cook, they've been off for a few months.
-
Excellent! :-+ Glad to see some action on this project and thread. CM, isn't TiN prepping another reference for our club? Do you have any status update on that?
-
vindoline
CM, isn't TiN prepping another reference for our club? Do you have any status update on that?
Yep, there is "FX" 10V LTZ1000A-based reference for USA Club in cooking, it's on since October, and still need bit more time before final calibration test and shipping out.
I'll do my best to get it delivered in January'19. Sorry it's taking so long, LTZ references don't like haste. :phew:
-
TiN,
take a nap, meanwhile ;)
-
vindoline
CM, isn't TiN prepping another reference for our club? Do you have any status update on that?
Yep, there is "FX" 10V LTZ1000A-based reference for USA Club in cooking, it's on since October, and still need bit more time before final calibration test and shipping out.
I'll do my best to get it delivered in January'19. Sorry it's taking so long, LTZ references don't like haste. :phew:
Thank you TiN. You're generosity is humbling.
-
vindoline
CM, isn't TiN prepping another reference for our club? Do you have any status update on that?
Yep, there is "FX" 10V LTZ1000A-based reference for USA Club in cooking, it's on since October, and still need bit more time before final calibration test and shipping out.
I'll do my best to get it delivered in January'19. Sorry it's taking so long, LTZ references don't like haste. :phew:
Just tossing this out there, I would be willing to take some measurements and compare the "FX" to my references (recently calibrated at CalMachine's lab). Up to you and Jason.
-
Happy New Year, cal club members!
I've been struggling with trying to get the temp/humidity logging working with the included Arduino Uno, has anybody else with a Windows machine successfully used Jason's script, as found below? Using Python via Windows command prompt the script seemingly runs, (I am presented with a blinking curser and no ability to enter further commands), however no "log.csv" file is being generated.
https://gist.github.com/cellularmitosis/a509a6470c068fd0746a7c7b49bd9a90
Likewise running the below script to simply write serial data to the console, it gets no further than line 19 stating that "typeError: write() arguement must be in str, not bytes"
https://github.com/cellularmitosis/logs/blob/master/plotting-scripts/trivial-serial.py
If I change line 19 from "sys.stdout.write" to "sys.stdout.buffer.write" (as suggested on a python forum for a similar issue) the error no longer pops up, but again all that is presented is a flashing curser which cannot accept any further text/commands.
I am _not_ a code guy and after many hours of frustration only to get this far, am about to give up on the logging so that we can keep this thing moving. Serial data can be seen from the Uno via the actual Arduino program, so I don't think it is an issue with my system- Just my understanding of what I, or these scripts are doing wrong!
Regards,
Chris
-
Perhaps you could send Cellularmtosis a PM and ask for help? He used to be very active on the forum, but he seems to have other things going on lately.
I'm afraid I can't help, as I'm not a "code guy" either. Also, I use a Mac and Raspberry Pi for my logging solution. It took me a long time and a lot of tortured hack coding to get it working. I started with the Python scripts the Muxr posted a long time ago. Good luck
-
Yeah, we tried to get it debugged over email, but I’m out of town at the moment and don’t have a windows machine handy :(
I’m not sure what to make of the error about binary data.
-
I haven't seen/used the latest logging setup, but since it's Python, is the version of Python that's installed compatible with the code? I've seen some programs only work correctly with a particular Python version.
-
Ha, I just happened to think the same thing yesterday afternoon, and am waiting Jason's reply to an email. I just assumed everyone is using Python 3 instead of Python 2, since it was released over a decade ago. I'm not familiar enough to identify any difference via the code itself, so I'll certainly try downloading 2 and see what happens if nothing further is known before then.
-
Yeah, 3.x has been around for a long time, but a surprising amount of code still requires 2.x. I don't know the versions well enough to point out anything that would specifically cause the problem you're running into, but it was the first thing that came to mind. Hopefully, that takes care of it.
-
Just a quick update, the script is currently running via Python 2.7.xx. this was the main problem :-+ Will try to wrap this up as quickly as possible and git 'er back to Jason!
-
Thanks for the update, Vac. Glad that was it. :-+
-
It will be quite a while before my name comes up but I have recovered from my disaster back in June, 2018, where I accidentally destroyed my first LTZ1000 by forgetting to disconnect the power before making some wiring changes. My new reference looks good as far as I can tell with the HP3457A I have. The battery supply and charger are built into the case I'm using. No switches were used and just A.C. in, red/green bi-color LED for charging/charged and live indication, 2 sets of gold banana jacks on the front panel for both 7 and 10 volts, and 30 turn pot. Batteries feed the reference through a 12 volt regulator and will keep it powered for over 36 hours.
Everything is a compromise and I felt this was good enough for my first go around. After I use it for a while I may make some changes to the next one. I have another case and identical metal box for version II. Here's a couple of photos of the reference.
-
:-+
I hope you have air baffle over LTZ chip on both sides.
-
Even have a foam washer around legs on front side and flat foam square on back side between board and thin fiberglass insulation sheet under entire board.
-
Nice construction! Where did you source the steel box from?
-
Nice job! What resistor values did you decide to use?
-
The resistors I used for the most part were Vishay RNC90Z 2ppm tempco and bought surplus at a low price. I used 988R instead of 1K, 12K931 instead of 13K, 68K000 instead of 70K, whatever available values were close and I figured would work after reading a lot of post about choosing resistor values. The 10 volt adjustment pot is a 5K WW 30 turn miniature one with 2K5 in series and that combination is across an 88R68 ohm WW resistor. The pot and the 2K5 resistor are the only two precision resistors that aren’t really low tempco but where they are in parallel with the very low value 88R68 ohm resistor their effect isn’t that great.
The metal boxes I used that the Dr. Frank boards fit perfectly into were a couple I got with Motorola GPS boards where I needed the boards at the time and the boxes went into, let’s say, my ‘stock’ pile so they are an item that isn’t really available. Inside the steel box is a thin sheet of fiberglass board for insulation and the box is mounted on a piece of brown phenolic board which is mounted in the case. The case is from a scrapped HP bus extender and I got two of them for cheap just for the case and the power connector.
-
... The battery supply and charger are built into the case I'm using. No switches were used and just A.C. in, red/green bi-color LED for charging/charged and live indication ...
Very nice box.
Is the charger 110v in? Is it linear or switch mode? Does it charge 4S 18650 Li-ion (ca. 14.8v) which feeds the 12v-regulator for the ref? And you just added wires to the PCB in the plastic box for external LED indicators (charging/full)?
Do you (or Does anyone) know about an ebay-orderable plug-n-play linear charger for 4S batteries?
-
The charger I used is a 100-240VAC switcher designed for 4S1P 18650s and the four battery 4S1P holder has built-in protection circuitry. I removed the A.C. pins for the charger plug and mounted the plastic charger case inside the metal case everything is built into. I brought the A.C. leads out to connect through a back panel fuse holder to the line filter/IEC connector. I also brought the 3 leads for the red/green LED out and mounted the LED on the front panel. The charger LED is designed to show red when charging then green when it switches over at 95% charge and trickle CV charges at 16.8VDC. The green indication will then stay on until the battery voltage drops whether the charger is plugged in or not.
The protection circuit in the battery holder will disconnect the batteries from the load when the voltage drops below a certain point so the batteries can never be run flat and the green LED goes out at this point as well. The battery holder output and charger output are connected to the 7812 regulator I used because I have a buffered 10 volt output and that 12VDC gave me some headroom. I'm not concerned about using a switcher instead of a linear charger because I never plan to use the reference while it is plugged in and charging, so the type of charger makes no difference to me. Using the reference while it is plugged in could cause some offsets or noise in the output voltage due to capacitive coupling or whatever so I'm avoiding those problems by only using the reference on battery power.
-
Someone know, what happens to Jason?
Now 2 weeks ago, I sent him a pm, no reaction.
-
Oops, I must have missed that one, sorry!
-
I am officially submitting my request to join the CAL Club. I am very close to completing my 'mini CAL lab' in my home office and I am ready to take the next step.
-
Welcome aboard! The conductor (cellularmitosis) will be by to punch your ticket. ;D
-
Inverted18650 - "I am officially submitting my request to join the CAL Club. I am very close to completing my 'mini CAL lab' in my home office and I am ready to take the next step."
That sounds like your introduction into a support group. :-DD
You are among kindred spirits.
-
Glad to have you on board! Can you send me a direct message with your shipping address?
-
Hello everyone, my name is Chris, and I am a volt-oholic. I usually have my first glance at the DMM's shortly after waking up and again right after breakfast. Sometimes I try to hide my addiction from others, but feel like this may be a good place to share my stories of obsession. :-+
-
Oops, the ref arrived last week and was being held at the front office. I’ll get it measured tonight and send it on it’s way!
-
May I ask again very carefully if the USA Cal Club idea continues? Jason was last seen more than a month ago and actually wanted to send the ref to me.
-
Well, let's see where we are. In August of 2018, nikonoid had the ref. Then, Svgeesus got his turn in November. Next, would've been dr.diesel, but he requested to take his turn in or after Spring 2019. Then Vacuuminded got it in December when hwj-d missed CM's message. So, hwj-d would be next.
The remainder of the list for this round, after hwj-d, is as follows:
flittle
bitseeker
orin
GEOelectronics
rhb
RandallMcRee
ArthurDent
edavid
CalMachine
Conrad Hoffman
VintageNut
vindoline
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
Since it's almost Spring, dr.diesel will need to get re-inserted sometime soon-ish.
I sent CM a PM. ;)
-
Yes, i missed it shortly half a day, because i didn't see that dr.diesel postponed his appointment, in timely manner. In the meantime, however, this has been clarified with the order up to me for quite some time. Actually Dr. Diesel would have to line up again after me.
-
Sorry guys, I am flying back into town tomorrow. I had it packaged up and had intended to ship it before I left town but I didn't. I'll get it shipped out Monday morning.
I've been neglecting this venture :-[ . I don't like that reality but the truth is that my efforts have been focused elsewhere (I've been learning a new software platform for work, and I'm likely to be focusing on that for most of 2019).
Just putting out a feeler -- would anyone be interested in taking over the cal club? I'd be happy to donate the reference to the cause :-DMM
-
Sounds like cellularmitosis found a cure for voltnuttery! Please do not share the recipe :-DD.
-
Hi Jason,
I understand such time problems, didn't want to put you under pressure but only whether it goes on. It's a pity that you can't put your energy into this project so much anymore. But I'm pretty sure someone in the usa can be found to step in, or the list should become a self-runner by now.
All the best,
cu
:-+
-
Today, the ref reached me. :-DMM
Someone put 10$ into the package, thanks ... ;D ... no, i think that was actually meant for someone else who deserved it better.
A first look says, all ok. So i take a first measurement w/o detailed temp (more to ftp).
My 34461A is >100µV out now, so I take it with my DMM6500 wich is inside one year calibration.
-
Hehe, perhaps the $10 can become the calclub good luck charm :)
vindoline has graciously offered to take over the cal club. Vindoline, would you like to start a new thread?
-
Thank you for taking the reins, Vindoline! And thank you, cellularmitosis, for pioneering our cal club.
-
Thanks cellular for what you have done, and vindoline thanks for taking over. Let us know how to help.
-
Hi everyone, I don't see any need to start a new thread for this. I'm just here to help out cellularmitosis with the logistics of getting the reference moved around to the members. cellularmitosis has done an amazing job of setting the club up and getting things going. Thank you! Now take a well deserved break!
My first goal is to get the reference(s) moving around quicker to the members. I think this will re-kindle the interest and enthusiasm of the group. I'm open to any ideas or suggestions. My initial thoughts are to "recommend" that we try to limit each visit of the reference to 2 weeks or less. I would reach out to the upcoming recipients to make sure they were ready to use their slot. If not, no problem. Life happens. I would just ask them where they would like to be placed in the queue. What does everyone think?
-
Sounds good to me. Let me know when you get a sense for the timing so we can coordinate with you and be ready.
-
I don't see any need to start a new thread for this.
I agree, no need to spread entities.
vindoline, I guess I need a shipping address for new 10V CalClub ref. Please send me PM with details (and phone # for EMS).
Have it ready to go in next week or so. :-X
-
vindoline, I guess I need a shipping address
Me too, to ship it back :)
-
Well, let's see where we are. In August of 2018, nikonoid had the ref. Then, Svgeesus got his turn in November. Next, would've been dr.diesel, but he requested to take his turn in or after Spring 2019. Then Vacuuminded got it in December when hwj-d missed CM's message. So, hwj-d would be next.
The remainder of the list for this round, after hwj-d, is as follows:
flittle
bitseeker
orin
GEOelectronics
rhb
RandallMcRee
ArthurDent
edavid
CalMachine
Conrad Hoffman
VintageNut
vindoline
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
Since it's almost Spring, dr.diesel will need to get re-inserted sometime soon-ish.
I'm assuming that this list is correct? If anyone is missing, please send me a message
-
dr.diesel, if he's still interested, needs to be inserted back into the list since he postponed his turn. He's welcome to go before me, if desired.
-
Just my two cent, wouldn't it make sense to circle the reference inside one country before it moves to another? That could keep shipping cost lower and the risk of getting lost or damaged smaller.
-branadic-
-
dr.diesel, if he's still interested, needs to be inserted back into the list since he postponed his turn. He's welcome to go before me, if desired.
I agree. I'll reach out to him before we re-boot.
-
Just my two cent, wouldn't it make sense to circle the reference inside one country before it moves to another? That could keep shipping cost lower and the risk of getting lost or damaged smaller.
-branadic-
Huh? I don't have an address list from CM yet. I thought everyone was in the US for this club. If not, that's a great point.
-
Well, hwj-d has Germany's flag on his profile. I suppose the -d might indicate he's in Deutschland. Possible?
I second branadic's suggestion if we're actually going global. If that does happen, I guess it means the "USA" in "USA Cal Club" will then indicate the home base of the references. ;D
-
And I thought you'd already noticed that on my bad english... ;D
Yes, i'm german, living in the near from Schalke-Stadion (soccer, everybody should know ;) ), and befor I become an IT-Consultant, I'm working as a coal-miner here in the 'Ruhrgebiet'. Not a joke, but long ago. :P
I'm also for sending it back here in D first. It would also be much faster, more practical and, last but not least, cheaper overall, than sending it to the USA in a star shape there and back. But, of course, I'm also happy to follow the order originally agreed.
-
Btw, what actually speaks against an EU-CalClub?
That orange box in my first picture is full of VHP's ready for 4wire. Because these are overall relatively flat built mini binding-posts, I'm sure, there's a ref board with pomonas and extra voltage control, that fits in if I'm not fooled. A matching 4wire teflon cable (standard beryllium hollow-bananas > beryllium mini plugs) is already there. Wouldn't that be an ideal starting point for an EU-CalClub?
-
I haven't heard anyone being against an EU Cal Club. Maybe it just needs a few people to say "Yes!" and then get started. The Aussie Cal Club was in hibernation, last I heard.
-
Well, hwj-d has Germany's flag on his profile. I suppose the -d might indicate he's in Deutschland. Possible?
I second branadic's suggestion if we're actually going global. If that does happen, I guess it means the "USA" in "USA Cal Club" will then indicate the home base of the references. ;D
Ha! I'm so unobservant sometimes! Herzliche Willkommen! I have no problem with a global USA club, I just failed to realize that it had already happened. I have not yet gotten a club membership list from CellularMitosis, so I don't actually know where anyone is. Are there other people on the rotation list in Germany or Europe? That would change things from a practical sense.
-
And I thought you'd already noticed that on my bad english... ;D
I'm always humbled by how well people from all over the world communicate in English, and I'm alway embarrassed by my poor mastery of other languages. After 50 years I can still barely spell in English without a spell-check.
I'm working as a coal-miner here in the 'Ruhrgebiet'. Not a joke, but long ago. :P
I was a plumber, but then I gave that up to get a PdD in Organic Chemistry. I'm still not sure I made the right choice...
-
There's a joke,
Someone who speaks three languages is trilingual.
Someone who speaks two languages is bilingual.
What do you call someone who speaks only one language?
American :palm:
Geography does play a role, though. If, instead of us having neighboring states speaking the same language, they were neighboring countries with different languages, things might have turned out differently with respect to multi-lingual skills here.
-
The ref is cooking now all the time 24/7 since it came to me. The ref seems very stable at 7.045632V with a very little tendence to go down a bit in the >0.1 ppm range. But honestly, this is already beyond my possibility to measure this with my equipment. Probably I already measure the inaccuracy of my dmm6500 rather than that of the reference. The ambient temperature hardly changes, at most in the range of 21.7°C to max 22.0°C (49% RH) over the entire period, so that I can hardly notice any influence on the measurement.
So I will send the ref back to vindoline next week.
I would like to thank everyone involved,
it was fun. :-+
-
I am interested to participate in some sort of "Cal Club" although with only a Keithley 196 (6.5 digit DMM, 10 years out of cal) and one Geller Labs SVR-T reference (AD587LQ) available to me here, I would only find out how far my references are out and not contribute any meaningful measurement of the LTZ. That said... how can I apply to join the club? (Edit: I am in the USA, near Portland OR)
-
JBeale, Welcome! You're now in the club :D
We are currently going through a bit of a re-boot or reorg. CellularMitosis did an amazing job of getting the club off the ground and running it for a few years. I'm going to be taking over the logistical aspects soon. I'm still waiting for the references to find their way to upstate NY. At that point I think we'll see some movement!
At this point there are quite a few Cal Club members, so I'm afraid it will still be a bit of a wait for your turn.
By the way, the K196 was my first "good" meter! I still really like it.
-
Welcome, JBeale. I also have a 196 in the stable. It's hard to beat the BIG display. ;D
-
Thank you very much and it is good to know the club is still running. I understand it will be some time for my turn. In the meantime I discovered https://www.ebay.com/itm/LM399H-10V-Precision-voltage-Reference-Calibrated-LM399/254200638725 (https://www.ebay.com/itm/LM399H-10V-Precision-voltage-Reference-Calibrated-LM399/254200638725) as well as http://www.voltagestandard.com/ (http://www.voltagestandard.com/) which are not too pricey so I will try those while I'm waiting.
I'm embarrassed to admit I recently got a $10 AD584 board from Amazon, but it was what you might expect at the price. The annoying thing is it included what seemed to be a calibration sheet (mostly in Chinese) mentioning "Agilent 34401A" and listing some numbers out to 5 decimals but the 10V number printed (10.00066 V) was about 4.5 mV above what I measured! I don't imagine my Keithley is so bad as that, especially when my Geller SVR-T says it's a whole lot closer.
-
Some of those "calibration" sheets are just duplicated in bulk. Some have actual measurements made and hand written, though. It's a tossup. Does yours look like real ink on paper?
-
I think you will pleased with the one from voltagestandard.com. Mine has proven to be accurate and stable ... performance much better than its specifications. My lab is air conditioned, and that helps a lot.
On another note, is an updated list of usa cal club members forthcoming? It would be nice to know where I am on the list.
-
Indeed the voltagestandard ref looks good. It arrived today 4/20 and the documentation states it was calibrated just yesterday 4/19 to an in-cal Keysight 3458A.
I see it was burned in 1038 hours, has an added temperature compensation circuit and it has a measured tempco (+0.7 ppm/C). I consider this thing is an impressive bargain, at the price.
The only sanity check I can do myself so far is compare with my many-year-old Geller 10V ref which is now 50 uV (5 ppm) below this one, with the room at 19 C.
-
Just a little information update. The ref is on the way to vindoline. I changed the lid a little, to make the ref somewhat more useful. The ref itself doesn´t change by this. More after it reached vindoline (hopefully).
greets :)
-
Thanks for the update. Curious to see your refinement to the ref.
-
Ok, now it's time to lift the veil, what I did with the lid.
Now there are two precise resistors 1k and 10k under it, with the ability to take a real 4-wire measurement, but without changing the dimensions of the tekko housing. I put little but precise 4-wire measurement cabels in the package, hoping that was a good and usefull idea. Some pictures:
(edit translation)
-
I haven't heard anyone being against an EU Cal Club. Maybe it just needs a few people to say "Yes!" and then get started. The Aussie Cal Club was in hibernation, last I heard.
As you requested: Yes!
But all jokes aside, has there already been a EUCalClub? My 34401A is certainly in for a calibration. And I would be willing to help if so requested. In any case I hope there are enough interested people to start a european cal club!
-
Id like to participate in the EUCalClub as well. :)
-
It would be a good topic for stuttgart to organize an eu-calclub.
-
EUCalClub consider me in
-
See https://www.eevblog.com/forum/metrology/eu-calclub/msg2410806/#msg2410806 (https://www.eevblog.com/forum/metrology/eu-calclub/msg2410806/#msg2410806)
:-+
-
As HJW-d announced, the USA CalClub reference has arrived back in the USA! The package arrived yesterday and I haven't had time to do anything but power it up. It's looking good so far and I'll monitor it for 24 hrs or so. I'm still expecting the other new reference from TiN. My current thinking is to wait for both references so we can send them around together. I expect most people would like to be able to use and compare the two.
-
Ok, now it's time to lift the veil, what I did with the lid.
Now there are two precise resistors 1k and 10k under it, with the ability to take a real 4-wire measurement, but without changing the dimensions of the tekko housing. I put little but precise 4-wire measurement cabels in the package, hoping that was a good and usefull idea. Some pictures:
(edit translation)
I like it! :-+
It's also exciting to see interest spring up for an EU Cal Club. Thanks, HW!
-
Some info from Dr. Frank regarding LTZ references (from the new EU Cal Club thread):
Hello,
we may discuss the stability problems in depth on the Metrology Meeting.
I've had bilateral exchanges of LTZ references and resistors already with a number of volt-nuts.
I usually have sent one LTZ1000, un-powered, and one VHP202Z, 10k, with built-in temperature sensor, allowing precise resistance calculation over temperature.
That fits in a small DHL packet at low cost.
Transfer is not possible during winter, as the LTZ especially at temperatures < 0°C shows hysteresis of several + ppm, but during spring .. harvest, it will retrace its nominal value within < 0.5ppm.
Round trip measurements are necessary to confirm successful transfers, as well as systematic data evaluation.
-
I would think this would threaten transfer year round as most shipping methods have an air transport leg, and the cargo hold can easily reach this temperature.
-
Shipping powered reference is more expensive as it adds weight, size and requires quicker shipments.
What about doing hysteresis recovery procedure after shipping, like in Pickering patent? It has expired, I believe, so free to be used.
This way we can ship references unpowered year round.
-
CatalinaWOW
You'd think. In reality we now have a LTZ1000 Ref, which has already been transported twice across the "big pond". Now it depends on the extent to which the measured values actually differ. Certainly also the demand of accuracy plays a role.
;)
-
Shipping powered reference is more expensive as it adds weight, size and requires quicker shipments.
What about doing hysteresis recovery procedure after shipping, like in Pickering patent? It has expired, I believe, so free to be used.
This way we can ship references un-powered year round.
Yes, the LTZ 1000 can be recovered by cycling the temperature, following the Pickering Patent. For personal use, this method has always been 'free'.
I assume, that this method did never work properly in the Datron 7000 application, as its oven temperature is described as being 45°C, and symmetric temperature sweeps of +/- 40°C, as described in this patent US 5369245 then would never be achievable at usual room temperatures.
The LTZ 1000 shows an asymmetric hysteresis behavior, though.
The hysteresis at low temperatures is bigger, than at higher temperatures. This is also indicated inside a paper about the Datron 7000.
Therefore, such a technique has to be run kind of asymmetrically, and the correct procedure (contrasting the original patent) has to be evaluated.
Otherwise, a temperature logger could be added to the packet, and a mandatory round-trip-measurement might detect hysteric shifts, that is to consequently measure the voltage before and after the travel at the same laboratory.
Frank
PS: demonstration of this problem, with Pickering document:
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg1882325/#msg1882325 (https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg1882325/#msg1882325)
-
I'm sorry about all the delays, reference module for US Cal Club is shipped today.
Initial calibration data and measured stats:
(https://xdevs.com/doc/xDevs.com/FX/usac/final_spec_1.png) (http://xdevs.com/doc/xDevs.com/FX/usac/final_spec.png)
My lab will get new SI Volt (doing JJA cal for my refs soon), so I will be able to confirm/deny the absolute uncertainty of my setup. Two 3458A and K2002 were used to perform all measurements.
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_ctop_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_ctop_1.jpg)(https://xdevs.com/doc/xDevs.com/FX/usac/usac_cable_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_cable_1.jpg)
Output provided by two Pomona 3770 low thermal EMF posts. Input power and auxilary circuitry using circular shielded connector (fanout cable included in package).
Power requirements : +11 to +15VDC, -2 to -15VDC, current limited at least 80mA. Power input is protected by 125mA fuse and TVS.
Optional PCB temperature sensor : MAX6610 (U4 on PCB), with isolated power +5V.
Optional ROM data storage: I2C EEPROM with isolated +3.3V power (ground shared with temp sensor).
Calibration data provided for temperature range +20 °C to +50 °C
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_face_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_face_1.jpg)
Chip is LTZ1000A, setpoint 12.5/1K TCC VHP200 divider. Onboard regulator for LTZ1000A section is LT3042. Schematics is otherwise same as Linear's datasheet.
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_rnet_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_rnet_1.jpg)
Output amplifier is Analog ADA4522-2 chopper, together with bipolar transistor driver. Reference can survive short at the output for indefinite time.
FX module output driver can sink or source up to 20mA, but here module cannot maintain output accuracy because kelvin sense lines are not routed to the binding posts.
Inline fuses are hidden in yellow tube. Suggest not to break them :)
Just realized, somebody melted few film cap casing :palm:
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_wiring_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_wiring_1.jpg)(https://xdevs.com/doc/xDevs.com/FX/usac/usac_ldo_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/usac_ldo_1.jpg)
Looking forward to see results on this by US members. I would recommend not to open the box to avoid magic ppms leaking out :popcorn:
-
Exciting! Definitely worth the wait. Thank you, TiN! I promise not to let any ppms escape. ;D
-
:clap:
Ok, I'll sign in again... ;D ;)
-
To provide it against high import customs, declare it as 10$ selfmade electronics scrap ...
:-DD ;D :-+
(edit: excuse my joke TiN, the reference is overwhelmingly well done)
-
reference module for US Cal Club is shipped today. . . .
Looking forward to see results on this by US members. I would recommend not to open the box to avoid magic ppms leaking out :popcorn:
Wow! Just, wow. This is going to be exciting! A huge 'thank you' TiN!
-
Just read through the xdevs.com/article/792x/ page about that reference board; what a project! Impressively photographed as well. I'm curious how the PCB is held mechanically inside the metal case, assuming it does not simply float. Is there some compression gasket around the edge?
-
It is kinda floating, suspended in foam and pads. Rigid mechanical coupling to metal case could translate into board stress and cause problems.
No gasket around the edge, it's not hermetically sealed.
-
Thanks. It looks like you leave nothing to chance so maybe you have measured it already, I'm just curious if the foam may have a measurable conductance (especially if ambient humidity rises) if it contacts any traces, untented vias etc? Normally I would guess "many megohms" and not consider it further, but normally I have never encountered an analog circuit which is measured at the sub-ppm level!
-
Alien technology has landed in upstate NY! A huge thanks to Cellularmitosis for starting the USA Cal Club and providing so much of its initial enthusiasm, momentum, and the references! Also I can't thank TiN enough for his donation of an FX reference. I feel like our club is now world-class! hjw-d also donated precision resistors and Kelvin connections - thank you!
Compared to the more experienced and knowledgable participants, I'm just a beginning volt-nut. My "best" meter is an HP3456A (6.5 digit) I also have a Keithly 196 and a recently acquired Fluke 8506A. Also, my budget is more limited - a 3458A is not to be considered! I'm adding a donation of a rugged Nanuk shipping case for the references and a DIY Prologix style GPIB-USB adapter. Thanks to forum member WaveyDipole for re-writing the firmware! The GPIB adapter works great in my hands. I hope people find it useful.
The full kit contains:
An FX 10 volt ref and its power input cable
A PX 7 volt ref with a variety of input and output connectors
A 12 volt regulated power supply for the PX
Precision Vishay 1K and 10K resistors along with a custom Kelvin connection cable
An Arduino based temperature and humidity monitor with sensor cable
A USB-GPIB adapter with WavyDipole's AR488 firmware
Two blank USB-GPIB PC boards - if anyone decides they want to build and keep one!
The Gear ships out Monday morning!
-
Hi vindoline,
have you measured the PX? Because it was going abroud/oversea, had the ref relevant deviations from my measurements in the EU?
Background, see the discussion to the EU-Calclub
https://www.eevblog.com/forum/metrology/eu-calclub/msg2458485/#msg2458485 (https://www.eevblog.com/forum/metrology/eu-calclub/msg2458485/#msg2458485)
(and following)
Thanks a lot,
H.-W.
-
Looks great, can't wait to see it. My meters are also just 6.5 digit, but I've got three Fluke 731 references that I inter-compare, plus a few others. I can do good differential measurements, but haven't had a proper cal in many years. Probably still within 50 ppm or so. My meters are on a Prologix GPIB controller, but anything else can be plugged in as well. Subterranean lair/lab is usually cool, but very stable.
-
hwj-d, yes. I measured the PX reference before and after I calibrated my HP346A with TiN's FX. I'll post the data soon!
-
I've got my eyes peeled for the new cal package. Thanks, vindoline, for putting it all together and taking the lead. Of course, I second all the thanks that you itemized, as well. :-+
With regard to meters, my highest-resolution ones are also 6.5 digits. I'm also a volt-noob, having only started my metrology experience with this club. So, no worries.
One question about your HP meter. It seems to have lost a digit somewhere. So far, it's been referred to as a 345 and a 346. Perhaps by merging what we know thus far about it, might its true model be 3456?
-
All these model numbers make my old eyes spin! It's an HP 3456A 6.5 digit meter. It's great!
-
I hear ya...on both counts. :-+
-
I figure that there are basically two types of people in the USA Cal Club:
1) those who want to use the refs to check and callibrate their equipment, and
2) those with serious high end 8.5 digit meters who want to check our references!
I'm firmly in the former camp and I figured that I may as well bite the bullet and use TiN's FX reference to "callibrate" the 10V range on my two best multimeters :-/O.
So, before I sent out the gear to bitseeker, I adjusted my HP3456A (easy front panel trim pot) and Fluke 8506A (awkward "software" calibration in the mid-range of the 20 v range :palm:) to 9.999 960
I then collected data for several hours with these two meters measuring the PX and FX references as well as a couple of experimental references I've made based on the JVR and 2DW233 circuits. The plots are attached.
-
Thanks vindoline.
Hm, so your HP3456A shows the pretty same as my dmm6500, but with 5.5 Digit 10µV resolution. With 18-19°C your Lab is little too cold.
-
With 18-19°C your Lab is little too cold.
You should see it in the winter! It gets down to about 8C! :-DD
-
Package arrived today, but had meetings until just a few minutes ago, so I haven't even opened it, yet. The box arrived in good shape. That's about all I can report for now.
What all is on the agenda for this first round of the calibration set? Log measurements for each DCV and the resistors (along with temp)? For how long should each be measured? Any particular NPLC?
Just asking in order to make the data useful across members and over time.
-
I would love if members can upload all results and measurement setup photo (too see wires connection, power source, etc) to FTP.
Posted details and logins here in the post (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1502620/#msg1502620).
This will allow to track how measurements change from different people and shipping locations. So we can make a graph in the end of overall reference values to see spread.
Let's fight lack of transparency on these Cal Clubs ;)
-
Will do. I uploaded results last round to your FTP. I'll add setup photo as well.
-
I'm in the process of getting everything set up. This was a good way to force one to clear some space on the bench. ;D
vindoline packed everything really well. Inside the flat rate USPS box was another box with a Nanuk label and inside that was the actual Nanuk armored box and inside that are all the goodies.
I unstacked my 34401A and 34410A, put them side by side for even operating temperature, and hooked them up to my laptop via 82357B. Remote control was working fine.
Then, as I started looking at all the parts of the cal kit, I read on the FX reference that it should be plugged in and allowed to settle for 24 hours or more at constant air temperature. Unfortunately, as happened during round 1, the refs arrived at the onset of a heat wave. So, it won't get constant temp for the next few days. I'll try it out anyway.
As for power, I have a HP 6114A precision supply, but the FX ref needs +12 and -12. So, I'll try running it with an HP E3631A and use the 6114A for the PX ref (should I use its included regulator?).
So far, most things seem like they go in obvious locations. Thanks, vindoline, for marking polarity on the various connectors and making effective use of genders, too, to avoid mistakes. Nevertheless, I'm attaching photos of the PX reference setup, just to confirm that I have it right before powering up.
As for the Arduino and temp sensor (a.k.a., Tempduino), they go together in an obvious fashion (pic attached), but I'm not sure how to log from it. Does it just output to the serial monitor in the Arduino IDE?
On the FX power cable, there's an I2C/RPi connector and a pair of banana plugs for temp. I'm not sure if I should be using them in some fashion. Please advise.
Finally, there are two more cables, one with alligator clips and one with a header on one end and bare on the other. I'm not sure what they're for.
-
On the FX power cable, there's an I2C/RPi connector and a pair of banana plugs for temp. I'm not sure if I should be using them in some fashion. Please advise.
That is more of an optional feature. There are MAX6610 Temp sensor and EEPROM on FX PCB, but power and ground for them are 100% isolated from analog section with LTZ and output. So there is separate fanout for that power into Pi3 header, +5V for MAX6610, and +3.3V for I2C EEPROM, and digital ground. I thought of idea to have each member to record their results and write in short form (e.g. used meter, readout value, current calibrated value of ref, ambient temp, date) into I2C EEPROM, so we can have data stored within reference, no matter where it goes. But didn't do software bit for that. If you connect I2C SDA/SCL to RPi you should detect I2C EEPROM (24C64 chip), but right now nothing written in it. None of this is needed for LTZ1000 or 10V output operation, that just works from +/- 12V power supplied to banana plugs. Make sure to connect grounds properly (blue are grounds), red is +, black is -.
-
...
As for power, I have a HP 6114A precision supply, but the FX ref needs +12 and -12. So, I'll try running it with an HP E3631A and use the 6114A for the PX ref (should I use its included regulator?).
...
Hi bitseeker,
nice to see all that stuff by you now. Vindoline makes a good packaging job. :)
Yes, use the included regulator as on your picture for the px. This guaranteed equal conditions in terms of voltage. Give the regulator something like 15V. The bare and clips ended wires are only for connecting power to the regulator. The Ref output and the little connectors to the header are high quality ones only connect together, not to others. I don't use the 'tempduino', because the temp in my lab are always the same at 21.7 to 22.0°C. But you need some python script at the back-end.
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2084380/#msg2084380 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2084380/#msg2084380)
(edit: oh, i see, you are there ... :o)
-
Thanks for the background info on those optional features, TiN.
HW, OK I'll use the regulator for consistency. I didn't realize that the tempduino was from the previous round as it was added after my turn that time. Thanks for the pointer. I'll try the Python code from that discussion.
-
Oof, it was almost 40 °C today. My AC can't maintain an otherwise modest 26-27 with that much heat outside.
-
I figure that there are basically two types of people in the USA Cal Club:
1) those who want to use the refs to check and callibrate their equipment, and
2) those with serious high end 8.5 digit meters who want to check our references!
I'm definitely in the "want to use the refs to check and callibrate their equipment" camp. Eventually I'll get to #2, maybe.
-
Is it possible to be added to end of current list, not random, as need time to implement GPIB control of 3478A/1410A? Thanks for any consideration.
Dick
-
Still alive, still following the list, and still aging my LTZ1000A.
-
Is it possible to be added to end of current list, not random, as need time to implement GPIB control of 3478A/1410A? Thanks for any consideration.
Dick
Hi VNUTDENYER! Welcome to the club. You're added to the list!
-
Thanks much. Hope to add useful data in return for measuring mine.
-
I've got a recently acquired 3456 with unknown history and a DIY LM399 calibrator calibrated at a cal lab in 2006, and recently measured as 9.99931 as an average of two HP 6.5 or more digits out of cal voltmeters.
I'd also like to be placed near the end of the list to give me time to build a much better LM339 reverence than I built before. Reading here, I see details I could have done better.
In addition to the 2 unused LM399H's Ive had since the dark ages, I've ordered a few used LM399H's from ebay. Being pulls from equipment, they may have many hours of operation and be more stable than what I have now - only time will tell.
Or if anyone here lives near San Diego and has a known accurate voltmeter, I could get a head start.
John
-
Greetings folks, sorry for the late update. The heat wave is over and then I was traveling for a bit. Got everything except the tempduino (looking for an old-fashioned USB cable) set up tonight and it's now beginning stabilization.
Meanwhile, I reacquainted myself with BenchVue and took a screenshot of it logging from the two DMMs. I don't know the characteristics of these references and don't remember the behavior of the meters, so it was interesting to see them settling in opposite directions. Once everything levels off, my plan is to log the data in this configuration, then swap meters and log again to see how similar they are.
The PX ref is sitting on top of an old Keithley 870 since the power lead is too short to get to the HP 6114A.
So, here's the setup:
HP 6114A @ 15V -> PX vreg @ 11.93V -> PX ref -> Agilent 34401A -> GPIB cable -> Agilent 34410A
HP E3631A @ ±12V -> FX ref -> Agilent 34410A -> Agilent 82357B -> laptop
I don't have low-noise cables (except for a Triax one), so I used the shortest Pomona banana leads I have for the FX output connection.
-
Looking promising, log some ppms and we can run some analysis on CSV :)
We need another 10V ref in the pack, so FX can be logged in opposition, then one could get sub-ppm resolution even with 34401A. ;)
-
After dropping during the first hour or so, during the second hour the 34410A at 100 PLC shows the FX between 10.0000349 and 10.0000409. I'll take a look at the CSV after the log is done to see how much resolution is in there.
Going to bed now, but will leave the refs powered overnight and see where they are tomorrow.
-
That makes your 34410A +8 ppm off, if everything else assumed stable.
Well within even 24 hour specifications of 34410A. :-+
-
Wow, that's pretty cool considering it hasn't been adjusted in ages.
OK, noob question now. How did you arrive at 8 ppm? 1 ppm of 10V is 0.00001V. The FX ref's sticker showed 10.000000 @ 23 °C. So, at max, it's +0.00004 or 4 ppm. Where did I go astray?
-
Sure, 10V is only nominal output (specification). Actual real output comes from sticker (https://xdevs.com/doc/xDevs.com/FX/usac/final_spec.png), field EMF, T23.
Now I got traceable calibration to SI Volt just 2 weeks ago, which corrected "xDevs Volt" by -0.22 ppm :). Thread about that exercise will be published soon.
With huge leap of faith, we assume that FX ref was not drifted/shifted from shipping stresses/etc, and stayed ~0.1ppm from my measurements. If we entertain that idea, than this FX should provide EMF = 9.9999568 +/-0.63 ppm.
Rest is simple, 10.0000379 (I took average between two numbers you provided, as without raw data files we can't do any better) / 9.9999568 = +8.1 ppm high.
But I wouldn't bet more than $10 on this number >:D.
-
Got it. Thanks for the background info and summary, TiN!
I've logged a couple hours in the same configuration after the refs had been powered on for about 12 hours. Then, I swapped the meters and am capturing a couple more hours. Currently on the second hour.
So far, the 34401A is closer (~10.000229) to the FX ref than the 34410A was.
I'll be uploading the data files to the FTP for further analysis.
How shall I handle the reference resistors in the PX box? Just take a 4-wire reading with each DMM or record a log? If a log, for how long?
-
I created a new subdirectory, Round2, on the Cal Club FTP at xdevs (thanks, TiN!) and uploaded my logs for the FX ref, PX ref, and some short logs of the 1kΩ and 10kΩ. I also included copies of the images I posted to this thread.
Attached is the 4-wire setup. The hollow plugs on these wires are a bit fiddly to get positioned correctly since the DMM's jacks are very deep.
For fun, I also briefly measured the PX and FX refs on the 100VDC and 1000VDC ranges. Both bench DMMs agreed with their 10V range. I also checked my Keysight U1282A, Agilent U1252B, and Greenlee DM-820A. They're all within tolerance. So, quite the happy family of multimeters.
For the resistors, both Agilent U1732B and DER DE5000 LCR meters agree with the bench meters.
-
And, no calibration check is complete without a little vintage Nixie fun. :-DMM Looks like the resistance measurements could use just a bit of a tweak. Maybe after I restore it and clean all the switches it'll be spot on again.
-
OK, let me know if you guys want me to do (or redo) anything else with the cal kit. Then, I'll pack it all up and send it back to vindoline for verification.
-
34410A leaked out ppms (wrong range for 1Kohm resistor) ;)
Also good idea to use DCV monitoring with Hi-Z input impedance (but it's negligible error with FX).
Graphs were plot so fast! I am not used to so short sample files. My logs often span days, so I see little help of 1 hour log :-DD
I only wish temperature/humidity to be logged along the readings to determine TC correlation. Then we can measure tempco of DUT, like DMM under test.
-
34410A leaked out ppms (wrong range for 1Kohm resistor) ;)
Regarding the range for 1kΩ resistor, I left the DMM on auto (unlike when I did DCV) and the CSV shows it chose the 1kΩ range. What should it have been?
Ah, I looked at the CSV for the 10kΩ resistor and the DMM chose the 1kΩ range. I guess that means it could've been forced to use the 100Ω range for the 1kΩ resistor. I guess auto mode was overly conservative.
Also good idea to use DCV monitoring with Hi-Z input impedance (but it's negligible error with FX).
OK. It seems some Cal Club documentation for recommended test setups, procedures, etc. would be helpful, especially for those of us with less metrology experience (*raises hand*).
Actually, TiN, is there a place on xdevs where I could at least start to document the inventory of what's in the club's cal kit, what goes where, and how it all fits into the shipping container? Then you or other experienced members could add a section of recommended settings, configurations, procedures, etc. for the most useful and consistent results.
Graphs were plot so fast! I am not used to so short sample files. My logs often span days, so I see little help of 1 hour log :-DD
I'm going to have to write my own logger since BenchVue stops after every hour (plus it isn't all that reliable; I crashed it a couple times). :palm:
I only wish temperature/humidity to be logged along the readings to determine TC correlation. Then we can measure tempco of DUT, like DMM under test.
That's where it'd also be good to have my own logging software to synchronize the capture of the data from multiple DMMs, temp, and humidity. I can explore extending the tempduino's Python code to incorporate GPIB/SCPI.
Question for vindoline and TiN: Should I do more right now or move the kit along until I have my own logging setup for better data capture? I feel that I'm delaying others, but having good data is also important.
-
For ranges, should use fixed range for best fit value. 1KΩ resistor measure at 1kΩ range, etc. 100Ω for 1kΩ resistor wouldn't work, and 10kΩ for 1kΩ will give worse results, because it's only 10% of the scale and resolution meter otherwise have.
Usually idea is to collect as much data as possible, to determine correlations and to build up confidence.
I'm not in position to decide the flow, I'm on the wrong side of the globe :=\.
You could document and take notes, photos and what not, uploading into folder together with sample data, like you already have.
Then it can be taken by next member and updated/posted here in thread as well, so everyone have transparent trail of data.
It would be hard to recommend specific setups, as that likely to start limiting involved members, and ultimately defeats one of the purpose of a club?
Key point is to share everything, even something you think as obvious, as the next member in the loop may be not as familiar with setup/methods, and there are multiple ways to do same measurement. This way when all is transparent, it will be a good source to learn for everyone.
-
Thanks, TiN.
So, for the 1kΩ resistor, the CSV indicates that it was measured on the 1000Ω range. I still don't get where the ppms got leaked (back in your original comment). The DMM chose the correct range, did it not? (I'll be sure to use manual ranging next time, regardless.)
-
Didn't look on CSV for resistance, I looked at photo of setup in post #300. :-//
-
I see what you mean. That's interesting. What's also weird is that the CSV for both the 1kΩ and 10kΩ resistor indicate that the 34410A was on 1000Ω range. :-// I'm really not liking BenchVue.
-
I tried using bench view for logging a 34401A. I had lots of problems with drivers and bugs. I was also unhappy that it had no obvious way in incorporate non-HP equipment such as a temperature sensor. There was also no (obvious) way to turn the VFD displays off.
In the end I ended up using a simple python script (I recently posted a version of it on the Datron 1081 thread). This is simple and easily configured for non-standard instruments. The result is a tab-deliminated file that you can then analyze with the tool of your choice: python, excel, alpha, whatever.
The only downside is that there is no real-time logging but you can view snapshots of the logging at any time.
I know others have built much more elaborate logging systems but for me simpler is better.
-
Same here- simple data collection program with a field to put instrument specific codes in. It sits and collects data from my HP meters for later display in Excel. I do have a window where I display the last dozen readings so I know if anything has messed up.
-
Thanks, guys.
Spent some time today learning Python, PyVISA, SCPI, and how to configure, trigger and fetch data from two DMMs efficiently (i.e., integrate 100 PLC in parallel instead of waiting for each instrument to finish serially). Works great!
Next up, serial communication with the Tempduino and writing everything nicely to files.
-
If you have commercial GPIB dongle (Agilent one or more reliable NI GPIB-USB-HS) or GPIB-LAN VXI bridge there is no need to learn much.
10V logger python, that outputs values to console and saves into file:
#!/usr/bin/python -u
import sys
#import Gpib
import vxi11
from datetime import datetime
logfile_name = "test_scriptdata.dsv"
#inst = Gpib.Gpib(0,8)
inst = vxi11.Instrument('192.168.1.125', 'gpib0,13')
inst.timeout = 30
inst.write("*IDN?")
resp = inst.read(100).rstrip('\r\n')
print resp
inst.write(":FORMAT ASCII")
inst.write(":CONF:VOLT:DC")
inst.write(":VOLT:DC:DIG MAX")
inst.write(":VOLT:DC:RANG 10")
inst.write(":VOLT:DC:NPLCycles 100")
while True:
ourdate = datetime.now().isoformat()
inst.write(":READ?") # read will wait the configured nplc time
#inst.write("FETCH?") # fetch will not wait
data = (inst.read(100).rstrip('\r\n'))
with open(logfile_name, 'a') as logfile:
logfile.write (ourdate + (";%s;\r\n" % data))
logfile.close()
print ourdate, data
-
I'm using an Agilent 82357B for GPIB and USB (i.e., Python serial module) for the Arduino temp/humidity.
Yeah, not much to it once you learn the language, what modules and methods to use, SCPI commands, etc. Anyway, I have it doing what I want and will deal with the Arduino and file handling tomorrow.
-
@TiN,
first of all, big thanks for your great support.
Short question, your R5 (orig R2 70k) hasn't the same quality as R4 (orig R3 70k). I heard about, but don't really understand why it isn't really necessary to have (orig) R2 as VHP too. I too have two 75k VHP100 and four 71.5k RNC90Y laying arround, which I originally didn't want to mix on a pcb. But is it better to have one VHP 75k and one RNC 71.5k on one board to have the VHP's at the crucial point?
Thanks,
H.-W.
-
Maybe it's better to measure 4-wire the resistors without powered Ref.
-
This is what I received for USA Cal Club ref :).
(https://xdevs.com/doc/_Passives/usac_resistor/vhp_calclub_1.jpg)
In reality using expensive VHP for 70K resistors is just waste of money. Sure they look nice, but there even 5ppm change of these resistors will be invisible in any typical setup, including 8.5d 3458A measurements. I have few LTZ refs with them as just 15ppm PTF56, they work just fine. Same goes to original HP 3458A A9 design. So you see fancy VHPs for these because "we can" not due to actual need.
-
Now I'm at the same point... :o ;)
-
IMO, one is far better off to do comparisons with their own personal reference resistors by current and voltage, rather than just using a meter. I have one L&N resistor (100 ohm) with a known value that I trust. Though they're not the greatest, they do well at constant temperature.
-
@TiN,
first of all, big thanks for your great support.
Short question, your R5 (orig R2 70k) hasn't the same quality as R4 (orig R3 70k). I heard about, but don't really understand why it isn't really necessary to have (orig) R2 as VHP too. I too have two 75k VHP100 and four 71.5k RNC90Y laying arround, which I originally didn't want to mix on a pcb. But is it better to have one VHP 75k and one RNC 71.5k on one board to have the VHP's at the crucial point?
Thanks,
H.-W.
Cellularmitosis has collected the influence of resistors from several users. (https://github.com/pepaslabs/ltz1000-info/blob/master/README.md)
The 70k have different influence on output deviation.
-
@MiDi
Ah, the Datasheet doesn't show. The impact of R3 is 10x vs R2.
Thanks.
But because mixing R3 75k with R2 71.5k doesn't matter?
-
Finished my Python logger and remote control program. PySerial was giving me some grief in Win 10 when trying to read from the Tempduino. Grrr. Then, I had to decode the goofy Python byte string. Ah, the joys of an unfamiliar programming language/environment. But it's working now.
I added some extra features such as the ability to turn off the displays to reduce wear on the VFDs or show some status info (see pic for temperature, relative humidity, and elapsed time).
Oh, I forgot to mention that this time the meters are set for 10G input impedance.
-
Impatiently awaiting for few day long CSV files to feed my matplotlib python (https://xdevs.com/doc/MX_Reference/ramon_tcmx1.png) plotter :popcorn:.
-
Haha, yes. Waiting certainly isn't the fun part.
Each measurement takes three to four seconds for somewhere around 25K records per day. How long should I let it log before swapping the vrefs?
-
Depends on your patience and USA Cal Club schedule.
Full day or so usually good idea, so you can pick sweetspot data when ambient temp is about constant.
I like to log for a week to get decent confidence.
-
Ok, we'll see how it goes. Thanks, TiN.
-
I saw an anomalous temp sample scroll by today: 86°C
For a moment, I thought something had gone Fahrenheit all of a sudden.
-
I grabbed a copy of the log now that it has gone past 24 hrs to see what the data looks like. There's definitely something up with the Tempduino. It gets huge spikes in temp data (cropped them off in the attached charts) on a regular basis and the measured temperature thereafter suffers an offset until the next spike. Very strange.
The temperature value also seems to be about 2°C higher than my DMMs report.
TiN, I uploaded the temp copy in case you want to chart it: log_2019-06-24_190620.tmp.csv
-
Vertical axis labels making those charts rather useless :bullshit: :box:
I prefer scaling vertical axis on my plots in ppm deviation from the reference point for two reasons:
* Graph difference 2.5 ppm vs 5.5 ppm is easy to follow, then something like 7.2345089 to 7.2345306 V :o. At least for me 8)
* Showing absolute values like 7.2345306 without expensive traceability to SI volt is rather pointless, as typical accuracy of hobby labs ends at 5th digit, so providing higher resolution numbers have only relative value.
According to the data, your 34401A box TC is -0.79 ppm/K.
Maybe something turns on nearby, like AC or dehumidifier or neightbour's welding supply, causing your thermometer to go nuts? For time being, before you fix the root cause you can add a hack to provide clean data, add a check "if (temp > 40c) : retry temp measurement again". Also why temperature changes in "steps". Almost looks like you restart data capture every X samples, missing some time in between.. :popcorn:
Here is code for plotter I used (https://xdevs.com/datashort/USA_Cal_Club/Round2/bitseeker/abcalc_bitseeker.py).
It needs Python 2.7, matplotlib, numpy, scipy.
-
Vertical axis labels making those charts rather useless :bullshit: :box:
Those were just quick Excel charts to see, qualitatively, what the data was like, and revealed the recurring temperature data anomalies. So, not useless for what I wanted to see. Hence, I uploaded the raw data for your much better plotter program.
I prefer scaling vertical axis on my plots in ppm deviation from the reference point for two reasons:
* Graph difference 2.5 ppm vs 5.5 ppm is easy to follow, then something like 7.2345089 to 7.2345306 V :o. At least for me 8)
* Showing absolute values like 7.2345306 without expensive traceability to SI volt is rather pointless, as typical accuracy of hobby labs ends at 5th digit, so providing higher resolution numbers have only relative value.
Yes, definitely better axes.
According to the data, your 34401A box TC is -0.79 ppm/K.
34401A or 34410A? Seems the 34410A varies inversely with the temperature.
Maybe something turns on nearby, like AC or dehumidifier or neightbour's welding supply, causing your thermometer to go nuts? For time being, before you fix the root cause you can add a hack to provide clean data, add a check "if (temp > 40c) : retry temp measurement again". Also why temperature changes in "steps". Almost looks like you restart data capture every X samples, missing some time in between.. :popcorn:
I'm not familiar with this Arduino-based temperature/humidity unit that's part of the cal kit. It's the first time I've used it. So, I'm not sure what this might be. There's no nearby AC. Maybe my neighbor does have a new welder. ;D
My logger fetches a sample from the Tempduino every time it gets readings from the DMMs. So, no lost time on the logging end. It looks like the firmware or the sensor may be resetting something at the time of the glitch and then the temperature reading "corrects" itself. Interestingly, the humidity readings do not exhibit this behavior.
The anomalous temperature reading goes too low as well as too high. However, the glitch appears as an integer value instead of the expected float. So, I'll check for that condition and retry when it appears.
Here is code for plotter I used (https://xdevs.com/datashort/USA_Cal_Club/Round2/bitseeker/abcalc_bitseeker.py).
It needs Python 2.7, matplotlib, numpy, scipy.
Cool. I'll check it out.
Thanks for the feedback and tips!
-
Added the anomalous temp check, swapped the vrefs, and started logging again.
-
There's definitely something up with the Tempduino. It gets huge spikes in temp data (cropped them off in the attached charts) on a regular basis and the measured temperature thereafter suffers an offset until the next spike. Very strange.
The temperature value also seems to be about 2°C higher than my DMMs report.
Yeah, that's defiantly not right. To be honest, I didn't check out the Tempduino at all (I already have a temp/humidity logging setup). I just plugged in 5v to see that it "worked" and put it back in the box. The weird "noise" spikes are one thing, but I really don't like how the baseline shifts around. That makes it pretty useless :--
-
I don't have any temperature control or logging facility. Both are on the ToDo list. So I'm planning on just making differential measurements on my 8 references and checking my DMMs.
When the kit reaches me I'll spend some time investigating the Tempduino. I should have all the parts for one, so I can probably test everything and sort out what is going on.
I'd only expected to keep the kit for a few days, but if the Tempduino needs some attention I'm happy to do that. IIRC I got a bunch of high quality thermistors a few months ago as part of a larger order. And thermocouples are readily available from the hardware store.
-
Hi rhb. Are you next on the list? That'll be great if you can check out the Tempduino.
I took a closer look at the temp and humidity data since I noticed that I didn't put my fix for the bad temp values in the right place last night. Corrected that and continued logging.
Both sensors show baseline offsets at the same time. However, only the temperature has the crazy values. Humidity values never spike (neither up nor down), they just shift at the same time that the temp shifts.
The anomalies occur every 2 hours, 20 minutes, and approximately 10 seconds (~8410 seconds). The number of seconds varies a bit because my logger records every time the DMMs finish their 100 PLC and place their values onto the GPIB bus (approximately every 3–4 seconds).
Oof, my script just died because it got an incomplete record from the Tempduino (no humidity data). That's the first time that's happened after approximately 40 hours of logging. I'll have to add a regex pattern to verify that the string received matches the expected format prior to extracting the values from it. Actually, it looks like the Tempduino died. The display went blank. Rebooting!
What do you need for logging, rhb? There's a GPIB interface in the kit and a couple of PCBs to make your own, if you want.
-
Perhaps it's an arduino software issue. In any case, one should look at its source code and the data sheets of the sensors, regarding time behavior over i2c or spi.
-
Yep. This time, I had the received values logged to the console when bad data arrived. The string format is "temp,humidity,checksum". Looks like there were a couple of cases when the data was incomplete.
(2:20:14h) 2019-06-26 14:37:10 | 34401A: +1.00002190E+01 34410A: +7.04570976E+00 24.994°C 46.904%
Error: Received unexpected output from Tempduino [34,46.891,3ba6]
Error: Received unexpected output from Tempduino [91]
(2:20:17h) 2019-06-26 14:37:14 | 34401A: +1.00002180E+01 34410A: +7.04571001E+00 24.997°C 46.872%
Error: Received unexpected output from Tempduino [39,46.851,28af]
Error: Received unexpected output from Tempduino [dd]
(2:20:21h) 2019-06-26 14:37:17 | 34401A: +1.00002180E+01 34410A: +7.04571120E+00 25.000°C 46.855%
Error: Received unexpected output from Tempduino [37,46.812,4a19]
(2:20:24h) 2019-06-26 14:37:21 | 34401A: +1.00002190E+01 34410A: +7.04570987E+00 25.439°C 46.794%
-
Hi all, just need some info on test setups. What interface to GPIB, which SW preferred for logging, tolerance / tracking of dual supply, temp stability needed/preferred, and ? Plan to use E1410A in VXI mainframe with 3478A for PS and Tempduino measurements both under GPIB control (if I can get a USB dongle to check operation). Is the list on first page original or already random sort (can figure how much time left for setup)? Thanks again for any upfront advice on making tests useful (maybe not just to me).
-
I'm next. I need another project like I need another hole in my head, but I'll look into the problem. Major shift from lathe spindle bearing design though. I'm rebuilding a new 7x14 Chinese mini-lathe as a precision instrument lathe so I can make go/no-go gauges for Chinese RF connectors. The lathe was delivered as a complete set of partially finished castings which was assembled at the factory to make sure they don't leave out a part.
I have 2x 34401A, 2x 3478A, 3x LM399 from Jason, 3x AD584JH from eBay and a couple of Doug Malone's references. And I've fixed bugs in 2.5 million lines of other people's code. So I have a reasonable chance of resolving the logger issue.
-
Hi all, just need some info on test setups.
Hi there. Sure, I'll answer what I can. Others will hopefully fill the gaps.
What interface to GPIB,
I'm using an Agilent 82357B USB-to-GPIB adapter.
which SW preferred for logging
I started out using BenchVue, but writing my own in Python with PyVISA enabled me to synchronize the DMM readings with temperature and humidity data. I can also extend it to log non-HPAK instruments (i.e., anything VISA-compatible), which BenchVue doesn't support.
So, there's no particular preference/requirement. It's up to your personal preference and/or equipment requirements.
tolerance / tracking of dual supply,
For the FX reference, I'm using an HP E3631A set at ±12 VDC. TiN, who provided the reference, stated that a "Decent quality linear bench PSU will be fine. For critical measurements (e.g. noise) I use two VRLA 12V batteries."
For the PX reference, I'm using an HP 6114A precision power supply set at 15 VDC. It powers the reference's own voltage regulator, included in the kit.
You can see pics in this post: https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2498784/#msg2498784 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2498784/#msg2498784)
temp stability needed/preferred
I don't have a strictly temp-controlled lab at home. I guess this depends more on what you want to measure. Different environments could be interesting to see how the equipment behaves.
Plan to use E1410A in VXI mainframe with 3478A for PS and Tempduino measurements both under GPIB control
Note that the Tempduino communicates via USB.
(if I can get a USB dongle to check operation).
The kit includes an ATmega 328-based USB-to-GPIB adapter. I'm not familiar with its operation.
Is the list on first page original or already random sort (can figure how much time left for setup)? Thanks again for any upfront advice on making tests useful (maybe not just to me).
I read back through the thread and this was the last list that was posted: https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2321592/#msg2321592 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2321592/#msg2321592)
Vindoline is managing, so he would have the most up-to-date information on the participant list and the order.
-
Since I've just gone back through all the posts in this thread, I gathered a "Getting Started" list of posts about the USA Cal Club kit, its contents, and setup. If I missed anything, PM me and I'll update this.
USA Cal Club kit contents (what's in the box and how it's arranged)
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2459097/#msg2459097 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2459097/#msg2459097)
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2472399/#msg2472399 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2472399/#msg2472399)
Resistor references (installed on the lid of the PX voltage reference)
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2409894/#msg2409894 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2409894/#msg2409894)
FX voltage reference
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2423187/#msg2423187 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2423187/#msg2423187)
PX voltage reference
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1501783/#msg1501783 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1501783/#msg1501783)
PX reference's power regulator (use this to power the reference for consistency between participants)
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1516537/#msg1516537 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1516537/#msg1516537)
Tempduino (temperature and humidity)
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1511497/#msg1511497 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1511497/#msg1511497)
-
I'm next. I need another project like I need another hole in my head, but I'll look into the problem.
I hear ya about projects. Your help is very much appreciated. I'm not sure what code is running on Tempduino or where to get it. Since cellularmitosis made it, I checked his two GitHub accounts, but didn't see anything that looked like it might be it.
https://github.com/cellularmitosis
https://github.com/pepaslabs
Major shift from lathe spindle bearing design though. I'm rebuilding a new 7x14 Chinese mini-lathe as a precision instrument lathe so I can make go/no-go gauges for Chinese RF connectors. The lathe was delivered as a complete set of partially finished castings which was assembled at the factory to make sure they don't leave out a part.
What is this go/no-go gauge: To see if an RF connector was made so far out of spec as to damage the connector it would be mated with?
I have 2x 34401A, 2x 3478A, 3x LM399 from Jason, 3x AD584JH from eBay and a couple of Doug Malone's references.
Cool. If you have GPIB cables, then give the included USB-to-GPIB interface a try. I haven't tried it, so I'm not sure what software is needed to interface with it. It doesn't have VISA drivers that I'm aware of. Maybe it's Prologix compatible.
And I've fixed bugs in 2.5 million lines of other people's code. So I have a reasonable chance of resolving the logger issue.
Excellent! Just the right man for the job. :-+
-
Do we have an up to date list of participants?
Also I think it's a great idea to have documentation regarding best practices to test and whatnot for us people new to the cal club.
-
Thanks bitseeker. Will plan on dual deep cycle SLA for +-12v, 15v linear for 10v, USB for temp with included firmwar, etc. Still need to load some logging app and test it and rig some shade on lab (camper). May have to move into basement if too hot. Do "we" have a feel for how long minimum temp should be stable before meaningful readings can be taken? Is detoxit needed on copper leads for ultimate resolution (not by me :( )? Thanks again.
-
I hear ya about projects. Your help is very much appreciated. I'm not sure what code is running on Tempduino or where to get it. Since cellularmitosis made it, I checked his two GitHub accounts, but didn't see anything that looked like it might be it.
https://github.com/cellularmitosis
https://github.com/pepaslabs
I'll pester Jason if needed. Can't be much to it.
What is this go/no-go gauge: To see if an RF connector was made so far out of spec as to damage the connector it would be mated with?
Precisely. I was testing a batch of Chinese 50 ohm SMA terminators with a Tek 11801 and 20 GHz SD-26 and the male pin pulled out of one! The 11801/SD-26 combination is amazing for checking RF stuff. You can evaluate the devices as fast as you can connect them. At least, you can if you know what the time domain response should be.
Making the gauges is easy relative to rebuilding a 7x14 Chinese minilathe to hold tolerances of 0.0002". I'm still collecting the required tooling, but I think it will be a fun project. I don't expect to get much better than 0.001" on the first pass as I don't have adequate temperature control. But that's OK. The whole point is to take on a project sufficiently difficult that I learn lots of things in the process.
I've owned a 10x20 Clausing lathe and 6x24 mill for 20 years. So I have a pretty good notion of what I'm attempting. What's amazing to me is the nasty attitude on the 7x12minilathe list on groups.io. They bad mouth the machine at every opportunity and my post analyzing the spindle bearings was greeted with contempt as being "reading rather than doing". I'm not an ME, but the more I study the machine, the more impressed I am with the design. It was designed to meet a *very* low price point. Every decision was guided by manufacturing cost and expected level of buyer's skills. As delivered for under $550 US it's just a complete parts kit, assembled at the factory to make sure they don't leave out any parts. But with appropriate modification and a skilled user it need apologize to no one. There is plenty of iron where it is needed. The major compromises are in the ergonomics of infrequent adjustments such as tailstock setover for cutting tapers.
In the late 1940's, David Broadhead, an amateur, was engaged by John Strong, one of the preeminent experimental physicists of the day, to make the screws for a ruling engine Strong was building at Johns Hopkins. In preparation for the work, Broadhead rebuilt his 13" Southbend to eliminate all periodic spindle errors and make the ways straight and parallel to 0.0001". The screws were tested by interferometer to less than a millionth of an inch. After the screws were turned, they were lapped. The June & July 1952 issues of Scientific American have some fantastic articles about it.
At 6e-6"/"-F expansion, it doesn't take a lot of temperature differential to blow through 200e-6". The really attractive part of rebuilding a mini-lathe to such tolerances is it's a *lot* smaller. I plan to recondition my Clausing lathe , but I probably won't go past 0.0005" because of the size.
All the fasteners need to be replaced and some serious work done to true it up, but I know of nothing else that can touch it. A used name brand machine will require as much or more work and parts are much harder to get. I had planned to document the rebuild on the groups.io forum, but I'm looking for an alternative which is more pleasant. The list has 15,000+ members, but only about 2 dozen who are active and *very* unpleasant and outright hostile. As best I can tell, they simply BS about YouTube videos they watch. And occasionally mangle there lathes a bit more with some poorly thought out modification.
I have IDC connectors and ribbon cable. I just have not made GPIB cables up yet as I don't have an interface or a multiplexer. I've got all the parts, but have not gotten around to it.
-
Sounds like a fun project and you know full well what's required. Stay away from Practical Machinst, as they'll trash any of the low cost imports, though some have gotten away with it recently. You might find The Home Machinist much friendlier, though the traffic is lower- https://www.chaski.org/homemachinist/index.php (https://www.chaski.org/homemachinist/index.php)
What PM does have is Richard King, who probably knows more about scraping than anybody.
-
Also I think it's a great idea to have documentation regarding best practices to test and whatnot for us people new to the cal club.
I'll get something together before I send the kit back to vindoline. I may post it independent of this thread (e.g., USA Cal Club Getting Started Guide), since this thread is titled for Round 2. That way it can be evergreen regardless of round number.
Do "we" have a feel for how long minimum temp should be stable before meaningful readings can be taken? Is detoxit needed on copper leads for ultimate resolution (not by me :( )? Thanks again.
Temp stability requirement depends on your equipment. Unless you have some really high-end gear, the vrefs are much more stable than the equipment measuring them.
No DeoxIT necessary that I'm aware of. I didn't have to do anything with any of the connectors. Several of them are gold-plated. :)
-
Thanks again bitseeker.
-
Quick question as this doesn't really belong in this thread, but does anyone here mind if I message them about some equipment questions?
-
I would be happy to answer, or try to answer, questions.
-
I would be happy to answer, or try to answer, questions.
I appreciate the offer, but I was looking to talk to someone that's been actively involved with the USA Cal Club for quite a while.
-
I doubt if anyone here would mind such questions. I know I don't. But one of the values of the forum is that you encounter questions you didn't know you had, along with answers to same. If you message your questions no one else benefits from the learning. You should not be embarrassed by asking publicly, there are bound to be many others who also do not know the answers to your questions. You will be in good company.
-
Yes, especially since it sounds like the questions are related to the Cal Club, ask away.
-
Yes, especially since it sounds like the questions are related to the Cal Club, ask away.
Message Sent.
-
Just so folks aren't left hanging, SirAlucard's questions were regarding test equipment for specific active eBay listings. So, I just replied to his PM.
-
Just so folks aren't left hanging, SirAlucard's questions were regarding test equipment for specific active eBay listings. So, I just replied to his PM.
Well there goes the mystery. >:D
Tho it did bring up a question for the club. Would it be possible to add on other references for other ranges? Being able to send around a package that allows people to calibrate their own equipment to each other's. Is that to difficult to do?
-
Sounds like a fun project and you know full well what's required. Stay away from Practical Machinst, as they'll trash any of the low cost imports, though some have gotten away with it recently. You might find The Home Machinist much friendlier, though the traffic is lower- https://www.chaski.org/homemachinist/index.php (https://www.chaski.org/homemachinist/index.php)
What PM does have is Richard King, who probably knows more about scraping than anybody.
I just abandoned PM. A request for a good reference on conical hydrodynamic bearings produced nothing but BS. physicsforums.com provided contact with a Kiwi with a similar background and interests who provided the magic incantation to get google to cough up what I was looking for. Now I just need to figure out how to get ASME papers and choose a couple of good graduate level textbooks on tribology.
As for Richard King, he condemned me for saying I intended to scrape the cross slide first for use as a template for scraping the saddle and bed. I reached this conclusion independently of Connelly just from considering the task, though my analysis was confirmed by rereading Connelly this morning. Richard dismisses Connelly as a "mere votech instructor". Which is fairly weird given that he makes his living teaching courses in machine tool reconditioning.
While I still consider the title of "machinist" and "mechanic" on a par with PhD, as it takes a similar amount of time and effort to master either, I'm forced to conclude that most of the people claiming those titles are like my PhD grad school classmate who did not know Snell's law. But he did get the degree anyway.
"Turning and turning in the widening gyre
The falcon cannot hear the falconer;
Things fall apart; the centre cannot hold;
Mere anarchy is loosed upon the world,
The blood-dimmed tide is loosed, and everywhere
The ceremony of innocence is drowned;
The best lack all conviction, while the worst
Are full of passionate intensity."
W. B. Yeats.
The more I fool around on the Internet the more I am impressed by the EEVblog crowd.
Sigh...
Reg
FWIW My mini-lathe arrived broken courtesy of UPS and my spindle test bars were delayed because of stupid shipping. So I have spent all my time studying the construction other than disassembly and cleaning. The more I study the design; the more impressed I am. As received they can do excellent work in skilled hands and with a rather modest amount of work I am confident they can match anything of comparable construction.
-
What PM does have is Richard King, who probably knows more about scraping than anybody.
I just abandoned PM... As for Richard King, he condemned me for saying I intended to scrape the cross slide first for use as a template for scraping the saddle and bed.
"You're on your own Reg as you seem bent on 'teaching the forum' how to, when you're guessing at it and have never scraped a machine yourself. I was willing to help until your last speech."
-- Richard King
https://www.practicalmachinist.com/vb/machine-reconditioning-scraping-and-inspection/scraping-lathe-limits-mensuration-364598/ (https://www.practicalmachinist.com/vb/machine-reconditioning-scraping-and-inspection/scraping-lathe-limits-mensuration-364598/)
-
OK, a little diversion to the topic on occasion is to be expected. I think it's time to get back on track with the Cal Club.
-
Does anyone have any information about the Tempduino other than cellularmitosis? I'm next, so I'd like to prepare insofar as possible. Also, what's in the kit? I got very busy buying test gear and have not followed the discussion closely. I've got an HP 16500 LA arriving Tuesday.
Can someone link to the message describing the current kit?
Thanks,
Reg
-
Hi Reg. I'll link to the getting started thread as soon as I've finished writing it. One page back, post 339, I collected links to existing info about the main components of the kit, which you can peruse for some familiarity.
-
That was sufficient information. I was following the thread closely early on, but then wandered off, even going to go as far as Colorado June 12-22.
Has anyone contacted Jason about the Tempduino source code? If not, I will, but I don't want to bother Jason if someone has already contacted him about it.
I have several Uno and Mega 256 units, so if I get the source before the kit arrives I can start work on the problem. The errors look as if it might be an overflow in an averaging routine buffer, but that's only a wild guess. In any case, whether hardware or software, I'll find the cause and fix it. That will assuage my sense of guilt for spending all my time buying more test gear rather than helping ;-)
-
I haven't attempted to contact Jason, so no worries about ganging up on him. Please proceed. I concur with your hypothesis of an overflow of some kind due to the regularity of the anomalies. However, I'm not sure what's causing the hysteresis-like behavior in the values unless it's just an error in the accumulation that subsequently leads to the overflow and reset to a new baseline.
No need to feel guilty. We're all doing what we can and I'm sure could all feel guilty about some aspect of our participation. Yet, we've been able to keep this going for a few years now. That's quite an accomplishment.
I'm thankful that you can help out with the Tempduino so that subsequent logging sessions for other members will provide better data than what I was able to collect. I'm happy I was able to improve my logging functionality this time in order to get everything synchronized in one file and record for 24+ hours vs. the single-hour logs I had done in the past. Next round, I look forward to being able to have good temp/humidity data incorporated.
-
I've contacted him about something else entirely quite a while ago, and haven't heard back. So he may not be around?
-
That's the sensor-breakout. Pickup the datasheets from the sensors.
-
I just sent Jason a PM. He's quite busy with work, so only very limited BW for other things.
A buffer overflow will in general roll over to a different number. If the temperature is being read in one interrupt routine and averaged and reported in another, a race condition can arise which would exhibit the observed behavior.
There are other possible causes, but that's an obvious possibility. I suspect it's a very subtle bug and that if you have the right circumstances the bug is not apparent.
If I've not heard from Jason when the kit arrives I'll just write new FW for it. A bug fix would be quicker than new code, but reading sensors and writing them to the output is not a large program. Even at the absolutely abysmal state of my programming skills it's a couple of days work.
You know you have not been programming for a *long* time when you don't remember the syntax of your vi abbreviations and editing macros. A best guess is that it is over 4 years since I wrote a significant program.
So, as in all such cases, I'll be a lot slower than I would be if I were working regularly.
-
Do we have an up to date list of participants?
Here is the current membership list:
Vgkid
kj7e
martinr33
nikonoid
Svgeesus
bitseeker
Vacuuminded
hwj-d
rhb
RandallMcRee
orin
ArthurDent
edavid
CalMachine
Conrad Hoffman
VintageNut
vindoline
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
JBeale
Grandchuck
t1d
VNUTDENYER
KK6IL
Mark03
maxwell3e10
dr.diesel
GEOelectronics
flittle
dr.diesel, flittle and Geoelectronics asked to be skipped this go-around. rhb is up next.
-
For the curious, here's the Arduino sketch:
https://gist.github.com/cellularmitosis/4bc0de74b6c3dd03d57ef5ce09c51896
Drop the crc files into the same folder as your Arduino sketch and it should compile correctly in the Arduino IDE.
After I implemented this, I found out about Fletcher's checksum, which is much simpler, and will do just fine for an application like this. I'll give that a shot for my next project. https://en.wikipedia.org/wiki/Fletcher%27s_checksum
I believe that this is the Arduino sketch for CM's temperature/humidity data-logger.
-
I just sent this message to nikonoid, but I thought I'd echo it here for the benefit of the next recipients.
I've included a little Arduino shield which uses an Si7021 temperature and humidity sensor.
Here's the Python script I use to read the data from the Arduino and log the data in CSV format:
https://gist.github.com/cellularmitosis/421781b4f89636b612378f33d705f8c3
(also attached to this post)
Usage:
The script takes one argument, which is the serial port it should read from.
For example, using this script on linux might look like:
./Si7021-logger.py /dev/ttyACM0
The script will print output like so:
timestamp,temp_c,humidity
1500422021.565,25.4043,68.5087
1500422022.565,25.4257,68.5239
1500422023.565,25.4150,68.5239
...
That's a timestamp (seconds since the 1970), temperature in celsius, and humidity in percent.
On Linux, your serial port will probably look like /dev/ttyACM0 or /dev/ttyUSB0. On Mac, it might look like /dev/tty.usbserial-DN02TIYO.
On Linux, after you plug in the Arduino, you can run the dmesg command to figure out which device name it got associated with. I don't know of the equivalent of dmesg on Mac, so I just open the Arduino IDE to figure it out. I'm guessing you can do the same on Windows, but I haven't tried it.
Since the output of the script is in CSV format, you can just redirect the output of the script to a file in order to log to a file, e.g.:
./Si7021-logger.py /dev/ttyACM0 > temp-log.csv
If you'd like to watch the output while also logging to a file, you can use the tee command:
./Si7021-logger.py /dev/ttyACM0 | tee temp-log.csv
Technical info:
The Arduino spits out 9 bytes of binary at a time: a 32-bit floating point temperature, a 32-bit floating point humidity, and an 8-bit CRC.
The Python script reads in 9 bytes and checks that the CRC is valid. If it isn't valid, it tries reading another byte until the CRC works, so even if the script starts up in the middle of a 9-byte sequence, the script will recover after throwing away the first reading.
And here is some info on reading the data from the arduino.
-
Well, no wonder why I couldn't find the code. It's not checked in as a project, but as a Gist snippet. Thanks for the link, vindoline.
After looking at the code, this firmware is different than what's on the Tempduino. This firmware uses a fixed-length packed-byte output and the accompanying Python code converts the packed data into CSV (as confirmed by the "Technical info" section of the quoted documentation).
The Tempduino, when it first starts outputting, may send a little garbage followed by a few newline chars (i.e., blank lines) and then a line with the field names as a CSV header. Subsequent lines of output are the values, also in CSV format.
Aha! I think I found it. This one appears to send the expected output and also supports an LCD display, which the Tempduino has.
https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97
Thanks, vindoline, for pointing me toward Gist. I had previously only searched around the normal Github area.
-
OK, my C is rusty and I don't know the exact behavior of the compiler, but here's a snippet from thermometer.ino (line 27) that is averaging samples from the Si7021 before outputting a result record:
void loop() {
double rh_accumulator = 0;
double c_accumulator = 0;
for(uint8_t i=0; i < oversample; i++) {
rh_accumulator += si7021.getRH();
c_accumulator += si7021.readTemp();
}
float ambient_rh = rh_accumulator / oversample;
float ambient_c = c_accumulator / oversample;
The oversample variable is 7, so it's not accumulating a lot of samples for the averaging, especially for doubles. I'm not sure how much of an issue assigning the division from a double to a float would cause. You lose precision, but it's not repeated.
Note: If I understood correctly, I believe readTemp() is used here instead of getTemp() because the former uses the value that was acquired at the time that getRH() was called, whereas the latter gets a new sample from the temp sensor.
Then, we have this:
char buf[32];
char *ptr = buf;
char t_buf[16];
dtostrf(ambient_c, 1, 3, t_buf);
char rh_buf[16];
dtostrf(ambient_rh, 1, 3, rh_buf);
The dtostrf() function expects a double. I'm not sure if it cares that it's getting a float since it only needs to use three decimal places for creating the string.
And then it puts together the rest of the output for the serial port.
ptr += sprintf(ptr, "%s", t_buf);
ptr += sprintf(ptr, ",%s", rh_buf);
uint16_t crc = crc16(buf, strlen(buf));
ptr = csv_append_hex_crc16(crc, ptr);
Serial.println(buf);
The remainder of the main loop just updates the onboard LCD.
-
Hey CM, not a big deal but was I on list, for last, and that's why not showing?? Just current list last is ok.
-
Hey CM, not a big deal but was I on list, for last, and that's why not showing?? Just current list last is ok.
Sorry about that! List corrected.
-Vindoline
-
I got a reply from @cellularmitosis and he's going to look for the Tempduino source code tomorrow.
-
Reg, for the source, see my posts starting at https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2521782/#msg2521782 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2521782/#msg2521782)
-
OK, my C is rusty and I don't know the exact behavior of the compiler, but here's a snippet from thermometer.ino (line 27) that is averaging samples from the Si7021 before outputting a result record:
void loop() {
double rh_accumulator = 0;
double c_accumulator = 0;
for(uint8_t i=0; i < oversample; i++) {
rh_accumulator += si7021.getRH();
c_accumulator += si7021.readTemp();
}
float ambient_rh = rh_accumulator / oversample;
float ambient_c = c_accumulator / oversample;
The oversample variable is 7, so it's not accumulating a lot of samples for the averaging, especially for doubles. I'm not sure how much of an issue assigning the division from a double to a float would cause. You lose precision, but it's not repeated.
Note: If I understood correctly, I believe readTemp() is used here instead of getTemp() because the former uses the value that was acquired at the time that getRH() was called, whereas the latter gets a new sample from the temp sensor.
Then, we have this:
char buf[32];
char *ptr = buf;
char t_buf[16];
dtostrf(ambient_c, 1, 3, t_buf);
char rh_buf[16];
dtostrf(ambient_rh, 1, 3, rh_buf);
The dtostrf() function expects a double. I'm not sure if it cares that it's getting a float since it only needs to use three decimal places for creating the string.
And then it puts together the rest of the output for the serial port.
ptr += sprintf(ptr, "%s", t_buf);
ptr += sprintf(ptr, ",%s", rh_buf);
uint16_t crc = crc16(buf, strlen(buf));
ptr = csv_append_hex_crc16(crc, ptr);
Serial.println(buf);
The remainder of the main loop just updates the onboard LCD.
Personally I would:
1- validate what si7021.getRH() and si7021.readTemp() return just to not integrate invalid data. Technically you should filter out everything outside a normal reading.
2- char buf[32] is too short. you are printing (worst case) 2 x 16 bytes plus the CRC in the 32 bytes.
3- make sure the reader of the serial port validate the crc and reject the data if the CRC is invalid.
-
FYI I just received verification from Jason that this is the correct source code. I'll get started on it tomorrow.
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2521782/#msg2521782 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2521782/#msg2521782)
Many thanks to @bitseeker for making life easier for Jason. He's done *way* more than his share in this enterprise.
-
Hey guys, so happy to see this club thriving! Glad you guys managed to dig up the tempduino source. I’m not sure what is causing the strange output. Sorry it is causing such a hassle!
As you may have guessed, I have taken up another hobby lately... :)
-
What an awesome view, Jason! Looks like a great new hobby. Thanks for confirming the correct source code and popping in to say "hi". :-+
-
Wow, I didn't realize we were still so high up on the list of participants.
Jason that looks like a fun hobby, probably cheaper too. :P
-
Does his new hobby mean it is possible to crawl back up out of the volt nut rabbit hole? ;)
-
Cool, but be careful out there. I inherited some nice stuff from the estate of a good friend with leanings in that direction.
-
OK, I think I have the basics in place for the new USA Cal Club: Getting Started & User Guide (https://www.eevblog.com/forum/metrology/usa-cal-club-getting-started-user-guide/).
I'll be adding more info such as how to do logging, log file formats, where to upload them, etc., in due course. I just wanted to get all the kit-dependent stuff in place while I still had all the goodies in order to take photos of them.
-
Does his new hobby mean it is possible to crawl back up out of the volt nut rabbit hole? ;)
Probably not. Looks like he hopped on that there bike and shot out the other end screaming. :-DD
-
Looks like the Austin Chalk to me ;-)
The son of a former business partner of mine was ripping through high grass on a bike near Houston and hit a stump. He now sits in a wheel chair all day :-(
Be careful! If you can't see what your front tire is going to see, don't go there at more than walking speed. The consequences can be seriously life changing.
-
People who ride off road on motor bikes get broken. Sometimes repairable, sometimes not. It is like bull riding that way. But sometimes you gotta do what you gotta do
-
In the User Guide, I added a table of contents and a new section, Packing the Box, with a checklist for getting everything packed back into the box(es).
The kit shipped out this morning and should be back to vindoline on Monday, then off to rhb.
-
+1 on the user guide- nice work!
-
Is it not too late to join this round? I have a few 6-7 digit meters to contribute.
I would also suggest to make a spreadsheet that each user can fill out with some specific numbers:
Instrument, last calibration, average voltage, short-term std., long-term p-p drift, ave temp, p-p temp.
This will make summarizing the results much easier later.
-
Good idea. Can you make a template since you have a sense for which parameters are desirable/useful? Attach it to your post and I'll add it when I make the how-to section of the user guide.
-
+1 on the user guide- nice work!
Thanks, Conrad. It will be ongoing work as there is much to be documented. Next, I want to add a how-to section regarding setup, logging, uploading files, etc., as I remember being a bit lost and figuring it out when I first started way back in Round 1 ^-^. I'd also like to have a more formal section about the club, member requirements (nothing crazy, but to set proper expectations), etc. An FAQ would be useful, too.
TiN will be adding a section for the Python plotting script and collecting data about the FX reference.
-
Here is a simple file. Maybe one can think of a few other columns to add.
-
Eeek! I'm now seriously under the gun and have not yet printed out Jason's code to study yet. Guess I know now what I have to do tomorrow. At least melting down 2.5 pounds of pure tin wine capsules won't take too much hands on time. Mostly just waiting. Though I should make ingot molds rather than leave it in the bean can crucibles. I think I may cheat and use a SS measuring cup as a mold. It's not that hot a pour.
-
Here is a simple file. Maybe one can think of a few other columns to add.
Maybe it's better to migrate it to google sheets.
I myself am not a friend of proprietary software like MS-Office, and therefore I always use Open-Document formats according to ISO/IEC 26300 standards, for example by using the free Libreoffice.
In this case it might even be better to use the cloud capabilities of Google Sheets in an appropriate way (access rights).
-
Is it not too late to join this round? I have a few 6-7 digit meters to contribute.
It's never too late, it will just be a bit of a wait! I'll add you to our list.
-
Rejecting Excel on proprietary grounds while proposing Google sheets is inconsistent.
I have no problem with odf, though for simple sheets such as this both MS and Libre can open and use either format.
If we are really concerned about such issues a backup for TiN's generous data repository might be worth discussing. Natural disasters or other circumstances might make it inaccessible. I don't personally feel that most cloud services address this need as they are vulnerable to business upsets that are far more likely than natural disasters.
-
Perhaps we should just keep the data in .csv format and let individuals use the tool of their choice. I agree with keeping the data on TiN's server - he's been very generous with time and treasure. :-+
-
Excel is ubiquitous in the commercial world and I've no problem with it. OTOH, CSV files should make everybody happy, or at least insure that nobody is left out.
-
People who ride off road on motor bikes get broken. Sometimes repairable, sometimes not. It is like bull riding that way. But sometimes you gotta do what you gotta do
You can't get to where I'm standing in this photo with a motor bike but you can still get broken if you're not careful 8)
-
Rejecting Excel on proprietary grounds while proposing Google sheets is inconsistent.
?
What should be "inconsistent" with both proposals if they have nothing to do with each other?
I've seen excelsheets, with rudimentary macros, that only provided their functionality under an original ms-excel. (edit:) Not to mention when using different OS.
That's one thing.
Google-sheets is a cloud service, which could be very handy under certain circumstances, and is therefore used by many people when they work together on the same sheet locally. CSV-format has nothing to do with neccessary access grants, that we need to do that.
One was a hint, the other was a tip. What we can do with it, we are just beginning to see.
-
Perhaps we should just keep the data in .csv format and let individuals use the tool of their choice. I agree with keeping the data on TiN's server - he's been very generous with time and treasure. :-+
For the setup information template, my intent was to convert the XLSX file to CSV format before attaching to the User Guide. So, no worries about accessibility. You can still choose the tool and platform you prefer. :-+
I use a variety of operating systems and tools for work and home. Thus, I try to make things as cross-functional as reasonably possible to minimize headaches for everyone, myself included.
-
Rejecting Excel on proprietary grounds while proposing Google sheets is inconsistent.
?
What should be "inconsistent" with both proposals if they have nothing to do with each other?
I've seen excelsheets, with rudimentary macros, that only provided their functionality under an original ms-excel. (edit:) Not to mention when using different OS.
That's one thing.
Google-sheets is a cloud service, which could be very handy under certain circumstances, and is therefore used by many people when they work together on the same sheet locally. CSV-format has nothing to do with neccessary access grants, that we need to do that.
One was a hint, the other was a tip. What we can do with it, we are just beginning to see.
I agree with the dangers of compatibility with Excel sheets containing macros, and definitely anything with VB functionality. But your comment recommended staying away from proprietary software (although you did go on to speak only of open formats for documents). Google Cloud is definitely proprietary and could go away, become restricted, or become a pay service at any time. And that is the basis of my inconsistency comment. Even though I do agree that they are somewhat different services. (There are collaboration tools in the Office products. How well they work depends on the work style of the teams using them, and they are certainly not the be all and end all of solutions to the collaboration problem. I haven't found cloud products to be that solution either, though they do definitely have their merits.)
I am not a zealot on Open Source, but find that many people are, and I therefore sometimes read more into comments on proprietary software than was intended. I apologize if you were in any way offended.
-
I can expose mercurial repository for data/plotting and everybody can contribute there, while everything will be automagically redundant and decentralized, not relying on single weak point/server/person. :)
-
...
I am not a zealot on Open Source, but find that many people are, and I therefore sometimes read more into comments on proprietary software than was intended. I apologize if you were in any way offended.
That's ok. We are not on an ad hominem side of discussion. That's not the problem ;)
Problem was, that the statment/cognition of part one in same thread, MS-Excel with all massive inherently problems, was transferred to the second completely independent proposal to use collaborative table software, which even seems to be particularly suitable for us, because a table kept centrally in the cloud, with subscriber-related rights can even be operated by web browsers. I've rated excel to that, but not google-sheets. You can't criticise inconsistency if consistency is not intended that way.
To be honest, I'm a little surprised by the headwind when I call ms-excel less suitable for us for the known reasons, when there is much more flexible and unrestricted software available everywhere like LibreOffice.
-
It's here!!!
Many thanks to all. Very nice set up. time for me to switch from mechanical engineering to electronics.
@bitseeker. Can you make a PDF of the instructions? I'd like to print them out and put them in the box when I send it out.
FWIW I started a group, machine-tool-rebuilding@groups.io as I got fed up with being trolled by people who could not pass a high school graduation exam. I have been amazed by the number of machine operators who claim to be "machinists" but are unable to solve practical problems in algebra, geometry and trigonometry.
WRT the spreadsheet stuff, a good friend from grad school who is now a professor at Drexel has published a number of professional papers about mathematical errors in Excel. I don't use Excel, but after many conversations with him about the subject, I would not trust Excel as far as I could throw my Ford Ranger.
-
@bitseeker. Can you make a PDF of the instructions? I'd like to print them out and put them in the box when I send it out.
I tried several different ways of generating a PDF from the User Guide thread, but none of them are ideal (images are too small, includes forum formatting, includes extra posts at the end, shows ad at the top, etc.). I tried Firefox's "simplify page" feature, but it only includes the contents of the first post. The worst, though, is the forum's built-in "Print" feature — it removes all the images! :palm:
Any ideas how to do a good conversion to print without having to maintain two separate copies of the guide?
-
@bitseeker. Can you make a PDF of the instructions? I'd like to print them out and put them in the box when I send it out.
I tried several different ways of generating a PDF from the User Guide thread, but none of them are ideal (images are too small, includes forum formatting, includes extra posts at the end, shows ad at the top, etc.). I tried Firefox's "simplify page" feature, but it only includes the contents of the first post. The worst, though, is the forum's built-in "Print" feature — it removes all the images! :palm:
Any ideas how to do a good conversion to print without having to maintain two separate copies of the guide?
Wrong workflow.
Nothing against your work. But in my opinion the forum software is not suitable for a PDF export of a single post, and not even intended for it.
The right way under these circumstances would have been to link to a PDF from the very beginning to avoid duplicate work.
-
hwj-d,
I completely agree and was hoping that, having already gone down this road, I'd not have to backtrack. I never expected the guide to be a physical document. Otherwise, I'd have started it in a word processor.
All,
So, is it better to have the guide as a PDF attached to a forum post (better for printing, but must be download to read)? If so, I'd rather redo it before adding more content.
What say ye? PDF or online?
-
I would vote pdf for a number of reasons. Perhaps most important, by starting with a document processor you will have much better tools to create the content. Should be easier on your end. Also will make it easier to collaborate or pass on the task when life inevitably intervenes. Another is on the receiving end. While I have a laptop near my instruments for data logging, it really is more convenient to page through a document while setting up. The document takes up less space than a laptop, can be moved around easily, is far less finicky about the surface it sits on and doesn't require as much worry about possible damage.
-
PDF for same reasons as CatalinaWow.
Also, lets go with csv as the universal data exchange format. We all have tools that ingest them....
-
Yes, CSV.
-
OK, it's showtime!
I've got my bench cleared of the 6" of debris that has covered it for the last few months, printed out and read the Tempduino source once and booted a Windows machine with the Arduino IDE on it.
The Tempduino seems to be giving reasonable values. I'm in a 7' x 10' closet with a 6000 BTU window unit running at low fan speed and set as cold as it will go. The door to the rest of the house is open.
The Tempduino on the bench reports ~69.6 F and the WWVB clock about 3 ft higher reports 74.3 F. I've moved it to the bench top and it seems to be moving lower as the Temduino moves higher. I've also got a Sunbeam analog temperature and humidity gauge on the bench. None of these have *any* pedigree.
The room is cool enough I'm wearing an insulated vest. There are now 3 Z400s and an LCD monitor running. Outdoor temp is around 90 F and it's quite humid.
So now it's time to learn to talk to the Tempduino by USB.
Have Fun!
Reg
-
I *think* I might have fixed the Tempduino. I explicitly reset the accumulators just before the averaging loop using another Uno I had so I did not disturb the one I received. If I understand loop(), that should not be needed, but C++ does "interesting" things when you least expect it.
I need to log data for 24 hours which Arduino 1.8.5 doesn't support that I can see. At the moment I'm using Windows 7 which is very much a foreign land for me. So I'll probably switch to Debian 9.3 to use the Python stuff.
I'm going to try to find a case for the Tempduino. If I can find it, I have one which should be the right size. Unfortunately, it is brutally hot here and even 5 minutes of looking through boxes leaves me soaked, so I need to get lucky.
Reg
Edit: Thanks to the wonders of Cygwin I am now logging data to disk *without* being forced into mortal combat with the computer ;-) I'll post a graph in the morning.
Where can I find hardware details? The LCD board is different from mine and I don't have one of the Si sensors. So rather than delay shipping the kit to the next person while I make a case, I think it makes more sense for me to just make another one in a proper case and swap them out later.
-
I think it's fixed. Here's a plot of data from running all night.
It appears that it is quite sensitive to when I am in the room. I'm just using the serial port to log data to a file adn plotting it with gnuplot. So I don't have time stamps. I'll get python working on Windows if I can. If not I'll switch to Debian and log the data there.
It seems to me that adding an RTC to the Tempduino would be a nice feature. I've got one and am willing to do that, but I don't want to delay sending it to the next person. So I suggest that after I cal my DMMs I send the existing unit out with just the code change and then send out a Tempduino Mk II in a case with RTC later to join the kit.
-
Is there a document that describes how to properly perform a calibration for particular equipment models?
It would be very helpful if one knew with certainty that they were setting up the meter properly. I have a pair of 34401As and a pair of 3478As to calibrate, but *no* confidence at all that I know how. I've never read the entire manual for either, and even if I had, I've been deep in mechanical engineering land for the past few months trying to get my head around such subtleties as the deflection of non-circular beams under combined torsion and point loading. That's *very* different from the mechanics of porous media and fluids. I've learned that it is all too easy to forget to do something if you have not done it in quite a while.
A checklist of settings and work procedure for each of the common instruments would help ensure that each person gets the most out of their time with the cal kit. It would be pretty heartbreaking to realize the day after you shipped it out that you had set something wrong. I'd feel a lot better if TiN or Andreas had reviewed my workflow for correctness.
Now to get the GPIB adapter hooked up so I can log data from the instruments. I'll be logging the room with the Tempduino while I work on that to make sure that I really have fixed the bug and will post the full series logging for 24 hours this evening.
Have Fun!
Reg
-
I *think* I might have fixed the Tempduino. I explicitly reset the accumulators just before the averaging loop using another Uno I had so I did not disturb the one I received. If I understand loop(), that should not be needed, but C++ does "interesting" things when you least expect it.
Not sure what you mean. Technically the accumulators are already reset on line 29 and 30 at the begenning of Loop(). Also since those variables are allocated on the stack they are going to be deallocated when the function exit. So next time Loop() is called the 2 accumulators are going to be re-allocated and set to 0 at the beginning of the function.
See https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97#file-thermometer-ino (https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97#file-thermometer-ino)
-
I *think* I might have fixed the Tempduino. I explicitly reset the accumulators just before the averaging loop using another Uno I had so I did not disturb the one I received. If I understand loop(), that should not be needed, but C++ does "interesting" things when you least expect it.
Not sure what you mean. Technically the accumulators are already reset on line 29 and 30 at the begenning of Loop(). Also since those variables are allocated on the stack they are going to be deallocated when the function exit. So next time Loop() is called the 2 accumulators are going to be re-allocated and set to 0 at the beginning of the function.
See https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97#file-thermometer-ino (https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97#file-thermometer-ino)
That's what it is *supposed* to do, but there is no guarantee that loop() is actually a function. Once I have a 24 hour run with the modified version I'll put the kit Uno back and log another 24 hour run. If the 2nd run shows the glitches then it is fixed. I just looked at the data and it still looks good. No glitches, so let's hope.
I couldn't find any documentation for the GPIB-USB interface in connection with WavyDipole. Google just handed me links to the NI and HP units. I was about to ask, but had the wit to look at the board and found the documentation which I just printed. We need to add a link in the Cal Club 2 user manual to the Twilight-Logic/AR488 page.
I'm still looking for instructions for checking my 3478A and 34401A DMMs. In particular, would it be OK to connect both the 34401As to the PX at the same time? That would cut the amount of time I need to keep the kit in half.
TiN? Andreas?
-
It's probably not quite a full 24 hour record, but it should be good enough.
The broad spikes are the result of my physical presence and the heat from the LCD monitor. It was 95 F before noon today here, so things look correct for a 7' x 10' closet on an ouside corner wall, a 6000 BTU window unit, three Z400s running and some fool wandering in to mess around on a computer from time to time.
I'm going to shutdown and switch to the Uno with the original code to verify that it really was a bug in the Arduino toolchain and not an environmental problem at @bitseeker's location. The original code is correct as I've documented in a thread in the MCU section. But the Arduino IDE didn't do what it was told.
I'm madly reading the EZGPIB, 3478A and 34401A documentation and hope to start making calibration runs tomorrow.
Have Fun!
Reg
-
Is there a document that describes how to properly perform a calibration for particular equipment models?
It would be very helpful if one knew with certainty that they were setting up the meter properly. I have a pair of 34401As and a pair of 3478As to calibrate, but *no* confidence at all that I know how. I've never read the entire manual for either, and even if I had, I've been deep in mechanical engineering land for the past few months trying to get my head around such subtleties as the deflection of non-circular beams under combined torsion and point loading. That's *very* different from the mechanics of porous media and fluids. I've learned that it is all too easy to forget to do something if you have not done it in quite a while.
A checklist of settings and work procedure for each of the common instruments would help ensure that each person gets the most out of their time with the cal kit. It would be pretty heartbreaking to realize the day after you shipped it out that you had set something wrong. I'd feel a lot better if TiN or Andreas had reviewed my workflow for correctness.
For calibration procedures, you need to get the service manual or calibration manual for the specific device you're calibrating. Also, you'll need references for all the ranges unless you only want to calibrate the one voltage and two resistance ranges covered by the current kit.
I couldn't find any documentation for the GPIB-USB interface in connection with WavyDipole. Google just handed me links to the NI and HP units. I was about to ask, but had the wit to look at the board and found the documentation which I just printed. We need to add a link in the Cal Club 2 user manual to the Twilight-Logic/AR488 page.
Links to the documentation, project page, and forum thread are in the User Guide under AR488 GPIB Interface.
I'm still looking for instructions for checking my 3478A and 34401A DMMs. In particular, would it be OK to connect both the 34401As to the PX at the same time? That would cut the amount of time I need to keep the kit in half.
Set both 34401As to 10GΩ input and you can connect them both to the PX reference in parallel. A net 5GΩ input will be fine.
It's probably not quite a full 24 hour record, but it should be good enough.
The broad spikes are the result of my physical presence and the heat from the LCD monitor. It was 95 F before noon today here, so things look correct for a 7' x 10' closet on an ouside corner wall, a 6000 BTU window unit, three Z400s running and some fool wandering in to mess around on a computer from time to time.
That behavior looks normal to me.
I'm going to shutdown and switch to the Uno with the original code to verify that it really was a bug in the Arduino toolchain and not an environmental problem at @bitseeker's location. The original code is correct as I've documented in a thread in the MCU section. But the Arduino IDE didn't do what it was told.
I'm also looking forward to seeing if the problem returns with the original code. It should as I haven't seen this kind of anomaly with the previous temp sensor we used for the kit.
-
It seems to me that adding an RTC to the Tempduino would be a nice feature. I've got one and am willing to do that, but I don't want to delay sending it to the next person. So I suggest that after I cal my DMMs I send the existing unit out with just the code change and then send out a Tempduino Mk II in a case with RTC later to join the kit.
That sounds cool. You could send it to vindoline since he'll have the kit after each member is finished.
-
I *think* I might have fixed the Tempduino. I explicitly reset the accumulators just before the averaging loop using another Uno I had so I did not disturb the one I received. If I understand loop(), that should not be needed, but C++ does "interesting" things when you least expect it.
Not sure what you mean. Technically the accumulators are already reset on line 29 and 30 at the begenning of Loop(). Also since those variables are allocated on the stack they are going to be deallocated when the function exit. So next time Loop() is called the 2 accumulators are going to be re-allocated and set to 0 at the beginning of the function.
See https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97#file-thermometer-ino (https://gist.github.com/cellularmitosis/7249b64a3b58010b7955e83dccb60e97#file-thermometer-ino)
That's what it is *supposed* to do, but there is no guarantee that loop() is actually a function. Once I have a 24 hour run with the modified version I'll put the kit Uno back and log another 24 hour run. If the 2nd run shows the glitches then it is fixed. I just looked at the data and it still looks good. No glitches, so let's hope.
The only way for Loop() to not be a function would by adding "static inline" in the definition. The compiler doesn't really have the choice, the way Loop() is defined, it's a function. I can re-create a small example here and show you some assembly code if you don't believe me :)
-
@bitseeker Hopefully we'll have a definitive answer on the glitches in the morning. I've got all the other stuff. It's just a lot to digest quickly. I'm raising a lot of this for the benefit of the community as a whole. Someone might get the kit and then get slammed at work and try to do a cal without enough time to think through what they need to do. There's a good reason pilots have checklists.
The plan is to send the new temperature and humidity logger to @vindoline when it is ready. I'm thinking about using either an MSP430 or an ESP32 for it so it can log to an SD card in transit. I've got a couple ESP32s, but zero experience doing anything with them. I can do a code review a *lot* faster than I can digest datasheets and code for a new processor. So if someone is proficient in coding the ESP32 I think a WiFi logger would be really nice.
I'd bet that between them TiN and Andreas have one or two of everything of significance. And the two of them have more experience (aka have made more mistakes) than all the rest of us combined.
@Kosmic I can transform *anything* into *anything* (edit: natural languages excluded). And that's without taking into account what a compiler does. I rewrote 500,000 lines of VAX FORTRAN into standard FORTRAN 77 with custom function calls to replace the VAX run time library calls. It took me about a week to write the program, but it worked flawlessly and was *much* faster than doing the edits by hand. It was also completely reliable which hand edits of that scale are not.
I have routinely transformed C89 into FORTRAN 77 if I discovered I needed to do complex arithmetic. Yes, C99 will do complex arithmetic, but it's ugly. I prefer to use carpenter's hammers for nails and mechanic's hammers for machinery.
Please feel free to investigate what the Arduino IDE is actually doing and report on it. I'd really like to know. So please take a look at the temperature plot I post tomorrow. If it's got the same glitches @bitseeker reported it needs to be fixed by the Arduino maintainers as it violates the language standards as I documented in the MCU forum.
Have Fun!
Reg
BTW if you ever feel that people are unpleasant here, try practicalmachinist.com or the 7x12minlathe@groups.io list. I greatly admire machinists, but have concluded that the majority of those claiming the title are mere machine operators who have been turning wrench handles for 30 years.
-
Please feel free to investigate what the Arduino IDE is actually doing and report on it. I'd really like to know.
https://github.com/arduino/Arduino/wiki/Build-Process (https://github.com/arduino/Arduino/wiki/Build-Process)
Long story short, they call your Setup() and Loop() functions from a main() and compile everything with GCC.
Should look a bit like that:
void main()
{
Setup();
while(1)
{
Loop();
}
}
So I really think the glitch come from something else.
-
Please feel free to investigate what the Arduino IDE is actually doing and report on it. I'd really like to know.
https://github.com/arduino/Arduino/wiki/Build-Process (https://github.com/arduino/Arduino/wiki/Build-Process)
Long story short, they call your Setup() and Loop() functions from a main() and compile everything with GCC.
Should look a bit like that:
void main()
{
Setup();
while(1)
{
Loop();
}
}
So I really think the glitch come from something else.
Yep, that's how I understood it. It's just a normal C program with a wrapper to "hide" main(). I have yet to understand how it's any simpler to someone new to programming. The IDE could've just put a normal code skeleton with two empty blocks for setup and loop. :-// But I digress.
I'm also looking forward to the unveiling of where the glitches are coming from, even if it ends up being my non-temperature-controlled environment. (Ah, yes, it was those pesky neighbors welding in the middle of the night. :-DD)
-
@bitseeker Hopefully we'll have a definitive answer on the glitches in the morning. I've got all the other stuff. It's just a lot to digest quickly. I'm raising a lot of this for the benefit of the community as a whole. Someone might get the kit and then get slammed at work and try to do a cal without enough time to think through what they need to do. There's a good reason pilots have checklists.
Yep, agreed. I'm happy to incorporate anyone's checklists into the guide and include attribution. Just let me know how you want your name to appear with your contributed content.
-
I just looked at the data after switching to the Uno from the kit. I don't see any glitches. I *should* have made note of when I changed over, but did not. But I'll know in the morning.
It's entirely possible it was EMI, but the regularity makes that difficult to believe. If it is not resolved before I send the kit back, I'll send the Uno I programmed along with the original so if the glitch reappears it's just a board swap to test the software change and we don't have to deal with different compiler versions, etc.
I think it's very important that until we resolve the actual cause we do *not* modify the FW on the Uno @bitseeker was using.
I already have far too much scar tissue from dealing with this sort of thing. At least in this case, I don't have a bunch of high level managers looking over my shoulder asking if it's been fixed yet.
I had that happen once when a program I wrote tickled an undocumented edge condition in X11 and would lock up user systems at random. I spent a lot of my time for a very painful week on the phone with top level Sun support trying to resolve that. In the end I made a wild guess it was an obscure race condition in X11 and made a small change to the order of operations in the code which solved the problem. Sun support was throwing possible fixes at me as fast as I could test them. I just happened to solve it because I got lucky first. The thing that really made me crazy was *most* of the time it worked properly. But I was working at the very lowest levels of X11. I never did figure out what the issue was, but very likely it was a X11-CDE/Motif interface bug.
I've fixed lots of major problems with other people's code. That was the only one I wrote and it stung. I've always taken great pride in getting it right first time, every time.
This sort of stuff is *not* fun,
Reg
-
I just looked at the data after switching to the Uno from the kit. I don't see any glitches. I *should* have made note of when I changed over, but did not. But I'll know in the morning.
It's entirely possible it was EMI, but the regularity makes that difficult to believe. If it is not resolved before I send the kit back, I'll send the Uno I programmed along with the original so if the glitch reappears it's just a board swap to test the software change and we don't have to deal with different compiler versions, etc.
It could be a problem with the IC. Did you change the 328 or the whole board? If the latter, there could be something going on elsewhere on the board, too.
I think it's very important that until we resolve the actual cause we do *not* modify the FW on the Uno @bitseeker was using.
Agreed. Otherwise you don't know for sure what fixed it.
-
Sigh... No joy.
No glitches all night. I swapped another board I had on an acrylic base with plastic feet for the attempt at fixing it.
I'm going to drop the Tempduino effort for now and focus on the cal work. I'll build a Tempduino replacement in a nice case with in transit logging to an SD card after I send the kit back out.
-
Regarding measurements - I never had 3478A and had 34401A only for few days to do a calibration/repairs for people.
I'm not familiar with PX loading behavior, but for FX it's no problem to drive multiple meters at once, it can source/sink up to 25 mA (however I'd say use High-impedance mode on meters, when available just in case).
You'd want to run separate set of wires from each DMM to the FX terminals. I usually log at least 5-6 meters at once, so when one (or two) start acting funny, others can reveal that business.
Log at least a day of data to see stable readings, so one can determine stability figures. Usually best bet to use highest NPLC, maximum resolution and fixed 10V range. Keep azer enabled, make sure mains frequency is correct.
Use high-quality copper wiring to reduce thermal EMF effects (no need super-pupper fancy cables, just twinaxial piece of LAN UTP cable, with freshly cut and cleaned ends will do great).
After you done, upload CSV to FTP, I'll run a graph python to plot some wonky lines.
-
TiN Thanks. I did a daisy chain, so I'll redo it.
My main problem right now is that something on the Windows system I'm using has loaded python 2.7, but it doesn't appear in the list of installed programs. I've got 3.7.4 installed and I see that, but Cygwin wants to use 2.7.
Sigh.. It just sank in that Cygwin is the culprit. Time to shutdown Windows and start Debian. I'll sort out fixing Cygwin some other time.
-
Sigh... No joy.
No glitches all night. I swapped another board I had on an acrylic base with plastic feet for the attempt at fixing it.
I'm going to drop the Tempduino effort for now and focus on the cal work. I'll build a Tempduino replacement in a nice case with in transit logging to an SD card after I send the kit back out.
Yeah, that graph looks wonderful. Thanks for digging into it. Perhaps we should see if anyone else runs into the same glitches. If not, then either it's something in my environment or a problem that might be caused/fixed during shipping (bad solder joint?).
-
After considerable confusion on my part I am now able to talk to both my 34401A DMMs using putty and the Twilight-Logic board. I'd bought a GPIB-USB dongle on Amazon, but could not get it to work and returned it. So I have *no* experience reading data via GPIB.
I *really* need a logging program for a pair of 34401As and a pair of 3478As if anyone can point me to one. I'll feel terrible about delaying sending it to the next person if I have to write one myself.
If I can just get the data in files with timestamps I can massage it with awk and generate all manner of plots with gnuplot. But without an RTC on the Tempduino I'm unsure how to do that.
I've got EZGPIB installed on a Windows 7 system, so I can use that if needed, but I am *much* more comfortable on Debian. It sort of behaves like Unix.
Edit: I'll be sending out a spare Uno with the modified code installed just in case it's an intermittent HW fault
Edit2: I've got it sorted using awk. Piece of cake, which is why I never learned python. I need to hack the
Tempduino code a bit, but that's all.
-
Replied to your PM with the SCPI codes to configure and get readings from the 34401A. Not sure if you still needed that, but it's there in case you do.
I haven't programmed a 3478A yet (mine are in the repair queue), so I'm not sure how different it is. Interestingly, the 34401A has a 3478A compatibility mode. So, in theory, you could treat all four DMMs as 3478A. I wonder if you'd lose a digit on the 34401A by doing that. ^-^
-
I *really* need a logging program for a pair of 34401As
All you need is a terminal program a serial interface and setting the GPIB-Adress to 31 on the 34401A. (Printer mode according to handbook).
Note that you will have a cable with crossed RxD and TxD lines for the serial interface.
With best regards
Andreas
-
It had been so long since I got serious with awk that I forgot about the rather odd semantics of writing to files. This works with the Twilight-Logic GPIB board on Debian 9.3. I spent a long time chasing my tail before I remembered the weird bit about needing to close files so that they get reopened. Otherwise you get the same value every time even though it has changed. A pipe has the same semantics.
#!/usr/bin/nawk -f
BEGIN{
system( "stty 9600 </dev/ttyUSB0" );
system( "stty 115200 </dev/ttyUSB1" );
for(;;){
system( "/bin/date +%s >/tmp/a");
system( "/usr/bin/head -1 /dev/ttyUSB0 >>/tmp/a");
getline sec <"/tmp/a";
getline th <"/tmp/a";
sub(/\r/ ,"" ,th );
close("/tmp/a");
print "++addr 10" > "/dev/ttyUSB1";
print "read?" > "/dev/ttyUSB1";
getline dmm10 < "/dev/ttyUSB1" ;
print "++addr 22" > "/dev/ttyUSB1";
print "read?" > "/dev/ttyUSB1";
getline dmm22 < "/dev/ttyUSB1" ;
printf( "%s,%s," ,sec ,th);
printf( "%s,%s\n" ,dmm10 ,dmm22);
}
}
the output looks like this:
rhb@Debian-9:~/wrk/cal-club/34401A$ ./tst2
1563661854,24.691,48.824,+1.00001120E+01,+9.99993960E+00
1563661858,24.722,48.824,+1.00001160E+01,+9.99993840E+00
1563661862,24.735,48.811,+1.00001110E+01,+9.99994090E+00
1563661866,24.722,48.827,+1.00001140E+01,+9.99994460E+00
1563661870,24.717,48.849,+1.00001160E+01,+9.99994090E+00
1563661873,24.699,48.853,+1.00001170E+01,+9.99994330E+00
1563661877,24.685,48.859,+1.00001170E+01,+9.99994090E+00
1563661881,24.662,48.863,+1.00001170E+01,+9.99994210E+00
The first value is the time in seconds from the start of the Unix epoch. I'm using the Uno I hacked rather than the one that came with the kit. I've been careful not to modify it.
I need to get a bunch of FTDI USB-Serial adapters for reading data from DMMs and save the GPIB for other instruments.
I notice that the GPIB-USB interface is returning 8 digits. Can anyone shed light on that? It appears to be the string returned by the 34401A.
Many thanks to all, but especially to cellularmitosis, vindoline and TiN for making this possible.
After 4 straight hours of tearing my hair out I'm going to take a break. But with modest luck I should be able to test my 4 DMMs, do differential tests with the references from Doug Malone and get the kit back in the mail on Wednesday.
Once it's in the post I'll write up the scripts and stuff. My sister is very proficient at making booklets on the computer, so I'm going to sit down with her and show her what to include from bitseekers links. I'm sure we can do better, but it will be a start.
Reg
-
The DMM is integrating over 100 power line cycles, so you get additional resolution. However, depending on the amount of noise included in the signal, the extra data may or may not be useful.
-
I'm planning on doing a bunch of noise analysis of the temperature logger once I recover. That's fairly simple as one can readily estimate the physically possible temperature fluctuations.
Voltage measurement noise is a lot harder. I'm sure I can estimate age related drift really well if I can ever get data, but short term noise is a whole other ball game.
-
Regarding data logging, I have found several available serial loggers helpful:
https://bitbucket.org/hyOzd/serialplot/src/default/ (https://bitbucket.org/hyOzd/serialplot/src/default/) - nice logger with real-time plotting and logging but no time stamping so far
https://www.compuphase.com/software_termite.htm (https://www.compuphase.com/software_termite.htm) - good for interactive debugging and logging to file with a time stamp
https://sites.google.com/site/terminalbpp/ (https://sites.google.com/site/terminalbpp/) - also can do logging and timestamping plus execute macros repeatedly
The AR488 USB-GPIB program also can execute macros repeatedly or perform continuous reads.
So for continuous logging one can often find a way to do it without having to write an actual program.
-
I powered off the bench yesterday once I had everything working as I was too tired to continue. This morning I made a new cable to connect both DMMs per TiN, hooked it up and powered up.
No workee!!!!!
The part that gave me trouble yesterday, the Tempduino works just fine. But I can't read data from either meter.
If I use putty and the GPIB-USB interface *IDN? causes an error on the front panel and does not show the device but *CLS clears it. I've been through all the menus checking that everything is set properly.
So, what was working doesn't and what wasn't working does.
The fact that I can select a meter, and the unit beeps and sets the error light in response to *IDN? without returning a string to putty but *CLS clears the error has me utterly baffled.
On one unit *IDN? clears all the digits but READ? restores them, but without returning anything to putty. The other unit also sets an error on *IDN?, but does not blank the digits. In both cases *CLS clears the error light.
-
Stupid question but why turning off the power on the bench? References for sure hate that! Fluke 732's shipped and transit with backup battery exactly for the reason of uninterrupted power.
I'd start with obvious - check the cables, make sure all GPIBs screwed down firmly, power cycle the meters, make sure addresses are all correct, and line termination are set right.
Issues like this are the reason why experienced *nuts prefer to use industry standard more expensive interfaces like NI or Agilent or GPIB-LAN bridges.
Yes, they cost much more than some one-off-DIY-cheap-PIC-adapter, but save your hair on the head and sleep at night.
-
I don't know anything about the Twilight-Logic GPIB-USB-serial interface" so what I am about to say is generically about GPIB interfacing instruments.
Individual instruments. Each must be set to different GPIB address. On newer instruments this is set via the UI, older ones required switches to be changed.
Cabling. The GPIB cable is a big fat parallel cable. Instruments can be connected in any topology, star, daisy-chain, misc. Cabling is used to interconnect instruments.
GPIB to computer connection. There are various kinds including cards that install in a computer or GPIB-to-usb converters. I've had good luck with a probably counterfeit agilent 82357B one. Unless you want alot of complications, only use one GPIB interface per computer.
GPIB driver. Each hardware GPIB interface requires a (different) software driver. The main ones I have used have been the National Instruments and HPAK. I recommend the NI one. Note that various home-brew GPIB interfaces will require their own drivers.
Warning. Both NI and HPAK by default install GIGABYTES of software that you can not *ever* uninstall and are incompatible with each other. Don't let them. Install only the parts you need.
Testing. For testing I use the "keysight connection expert" (but the NI GPIB driver). Start the software with the test instrument turned off. Then turn it on and refresh. Your instrument should appear in the software. This will confirm that the GPIB address is what you expect and that the cables are connected. Try sending "*IDN?" and reading the result.
Logging software. You need something on the computer that sends commands to and reads data from the various instruments. TIN and others have published very good logging "frameworks". I just write very short and simple python scripts that use the Py-VISA library and log to a CSV file.
-
Line termination? Please explain. Having had to kill a goat or two to get SCSI to work I was wondering about that.
I just cleaned all the flux residue from the board in hope it might be that as it is *very* humid here. No joy.
I've reseated the cable and tried moving the GPIB-USB to the other end. It worked yesterday with no issues. I hooked it up, added myself to the dialout group, installed putty and except for one meter which was set for RS-232 had no issues with reading data. The time, temperature and humidity beat the bejeezus out of me, but getting data from the DMMs was not a problem.
I can talk to the units, but they are unable to reply for some reason. Unfortunately, I don't know the electrical details of the GPIB interface. The only time I *ever* worked with it was the serial version of IEEE 488 used on the Commodore 64 which was a *very* long time ago.
IIRC I bought a handful of cheap RS-232 interfaces, but have never tried them to see if they work.
The bench is setup so that certain power feeds stay on and the rest are powered off. This is for general safety and to in particular to protect CRTs from burning. Doing it yesterday was a bad idea but I was exhausted and not thinking very clearly.
-
Verify puTTY communication with the USB-GPIB interface is working by typing "++ver" which should return the version of the firmware. You may have to type <ctrl>J instead of <Enter> to allow puTTY to send <LF> to the interface.
Make sure the above works, and after you send "*IDN?", your meter should respond. You have to type "++read" to get the response from the interface, or set automode ("++auto 1") to have the interface automatically issue a gpib read command.
Hope this helps...
-
Yeeha!!!!
"++auto 1" was the missing incantation. I had done that manually early in fiddling with the GPIB-USB device. After 4 hours of fighting lovely entertainments like '\r' and the odd business that awk is not nawk (which I'm used to in the Solaris world) and that I had to close the temporary file where I was storing the date, temp and rh, I'd completely forgotten about "++addr 1".
rhb@Debian-9:~/wrk/cal-club/34401A$ ./tst2 | tee tst2.log
1563747834,23.618,53.234,+1.00001100E+01,+9.99992850E+00
1563747838,23.615,53.195,+1.00001070E+01,+9.99993100E+00
1563747842,23.608,53.167,+1.00001060E+01,+9.99992850E+00
1563747846,23.600,53.143,+1.00001100E+01,+9.99993100E+00
@JXL I owe you a *lot* of favors. I was on the verge of giving up and falling back to the end of the queue to give me time to sort out the problem. I was feeling terrible about delaying the cal kit while I sorted out issues I should have resolved beforehand. I was quite surprised to learn I was the next person after @bitseeker and almost asked for a delay.
NB for those not aware of it, nawk does not strip '\r' when it reads a string. You have to do that manually with "sub( /\r/ ,"" ,str );" If you don't if you "printf( "%s,%s\n" ,str ,next_str );" all you will see is next_str. The '\r' will rewind and you'll print next_str over the top of str. It's one of those rare and obscure things you run into if you're dealing with both Windows and Unix centric hardware and software. I've run into it less than a dozen times in 30 years of Unix. It's *very* easy to overlook and it will make you crazy, especially in conjunction with the need to close a file so it will reopen.
-
rhb, I'm glad you've got the GPIB adapter sorted. It looks like you're making great progress! Please don't feel under pressure to finish and send the gear on. Take the time you need to learn and do the measurements that you would like, calibrate your meters, etc. It could be a while before the kit comes around again!
-
Yeeha!!!! "++auto 1" was the missing incantation.
For what it's worth, you can set "auto 1" as default in the AR488.ino firmware for the USB-GPIB adapter. It's pretty simple to do via the Arduino IDE.
-
I was on the verge of giving up and falling back to the end of the queue to give me time to sort out the problem. I was feeling terrible about delaying the cal kit while I sorted out issues I should have resolved beforehand. I was quite surprised to learn I was the next person after @bitseeker and almost asked for a delay.
I don't want anyone feeling stressed about their time with the club kit! If your time arrives and you're not ready for whatever reason, we can swap your spot with the person behind you in line. Also, please feel free to take the time you need to familiarize yourself with the kit and take the measurements that you would like. It's a long wait for your turn and you should have the time to explore, learn, and have fun.
-
I feel as if Rod Serling is standing somewhere around the corner smoking a cigarette...
Now *cls produces a -420, Query "UNTERMINATED" error and turns *on* the error indicator. Which I can clear with the syst:err? command.
You can store the desired configuration in the AR488, so no need to change it in the source.
Were it not for @JXL I think I would have completely lost it today.
I've put everything on hold while I get this working. And I'm leaving it all running even though I'm not logging data tonight. My initial script was generating errors and I don't consider that acceptable even if it appeared to be working.
I should be able to put together something which is portable across any system which can talk to the AR488 and which can be configured to talk to any DMM. I only have the 34401A and 3478A, but those have different command sets, so by the time I accommodate both of those adding another DMM should just require describing the commands for that meter.
In addition to the data logging, I'll add plots and data analysis using gnuplot. Though that will probably not get finished until after I send the kit back.
I'm quite amazed by the 34401As. The older, heavily used one reads 60-70 uV below 10 V and the newer, almost unused one reads 110-120 uV above 10 V.
-
What is the correct setup for the 34401A?
As I read the manual I want to send the meters the following SCPI commands before I start collecting data.
INPUT:IMPEDANCE:AUTO ON
SENSE:ZERO:AUTO ON
SENSE:VOLTAGE:NPLCYCLES 100
SENSE:VOLTAGE:RESOLUTION MAX
Should I also set?
SENSE:DETECTOR:BANDWIDTH 3
Is there anything else I've overlooked?
Thanks,
Reg
-
FWIW I had given up for the day. But decided to have one more go at it before turning in.
The AR488 was not working even from putty, so I power cycled the AR488. Much to my immense relief, it is now logging time, temperature, humidity and voltage for my pair of 34401As.
No idea at all what the problem was, but at the moment I'll settle for getting my calibration done. I'll find out what the issue was and fix it later.
Reg
-
Oddly enough I was just reading about this in the manual. I think resolution is a measurement of the smallest size. Here is what it says:
"Specify the resolution in the same units as the measurement function, not in number of digits. For example, for dc volts, specify the resolution in volts. "
and provides the example:
MEAS:CURR:AC? 1,1E-6 6-1⁄2 digits on the 1 A range
I think you also need to set the range explicitly. This sets both the range and resolution:
Sense:Voltage:DC:Range 10, MIN
-
Yeah, I found that after I posted. I modified my code, but didn't modify the post.
It's happily logging the FX reference that TiN built. I won't have 24 hours of data until bedtime, so I'll log the 34401As until tomorrow morning and then switch to the 3478As.
An interesting thought occurred to me as I crawled into bed. I'm collecting measurements from a single device with two meters over the same time frame.
As some will recall from when I first got active in the forum, I spent my career in DSP. I can separate the noise characteristics of the FX and the LM399s in the 2 meters It's not particularly exotic math, so it's likely that TiN and Andreas are already doing this, though I've not seen a discussion of it.
Briefly, because the 1/f noise is uncorrelated, if one computes the cross correlation of the two meters one gets the autocorrelation (time domain amplitude spectrum) of the FX. One can then estimate the autocorrelation of the 1/f noise of each meter by dividing the autocorrelaton of each of the meter readings by the cross correlation of the meter readings.
If this has been discussed, please post a link so I can post my data there. If not, then I'll start a thread once i have collected the data.
Edit: I realized I don't need the cal kit to do the 3478As as I've got 3x LM399s I can use as transfer standards from the 34401As. So this evening I'll switch to logging the PX reference with the 34401As and collect 24 hours of data on that. Then I'll do short runs on the resistors and have the kit in the post Thursday which will have it back to @vindoline Saturday with luck and Monday at worst.
-
. . .
Briefly, because the 1/f noise is uncorrelated, if one computes the cross correlation of the two meters one gets the autocorrelation (time domain amplitude spectrum) of the FX. One can then estimate the autocorrelation of the 1/f noise of each meter by dividing the autocorrelaton of each of the meter readings by the cross correlation of the meter readings.
If this has been discussed, please post a link so I can post my data there. If not, then I'll start a thread once i have collected the data.
Edit: I realized I don't need the cal kit to do the 3478As as I've got 3x LM399s I can use as transfer standards from the 34401As. So this evening I'll switch to logging the PX reference with the 34401As and collect 24 hours of data on that. Then I'll do short runs on the resistors and have the kit in the post Thursday which will have it back to @vindoline Saturday with luck and Monday at worst.
I'm interested also in the autocorrelation aspects of this. I don't know of any discussions on this, personally. I have been studying this thesis (hopefully I can attach to this post). It's called "High-Sensitivity Instrumentation for Low Frequency Noise Measurements” by:Gino Giusi https://pdfs.semanticscholar.org/ed82/b314dec4644069fb6a5f7c0bba84465405e9.pdf (https://pdfs.semanticscholar.org/ed82/b314dec4644069fb6a5f7c0bba84465405e9.pdf). Is this the kind of processing you are talking about? I'm not a DSP guy, just fyi, but I can test and am trying to come up with a DIY meter for noise measurements.
I'm next in line--take your time (as vindoline mentioned).
Randall
-
Cool!!! Thanks for the thesis! I need to pull my duplexing printer off theshelf and get to where I can work on it and see if I can fix it. I much prefer sitting in a recliner with printed paper for doing serious reading.
FWIW the canonical reference on such things is "Random Data" by Bendat & Piersol. I bought the 2nd ed in grad school 35 years ago which I almost wore out. I also have the 3rd and 4th which is the last as Alan Piersol passed away. While the latter is preferable, books have become depressingly expensive and the 2nd or 3rd ed is more than good enough.
It is wall to wall calculus all of which originated with "The Extrapolation, Interpolation and Smoothing of Stationary Time Series" by Norbert Wiener. That originally appeared as a classified document during WW II and was referred to as "the yellow peril" by the EEs who had to read it. The name came from the color of the covers denoting it's classification level. Wiener wrote well. I was recently sent a copy of his autobiography which should be great fun. So far I've just read the part about the Geophysical Analysis Group which Wiener set up at MIT after the war which pioneered DSP.
Fortunately, like much of mathematics, you don't have to dig very deeply to make use of it. Most of the agony is in the proof of why it works. Though there are always rare instances when the fine print is *very* important.
I powered up the PX reference to let it warm up and one of my 3478As reads *exactly* 7.0456 V. The other reads 7.0458 V. I was reading the specs on the 34401A last night and both of mine are well within the 1 year accuracy interval. The nicer looking one was a closet queen and has never been calibrated since it left the factory. I got the 3478As for ~$110 each and the 34401As for ~$250 each.
Looking at the really nice graph TiN put on the FX reference makes me want to investigate the thermoelastoplastic deformation of silicon and the construction of the LTZ1000. I think a fairly simple model would account for the hysteresis seen when it is cycled over the 20-50 C temperature range. Over the 33-50 C range it appears to be purely elastic without hysteresis. Then the heating and cooling curves diverge until crossing again at 22-23 C. Differential expansion between silicon and gold bond wires might well account for the differences and it should not be particularly difficult to calculate the stresses and elastic-plastic threshold.
Have Fun!
Reg
-
Oddly enough I was just reading about this in the manual. I think resolution is a measurement of the smallest size. Here is what it says:
"Specify the resolution in the same units as the measurement function, not in number of digits. For example, for dc volts, specify the resolution in volts. "
and provides the example:
MEAS:CURR:AC? 1,1E-6 6-1⁄2 digits on the 1 A range
I think you also need to set the range explicitly. This sets both the range and resolution:
Sense:Voltage:DC:Range 10, MIN
In my logger, I first reset the 34401A (and 34410A) in order to have a known starting configuration. Then, set up the range and other parameters that differ from the starting config.
i.e.,
*RST; *CLS
followed by
CONF:VOLT:DC 10,MIN; :INP:IMP:AUTO ON
-
I just plotted the logs for 22.5 hrs with my 34401As reading the FX reference. The old, very heavily used meter showed about 1 ppm total variation over that period with no visual correlation with temperature.
The nice shiny closet queen showed 2 ppm variation with strong correlation with temperature. I've decided to continue logging the FX until morning. That will pick up the 1-2 degree drop overnight.
I had been planning on sending the kit back out on Thursday, but this result is very interesting. How do other people on the waiting list feel about my keeping it a little longer and shipping it out Saturday instead?
I'll make plots and run some statistical analysis in the morning and post the results.
Reg
-
Upload some CSVs to FTP, so others can have fun too. :-+
Instructions (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1502620/#msg1502620).
-
Columns are:
seconds since start of run
temperature C
relative humidity
heavily used 34401A
closet queen 34401A
All data are from the FX reference. The spike in the humidity was my doing a load of laundry in the washer and dryer 10 ft away. There is a 6000 BTU window A/C cooling the 7' x 10 ' closet which is my current lab area. Air is deflected towards the ceiling as much as possible. There are 2x 34401A, 2x 3478A , 3x Z400 and a linear PSU running in the room which has the door open to the rest of the house. My presence in the room is very noticeable.
I'll post a longer copy tomorrow. I'm only doing it now because I woke up at 2 am and an hour later am still awake :-(
I'll upload to xdevs when I've finished the run.
FWIW the heavily used 34401A was wonky when I got it from ebay and did not produce any usable data. I "fixed" it by cleaning a 1 cm splotch of flux residue from where an LF357 had been replaced.
Edit: Data have been removed as the full run set is posted later with FX, PX and ratio in a single file
-
How do other people on the waiting list feel about my keeping it a little longer and shipping it out Saturday instead?
Reg
What you learn could help others on the list. Take your time.
-
I have turned off the A/C, so the room will warm up to 85 - 90 F during the day as there are 3x HP Z400s running to provide heat. I'll then turn it back on around 4:30 to finish off the FX run.
What I'd like to do is repeat the same run with the PX followed by doing a similar PX/FX ratio run Saturday and Sunday and ship the kit back on Monday. That will give me 3 series of ~100,000 samples spanning 48 hours and including a ~20 F temperature excursion. Not as good as an environmental chamber, but the best I can do as that build is only partially completed and is not large enough to hold the meters anyway.
I've been making a list of data plots to generate and those 3 data series will keep me busy for quite a while. Each of the data series will involve making about 20 plots and possibly double that. I have a suite of statistical tools I wrote when I was analyzing rock properties from several hundred GB of data from oil and gas wells, NOAA and other sources.
Edit: I've attached a plot showing the distribution of errors based on the nominal 10 V output from the FX. It's just part of the data, but I wanted to see what the distribution looked like. This is strictly just an initial peek at the error distribution. I'll repeat later for various temperature ranges. I had to modify my program for computing the histogram to handle doubles. Rock measurement errors are 1-2% *if* you're lucky.
-
I made new plots of the data and have attached the input file to this post. For some reason I got an argument from the server about saving it.
This is all just exploratory data analysis.
-
Some plots from rhb data.
-
I can't read the text on the plots. Would you please use a bigger font? And also explain what you're trying to show. The "closet queen" (col 5) has much more temperature dependence than the "work horse" (col 4). Your plots appear to show the opposite, but even using a magnifying glass I can't be sure about the curve keys.
I posted what I did just to give a general idea of what I was doing. in about 20 minutes I'll switch to the PX reference, at which time I'll post the full data series file for the 48 hour FX run. I'm not sure when I'll have time to do plots and other analysis as I've had personal matters intrude upon my having fun. :-( But taking care of family takes precedence over playing.
-
I don't see any issue with size, did you download large picture?
My numbering is from zero, so col3 (green) is more stable meter, col4 (blue) is less stable one.
Otherwise everything should be straight-forward. One plot in timescale, second is temperature (ambient) dependency.
-
col 1: time in seconds since start of run
col 2: temperature in Celcius
col 3: relative humidity
col 4 work horse voltage
col 5 closet queen voltage
My 1600x1200 screen is in portrait mode which makes the text slightly smaller, but it's still too small even if you use landscape mode.
I made my figures to professional society meeting presentation standards. Until I retired shortly before employment for very expensive old guys went away in 2008, I had missed 3 annual Society of Exploration Geophysicists meetings in 28 years.
I used gnuplot with "set term post color 'helvetica-bold' 14 lw 3". I *think* that's readable in a hall with 500 people in a normal projector setup. I've never had a chance to put one up and walk to the back of the room to check though. A size 16 font might be needed in a room that large. My sub-specialty had smaller rooms assigned that only seated 200.
That may seem like a minor detail, but if you fail to communicate with the audience it's a waste of everyone's time. I've seen far too many grpahics which were impossible to read even from the first 10 rows. The only justification for enduring it was you wanted to see the next talk.
Real life has intruded into my play time, so there may be some delay before I can post more analysis. I'll post data as it becomes available and ship the kit to @vindoline on Monday,
I'll probably spend 100+ hours on analyzing the FX, PX & PX/FX data. There are a lot of things to look at in that data set. So results will trickle out as time permits.
-
OK Here's the full 48 hour FX run. The PX is logging and I'm off to bed.
Many thanks to to all who made this possible, but especially @cellularmitosis, @WaveyDipole, @ vindoline and @TiN.
Have Fun!
Reg
Edit: Data have been removed as the full run set is posted later with FX, PX and ratio in a single file
-
That may seem like a minor detail, but if you fail to communicate with the audience it's a waste of everyone's time.
Image has decent enough resolution and size, so that is not an issue of the generated data.
If it is open and resized afterwards to tiny phone screen or obscure 1200x1600 resolution - how is that my problem? After all you can kill some tree and print on A4, where it will be all readable. :palm:
Since this is goodwill I'm not investing lot of time for best possible formatting, but rather reuse my existing plotter that was made for much bigger data set (18 DMMs) with your data. Hence the smaller font-size.
-
Auto-scaling will make it small. View the plots at 100%. They're nice and big.
-
Tin: Thank you for the very cool graphs. Can you say a bit about what they represent. I assume the shadow colors are the samples and the dark lines are some form of trend or average. How were they made?
Rhb: Here is what the legends say (you will need to click on the picture and then save it to view it full size). The originals are large enough resolution that you can enlarge them as well.
-
For the next person setting up logging for a 34401A:
One other important hint that bitseeker gave me is that the 100NPLC read takes approximately 3 or 4 seconds. I had trouble with pyvisa until I increased the timeout.
After I got it working, I changed from "read?" to "initiate" and "fetch?" so that I could communicate with other instruments while waiting for the 34401A to finish.
The final bit of 34401A SCPI is "display off" to save wear on the VFD display.
BTW, what is a good interval between samples? As fast as possible? 10 seconds? 1 Minute? I know some temperature sensors self-heat if sampled too frequently.
Next up, 34970A...
-
Over the course of my career I spent approximately 2000 hours in professional scientific meetings looking at around a million plots of almost every imaginable type of data. I offered a suggestion on improving the presentation of data.
How you present your data matters a lot. If your audience can't understand what your figure is showing in a few seconds they will probably not understand your presentation. Large complex datasets are very difficult to present well. Doing it well is one of the things that distinguishes the A list. If someone has to ask you to explain your figure, it needs work. And if you want to make the A list, both your writing and your graphics must make the grade. It is a lot of work. So your have to care about meeting A list standards to do it.
In the past I have not bothered to comment on poorly done figures posted here, but in this case, as I had trouble recognizing my own data, I thought I should.
I commend to everyone's attention the work of Edward Tufte on data visualization.
https://www.edwardtufte.com/tufte/ (https://www.edwardtufte.com/tufte/)
Much otherwise excellent work has gone unnoticed because of poor presentation.
Reg
-
How you present your data matters a lot.
I have complete agreement on that matter. Axis without labels or mismatched colors haunt me at night, because you forgot what chart shows next week already. :)
I only disagree about previous statement that graph is too small, which is not true, I always spend extra effort and storage space to generate higher resolution images.
MaxFrister
Your assumptions are correct. Dark lines are result of gaussian_filter1d from python lib over data, just to see a data trend.
All calculations, like median, min/max, sdev are done on raw unfiltered data, trend used only for graph purposes.
Bottom graph is traditional seconds timescale, imagine just as oscilloscope, but with very slow time/division. I have posted Python plotter generator app sourcecode few posts back, when we went thru bitseeker's dataset.
Top graph is relation between temperature (ambient sensor on X axis) and measurand (Y axis). This graph is not very clear, because temperature change was too little.
Usually better data have temperature swing >5c (e.g. AC on / AC off cycle).
Perhaps rhb could log FX with large temperature swing (e.g. one day AC on, one day AC off) if he is interested in temperature stability of his 34401A.
Because measured FX reference have TC <0.05ppm/K, it's own error usually only a tiny fraction of the 34401A's own TC error.
-
God says, "No.". The daily highs were in the mid to upper 90's just before I got logging working. but now it's low 80's and will be until I send it out Monday. It's nice, so I'm not complaining.
I've got a good cal on the 34401As and should be able to transfer that to my other instruments with careful thought and a few additional gadgets. The "Work Horse" is more than adequate for testing my voltage references.
Cellularmitosis offered me the "Work Horse" in an email one morning after he sniped it on ebay on an "if it doesn't work I'll take it if you pay the extra shipping" basis. Talk about an offer you can't refuse. I'd seen it, but was unwilling to bid on it because it looked to be wonky, which it was when I got it. It just produced random digits measuring a DC voltage source.
I opened it up to take a look around and saw a 1 cm splotch of flux around an LF357 which had been replaced. A series of VCR failures while living in Houston had suggested to me that the hygroscopic nature of rosin was a significant issue in humid environments. I had already verified that was the case with a half dozen other "repairs" which consisted simply of cleaning residue from consumer gear. So I took a spray bottle of 91% isopropyl and an old toothbrush and cleaned the residue from the board. It immediately worked like a champ. But until now I did not know just how good it was.
It has redoubled my interest in investigating the aging and thermal hysteresis of voltage references. I am certain that very fine work has been done at HPAK, Fluke and Keithley, but it's all trade secret stuff. Unfortunately, I have to have many years of data for such work. So even if I start collecting data, it may never get finished. My requests for existing historical data have been rebuffed. So a few weeks of analysis must wait on several years of data collection and the burden of building and maintaining a data acquisition system for the task. Prior to learning I was next on the list, all my attention was on learning to solve the strains of arbitrary prisms under combined torsion and bending moments. A subject to which I shall return as soon as the kit is in the post and I have set up logging on some LM399 references.
It took me 3 years of intensive study of 3000 pages to master the mathematics to solve the problem of forecasting aging. What are the chances that someone building and logging references as a hobby is going to put that much effort into the mathematics?.
That effort was without question the most challenging intellectual effort I have ever undertaken and I probably have over 30 semester hours of university level mathematics. But none of it even came close to "A Mathematical Introduction to Compressive Sensing" by Foucart and Rauhut.
And to setup the equations, you have to be able to write out the thermovisocelastoplastic constitutive equations for the wirebond to the reference die. For starters. It probably gets worse. And having dealt professionally in the oil industry with much simpler problems in the mechanics of materials, I don't think there is a snowball's chance in hell that anyone has much chance to do all that on their own. Humans don't live that long.
So without cooperative effort, it won't happen. At least in my lifetime.
Reg
-
Over the course of my career I spent approximately 2000 hours in professional scientific meetings looking at around a million plots of almost every imaginable type of data.
That's truly impressive. Assuming each talk at those meetings was 12 minutes (followed by 3 minutes of questions) that works out to a data plot being presented every 5.76 seconds. Every 12 minute talk would contain (on average) 125 data plots. I don't know how anyone would be able to absorb any information from that many data plots presented at that rate.
-
One other important hint that bitseeker gave me is that the 100NPLC read takes approximately 3 or 4 seconds. I had trouble with pyvisa until I increased the timeout.
Ah, yes. I forgot that I had also increased the PyVISA timeout. I'm glad my mention of the time required was enough to point you in the right direction.
After I got it working, I changed from "read?" to "initiate" and "fetch?" so that I could communicate with other instruments while waiting for the 34401A to finish.
What I did was send "read?" to both meters, but not fetch right away. While the meters were integrating, I then fetched the temp and humidity. Finally, I read back the DMM's values. It seemed to work effectively the same as using initiate and fetch. There seemed to be some overlapping functionality and behavior in the SCPI command set.
The final bit of 34401A SCPI is "display off" to save wear on the VFD display.
Yep. Same here. :-+
BTW, what is a good interval between samples? As fast as possible? 10 seconds? 1 Minute? I know some temperature sensors self-heat if sampled too frequently.
I took samples as fast as the DMMs would let me, fetching temp and humidity at the same rate. So, that'd be around a sample every 4 seconds.
The firmware on the Uno takes 7 samples and averages them every time you request a value, so if there is a self-heating issue, I'm not sure what interval might be problematic.
-
I've got NPLC 100 set and am getting samples from both meters at ~2 second intervals. For simplicity I just used the 1 second granularity of the Unix epoch.
I left the displays on so I could easily see what was going on.
Right now I'm as nervous as a cat in a room full of rocking chairs, but once the ratio run is finished and the kit is in the post I'll start looking into the various details. The Si7021 that Jason used for the Tempduino is very low power, so I don't think self heating is an issue. I plan to investigate adding an MSP430 based transit logger to the kit as soon as I can get the parts. As that will be battery powered, sampling at 1-10 minute intervals is probably desirable.
From a statistical perspective, the larger the number of samples the better. If you plot a histogram of a Gaussian distribution, it doesn't really look Gaussian until you get to around 100,000 samples.
TiN requested photos of the lab setups, so I've attached a couple, one from the PX run and another from the ratio run.
Have Fun!
Reg
-
Thanks for photos. I'd wish they would be bigger size, hard to see any details.
-
I'll upload a higher resolution image to your website. This camera does not offer a reasonable image resolution for posting. Eventually I'll get around to fixing the battery compartment door on my old camera and go back to using it.
Questions?
The leads are all from the same piece of CAT 5, length variation is under 3 mm.
Banana plugs are cheap Chinese with gold color plating. No idea if it's gold or not. The 34401As are on the bench surface because there is no way to reach the GPIB socket with them on the bench. Wires are secured to the banana plugs by a loop through the wire ring.
PSU is an Instek GPE-4323. I'm using 1 & 2 for the FX set at 12.04 V and 4 set to 12.20 V feeding the regulator you supplied for the PX. 12.04 V is what the PSU claims. I was so flustered I didn't stop to think about the regulator drop feeding the PX. I'll check both after the run finishes.
My major complaint about the Instek is that adjusting the output voltage closer than 50 mV is almost impossible. I've had it apart to see if there was room to change the pots, but it's too cramped. Though I might bring out leads for fine adjustment pots in a box on top.
I made a shorting plug for the FX from a dual banana plug with a piece of 12 AWG wire and a tag attached to the wire ring. Inserting a bare copper wire into the FX banana jacks risks an ESD. As the banana plug can be inserted without human electrical contact it reduces that risk. It also makes it less likely that someone with vision impairment might power up the FX with a dead short across the output. The tag reads "ANTI-STATIC SHORT" on one side and "REMOVE BEFORE USE" on the other. I plan to put heat shrink over the 12 AWG short wire and fill the sockets with hot melt glue before I ship it out.
I'm sending out my Uno with the FW hacked to just spit T & H without the display working. The original Tempduino board is untouched. I think that should stay that way until we resolve the issues @bitseeker ran into. The one I'm sending out is on an acrylic base with plastic feet, so very easy to identify.
"Work Horse" at address 22 is on top with the "Closet Queen" at address 10 on the bottom.
I have a stock of Cinch banana sockets and should be able to find a suitable plastic enclosure, so if it meets with general approval I'm willing to repackage the PX so that all the connections are via case mounted banana jacks before I place the kit in the post on Monday. General idea is to hot melt glue the PX and VR to the case, trim the existing leads, solder and label the jacks. If I do that, I'll make another shorting plug like the one I made for the FX.
A transit temperature and humidity logger will take some time, so I'll send that to @vindoline separately when it is ready.
-
I see the size limits have changed.
-
Are there something about verifying resistor values?
-
I am currently logging the 10K resistor on "Work Horse". I plan to log both resistors on both meters up until I take it to the PO.
I'd like to propose some changes to the cal kit:
1) box the PX reference with the voltage regulator TiN furnished as follows:
dual banana jacks for power in and voltage out
dual banana jacks for the resistors, one pair for each resistor
2) add the following cables
a pair of one dual banana jack to two dual banana jack cables
a ribbon GPIB cable to operate a stack of two 3478A/34401A meters
3) add a 2nd shorting plug like the one I made for the FX
I don't have the required female Centronics IDC connector, but can make up the cable with the males I have and let the next person add the female. Or I can remove the Centronics connector from the AR488 and box it with the cable coming out the side. Without a regular GPIB connector to space the AR488 out from the 34401A there is no way to connect it and the IEC power cord. I think that placing the AR488 in a box with the pair of male Centronics connectors would be the best, but it would delay mailing until tomorrow.
The PX changes would allow logging both resistors to two meters simultaneously in addition to doing voltage and ratio measurements. So in five logging runs with two meters the following tasks could be performed:
measure FX voltage on 2 DMMs
measure PX voltage on 2 DMMs
measure PX/FX ratio on 2 DMMs
measure 10K resistor on 2 DMMs
measure 100K resistor on 2 DMMs
Everything needed to do the above would be in the kit. I still need to rewrite the logging software and clean up the Tempduino code. But it would make the kit completely turnkey for the measurements listed above.
If the group wants me to make the changes I've outlined I need to hear by 4 pm CDT today as that is when I plan to post it back to @vindoline. If I make the changes it will go out tomorrow instead.
Have Fun!
Reg
-
The quality of hollow bananaplugs are lightweight goldplated berillium-copper ones. So they are not on the cheapest side. The little plugs at the other side too. Keep in mind, that the mechanical quality of the precision socket is actually not suitable for others than quick cautious measurements with this lightweight cables and shortly plug them out after that. So I suggest we leave it at that.
-
I'll leave it as is. I'd been logging the 10K on one meter and just moved it to the other meter before switching to the 100K. I only proposed the changes out of concern that the connectors are not very robust. I was a bit surprised when I saw the connectors on the PX.
I should have time to collect about 15,000 samples on each meter for each resistor before the trip to the PO.
-
I'll leave it as is. I'd been logging the 10K on one meter and just moved it to the other meter before switching to the 100K. I only proposed the changes out of concern that the connectors are not very robust. I was a bit surprised when I saw the connectors on the PX.
I should have time to collect about 15,000 samples on each meter for each resistor before the trip to the PO.
Thanks.
The socket isn't that robust. The cables and their connectors are. You have a 1k and a 10k resistor.
;)
-
Well, I'll actually have 5000 samples for each on the "Work Horse" and 1000 for each on the "Closet Queen". Much too late I spotted a typo in my script and had "RESISTANCE" instead of "FRESISTANCE". So I got lots of 2 wire data, but not a lot of 4 wire.
On the bright side, I should have an excellent measurement of the resistance of the wire leads ;-)
I'm just waiting on the 10K and will be heading to the PO in about 15 minutes.
BTW I started a thread on suggestions for round 3.
-
It's in the post and I've sent vindoline the tracking number.
I fitted a space for the 2nd Uno with a note on the acrylic base what it is for.
The Tempduino FW on the board I added is hacked so the display does not show time T & H values. I had to comment out a bunch of stuff to get it to respond quickly enough to a query. It's right around 2 seconds to read time, temperature, humidity and two 34401As set at 100 NPLC using my scripts.
I reattached the LCD to the original Uno which has a cardboard base. So everything is as received, but with a 2nd Uno added.
The one I added is just for backup in case the problem bitseeker saw reappears. That way we don't need to change the original FW to test fixes.
Thanks to all for their contributions and patience. I've shut down that computer for maintenance. Once that is done I'll post the full data sets to xdevs.com.
Reg
-
Here is the full data set. I'll let TiN transfer it to xdevs.com if he will be so kind. I'm going to delete the earlier copy of the FX run data.
As I botched the script for the resistor data and did not get much data I am just postnig the voltage reference data.
-
I have started on the data analysis. For the benefit of other members I should like to offer a few suggestions and comments. It is unclear what data I actually got and how usable it is.
1) If you do not have a GPIB interface already, make one using an RTC, Si7021, an Arduino and AR488 *before* the kit arrives.
2) Test your data collection setup with a 9 V battery if you have nothing else prior to the kit arrival. If you need help ask. I'll be working on this over the next few weeks.
3) For characterizing noise, NLPC=10 is preferable to NLPC=100 if the instruments will respond to a query fast enough and you are logging automatically. If you're doing it manually, use NLPC=100.
4) Synchronous sampling will aid in suppressing EMI. I think that can be done, but am not yet sure if or how.
Gross errors in my data acquisition procedure which I have identified so far. There are likely more.
1) Failure to query setting after sending the instrument a command to set a parameter.
2) Stacking instruments and failing to shield them and the reference from room air currents.
3) Use of 1 second timing resolution in logging. I'm not sure yet how to address this as my timing was the concatenation of the system, Tempduino and DMM response times. There is a lot of variation in the interval between my samples.
So far, the only thing I can say about my data is that the noise on the Work Horse FX data is Gaussian distributed.
Have Fun!
Reg
-
How important is the consistency of the sampling?
-
How important is the consistency of the sampling?
I don't have the foggiest idea. That's one of a long list of questions. At the moment the list is getting longer rather than shorter. So far I have only looked at Work Horse. Closet Queen is next. Work Horse has a nice linear ~0.3 ppm/C TC except at around 20 C there is a ~0.25 ppm jump above the mean based on an eyeball estimate. IIRC that roughly corresponds to hysteresis in the FX reference as measured by TiN.
The Closet Queen TC is quite non-linear and much steeper. I'm very curious about the ratio of the two DMMs vs humidity. Work Horse got a thorough wash with isopropyl. The Closet Queen did not that I recall. I'm pretty sure I opened it and looked at it for evidence of repairs, but I don't think I did anything other than put it back together.
I need to track down my post on the analysis of aliasing in a multislope ADC. Another question is how is NPLC implemented in the 34401A? Zero crossing count with a comparator or timer? Short term power frequency fluctuations resulting from grid level load balance changes would result in intermittent variations in the error statistics if it is timer based. There would also be a regular change in the error statistics as they adjust the frequency coming up on midnight.
Edit:
Unknown irregularities in sampling will lead to an erroneous estimate of the exponent of 1/f**a noise. It will introduce spurious noise at higher frequencies which will flatten the slope and lower the estimate of the exponent.
In the case of known irregularities in sampling, an L2 DFT is unable to correct the errors. I *think* that an L1 DFT will correct them, but have not done a test.
-
Received the cal club shipment today.
Hooked up the FX reference to +-12v and powered on. Going to wait 24 hours before logging.
It has turned into a very busy week at work and so does anybody mind if I don't touch the PX reference? My wife thinks that I am also working on her gardening projects all weekend. Sigh. :'(
Preliminary measurements for FX reference are
K2001 (last cal date 5/2018 by TiN): 9.999997 volts
R6581T (cal by me to Fluke731B): 9.9999725 volts
with a fair amount of noise for both...
Other meters to check, but no logging scheduled, are an HP3456a (self cal) and a K2015 last cal 2016. I won't report on these here unless there is some interest?
Randall
-
Why wait 24 hours? Log away from the start, this way you can also check how long it takes to settle..
I won't report on these here unless there is some interest?
Up to you really, but I wouldn't see any issue getting as much data as there can be. :)
-
I'd like to 2nd TiN. The startup data is very valuable for the investigations I want to pursue. You can always throw data away. But you can't get what you don't have.
-
Argh. After reading your messages, and after work, I tried mightily to get logging going. However, windows did its massive os update on my laptop and my prologix adapter no longer works—or its ezgpib, not sure. I will have to debug it or use the arduino. Very disappointing.
Sent from my iPhone using Tapatalk
-
Sounds like we need to rename USA Cal Club to USA Gpib Club.
No-one have working datalogging from the first try :-DD
-
It occurred to me at some point that one "modern" way of reading these instruments would be to use a webcam with image capture and OCR. I even looked a bit if anyone had written a program like that. There are GitHub programs for timed image capture and separately for OCR. It should be possible to write a program like that and it would also be useful for logging instruments that don't have a digital interface. ("modern" means to transfer 1 Mb of data to read less than 10 bytes).
Edit: Googling a bit more I just found a few possible programs:
https://github.com/skaringa/emeocv (https://github.com/skaringa/emeocv)
https://www.pyimagesearch.com/2017/02/13/recognizing-digits-with-opencv-and-python/ (https://www.pyimagesearch.com/2017/02/13/recognizing-digits-with-opencv-and-python/)
https://github.com/xyk2/Scoreboard-webcam-OCR (https://github.com/xyk2/Scoreboard-webcam-OCR)
https://www.unix-ag.uni-kl.de/~auerswal/ssocr/ (https://www.unix-ag.uni-kl.de/~auerswal/ssocr/)
https://github.com/LazizEx/WebAcm_Number_Detect (https://github.com/LazizEx/WebAcm_Number_Detect)
Edit2: The last program, available as an executable for windows here https://yadi.sk/d/FGzn5L6s3EK9Xr (https://yadi.sk/d/FGzn5L6s3EK9Xr) seems to be exactly what one would want. It even makes a real time plot and time stamp for saving to Excel. The only feature I did not find is switching between webcams.
-
However, windows did its massive os update on my laptop and my prologix adapter no longer works
Yes, very aggravating. Thanks Bill Gates!
I'm a big fan of unsing a stand alone, headless raspberry pi as my GPIB logging solution. The whole setup isn't much bigger than the little GPIB interface itself and it can be tucked away in back of your meter if you wish. I connect to it via WiFi and the built in VNC server to control it.
-
Made progress--confirmed that windows update "lost" several of my usb drivers and reinstalled the important stuff...wireless mouse included.
So now logging is getting started.
I use Linux at work, again I know why.
Thanks,
Randall
-
Glad to hear you got it sorted. I *really* felt beat up by the time I got stuff running.
I didn't participate in Round 1. I'm hoping that by the end of Round 2 we will have the process worked out so that we can smoothly provide participants with an annual cal with good traceablility even if not quite NIST level.
Have Fun!
Reg
-
Sounds like we need to rename USA Cal Club to USA Gpib Club.
:-DD I've got plenty of GPIB stuff to be a member of that club. I'm not sure if that's good or bad. ^-^
-
I've discovered that the Tempduino reported a bunch of 2000+ C temperatures towards the end of my PX/FX ratio run and an old research code of mine is giving me occasional NaNs. So I have some debugging to do. And a *lot* of cleanup in the form of adding bounds checking, etc. But I'm getting fairly close to being able to post plots and analyses of my runs.
I read through the Mega version of AR488 and it's *really* fine code. So it will be a real pleasure to work on extending the functionality.
Have Fun!
Reg
-
Some measurements, zipped, labeled m1-m6...
For first three measurement runs my temperature logging was to a separate file, I then realized how horrible correlating that was with the gpib file and subsequently created a gpib script (thanks CatalinaWow really) to put all-in-one.
So for m4,m5,m6 the file line is <Date, Temperature, Humidity, R6581T, K2001 data>
In all cases K2001 is measuring USA Cal Club FX reference.
m1-m4: R6581T is measuring delta between my in-house LTZ reference and the FX reference. Btw it has never been temperature compensated, unlike the FX. FX is negative lead.
m5,m6: R6581T is measureing delta between Fluke731B and FX reference. FX is negative lead.
Temperature measurement using TI EVM sensor HDC1010.
Bitseeker,TiN: where/can we post to the xdevs site? If so, can we put instructions in the cal club instructions thread?
-
...
It has turned into a very busy week at work and so does anybody mind if I don't touch the PX reference? My wife thinks that I am also working on her gardening projects all weekend. Sigh. :'(
...
Yes. Me.
Is it so difficult to take at least once a measurement of the PX voltage and the two resistances, and to announce the three values here, yours Win10 chaotics? :box: :popcorn: ;) ;D
I'm waiting for that, since I deliver it back ...
-
...
It has turned into a very busy week at work and so does anybody mind if I don't touch the PX reference? My wife thinks that I am also working on her gardening projects all weekend. Sigh. :'(
...
Yes. Me.
Is it so difficult to take at least once a measurement of the PX voltage and the two resistances, and to announce the three values here, yours Win10 chaotics? :box: :popcorn: ;) ;D
I'm waiting for that, since I deliver it back ...
Ok, my panic has subsided and I will supply measurements on the R6581T and K2001, tomorrow.
-
FWIW I have discovered what I think are serious EMI issues in my PX/FX ratio run in addition to Tempduino problems. I'm still investigating.
I've got the initial analysis of the FX & PX runs done, though I want to recheck them. I'll start a thread on the subject when they are ready. Probably late today.
It doesn't take a lot of time to switch references. The PX is much simpler to connect. I was super paranoid because of the split supply requirement of the FX. I must have checked everything 4-5 times before powering up the FX.
-
Bitseeker,TiN: where/can we post to the xdevs site? If so, can we put instructions in the cal club instructions thread?
Yes, the tutorial section is on my to-do list to add. I'll do that as I port the guide over to PDF.
-
I have sent the cal club package back to Vindoline.
I have posted all of my raw data and a bit of analysis to the xdevs ftp site per TiN's instructions. Thanks (no icon for gratitude!? :wtf:)
See directory RandallMcRee and note that there is a ReadMe.txt file there to start with.
I made two charts, using Ulrich's Plotter....
-
Thanks (no icon for gratitude!? :wtf:)
I guess the closest we currently have is :-+
-
I made two charts, using Ulrich's Plotter....
Your plots would be easier to digest if you subtracted off the mean (or nominal value) and re-labelled the vertical axis in µV.
-
After watching all of the logging problems I fired up my system and found myself in failure mode also. Mine is based on NI GPIB-USB-HS, and after Windows update and NI driver updates it labels mine as "not a genuine NI product" and refuses to play further. So far in my research I have found that if I can load the NI 4882_1700f0.exe driver set all will be well, but NI has removed this and earlier versions from its servers "to assure the quality of GPIB data collection". NI's solution for the problem is for me to buy a genuine NI product (Roughly $1000). Even though on line reports indicate even buying from authorized NI distributors doesn't guarantee genuine.
I have been using EZGPIB, which requires a working set of 488 drivers. I am now looking into other options including Python so I can use other low cost adapters.
In Round 1 I added a TI HDC1010 temp and humidity sensor to the package. I have logged many tens of hours using this sensor with no glitches. I have discovered a problem with this sensor which may be why it has been dropped from the club. In the delivered configuration the sensor is self heated by the microprocessor doing the conversion from the I2C serial interface of the sensor chip to USB resulting in a roughly 1.6 degree error. I didn't notice this prior to donating the unit. The board is designed to easily allow separating the sensor with a 5 wire cable. After doing this modification mine it meets the "typical" specification of 0.2 C error. On line you can find measurements on two sets of ten of the devices showing a distribution with a one sigma error of roughly 0.1C. It may be worth resurrecting this device if debugging the Tempduino continues to be a problem.
-
Now you see why I prefer using Raspberry Pi for data logging any day or night?
Unlike windows environment, you suffer only once (when learn how to setup scripts/logging) instead of every time MS "improves" and "optimizes" your experience.
-
I've been lucky so far. GPIB has continued working on my laptop running Windows 10. However, I do plan on migrating to a standalone RPi or BeagleBone Black/Green logging setup. :-+
-
Package received in good condition.
-
I have managed to find some older drivers and am back in business, but working a longer term solution.
-
I've been lucky so far. GPIB has continued working on my laptop running Windows 10. However, I do plan on migrating to a standalone RPi or BeagleBone Black/Green logging setup. :-+
Try out the itsy bitsy teenie weenie Pi Zero W ;)
-
Everything is powered up and stabilizing overnight and I am using some good supplies to power everything. I have a cheap temp/humidity unit that seems to agree with the fancy one which gives me a little more confidence that those readings are correct.
I did check the resistors on the PX and I could compare them against my standard resistors that I believe to be pretty close. The oldest physically beat up 1K resistor that I have hasn’t been checked for calibration in 30 years but agrees with the much newer and nicer HP 1K. The two HP 10K resistors read exactly the same so I tend to believe them as well. Here are the initial readings I got. I’ll check them again later on.
LT1616 999.997 ohms HP s/n40185 10,000.02 ohms
HP11103A 999.998 ohms HP s/n40189 10,000.02 ohms
PX 1K 999.997 ohms (label 1000.02 ohms) PX 10K 9,999.82 ohms (label 9,999.94 ohms)
The initial 7 volt reading from the PX was a little different than the label indicated but I found the sequence of numbers interesting and included a photo.
-
The USA Cal Club package has been sent back. I didn’t get a chance to do everything I probably could have but got enough done and gathered enough information to keep me busy for some time.
I had ordered all the parts for the GPIB interface a couple of weeks ago so I assembled the bare board I received and got to the point where the LEDs seemed to indicate a handshake and I got a 4 line message on my laptop indicating that the board was recognized but not programmed.
On measuring the 10 volt output of the FX supply, I probably did it differently than most people would. I wired the FX supply in series opposed with my LTZ1000ACH 10 volt Dr. Frank supply so that if the output of the two supplies was equal, the voltage difference between them would be exactly zero. To measure the difference I used my 50+ year old HP419A null meter which can measure down to 3uV full scale. To see a short video of the 2 being compared, go to: https://youtu.be/wTGxBJNNhWc
Although I had no history on it, I believed my HP3457A was pretty close but that was more of a gut feeling than hard fact and I don’t know when it was calibrated last. Judging by the CALNUM it had been calibrated about 20 times. The resistance readings of my standard resistors seemed very close to what I expected and the voltage readings also seemed to agree with my 5.5 digit Fluke 8840 (what do you expect for $16 at a hamfest). The HP3457A and my LTZ1000ACH 10 volt supply had been powered for several months and the readings seemed to be a stable 10.00000 VDC with maybe a 1 for the last digit.
When I connected the two supplies and the null meter up the difference was far greater than I expected. What I discovered is that if I connected the HP3457A across my supply while it was still connected to measure the null, that the difference on the null meter was much less. The adjustment on my supply is a 30 turn 5K pot that is wired so that 1 turn is between 5-6uV and exactly 10 volts is close to half way. By adjusting the pot I could regain a null on the 3uV full scale range and at null my HP3457A read about 9.99996 occasionally up to 9.99997 volts. If I read the graph on the FX supply correctly at 24°C the output would be 9.9999585 to 9.9999600 volts. This leads me to believe that my HP3457A is pretty close, or close enough so I don’t want to mess with the calibration. Apparently there is some small loading and/or leakage that affects the readings and where the 2 supplies agree that close with my HP3457A connected across my 10 volt supply, I assume that is pretty much the correct reading.
I decided to check the 2 supplies to see how they reacted to light loads and I knew mine could survive a direct short for some time because I designed the added transistor buffer within the opamp feedback loop to handle a direct short and also protect the opamp. With a 100K load both supplies dropped 1uV; at 10K, mine dropped 10uV, the club’s 20uv; and at 5K, mine dropped 20uV, the club’s 40uV so the output resistance of mine is less than the club’s but neither would be loaded like that in any testing situation. There may be some capacitive leakage from line powered devices causing part of the problem. Both my supply and the HP419A null meter can be run from internal batteries but the HP3457A is line operated and digging out 2 batteries and running the FX supply from batteries was more than I wanted to do. I’m satisfied that my HP3457A is reading close enough for my needs so unless I get a higher accuracy meter (which I don’t see in the budget) I’ll be satisfied with what I have now.
More to follow later tomorrow.
-
Sounds like your 3457A is spot on. I wouldn't have messed with it, either.
-
Sounds like your 3457A is spot on. I wouldn't have messed with it, either.
It should be right on, I paid $225 for it with the HP44491A scanner card! 8)
-
What are the possible error sources when performing that kind of measurement - bias current for the DMM, and any ground loop issues, plus thermals on all the connectors? It seems very hard to wrestle with electrons at the microvolt level...
-
What are the possible error sources when performing that kind of measurement
Hello,
additional it could also be some EMI effect (e.g. from a LED-Lamp, USB-cable, mobile phone or switchmode supply) nearby.
I had around 20-30uV "offset" between unbuffered and buffered output of my LTZ#3-LTZ#6 due to EMI.
After putting a 100 nF capacitor across the (capacitively decoupled) LTC2057 output the offset was within datasheet spec. (< 4uV)
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg846835/#msg846835 (https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg846835/#msg846835)
I usually try to put all equipment used for a measurement on a common metal plate ground to reduce EMI effects.
with best regards
Andreas
-
If you don't have a spectrum analyzer and an LNA for it, get an SDR. The RTLs are super cheap, but the extra 4 bits of the SDRplay RSP2 is worth the money IMO. I have one and am quite impressed by it.
Turn on the MW & VHF broadcast filters and take a look at the spectrum below 500 KHz. You are likely to be shocked at what you see with a 6" antenna or even *no* antenna.
Another option if you have a DSO with a good FFT such as the Instek GDS-2000E line, which can be hacked to enable the MDO-2000 spectrum analyzer option, is to make a small loop to stick on the end of a scope probe. That will allow you to locate sources as a 1/4" loop is very insensitive. I put 1/8" wire mesh over the openings on my PCs because of PSU noise they were emitting.
My next lab area is likely to be a full on Faraday cage. And I live in the boonies. LED lamps are a *huge* issue, even if you don't have any.
-
Who has the package?
What is happening?
Data! More data!
-
Arriving at CalMachine this weekend!
-
Arriving at CalMachine this weekend!
Everything arrived safely :)
I've taken some preliminary pictures and I am about to begin logging.
-
Picture Directory: https://xdevs.com/datashort/USA_Cal_Club/Round2/CalMachine/Pictures/
Safe arrival of the USA Cal Club References :-+
(https://xdevs.com/datashort/USA_Cal_Club/Round2/CalMachine/Pictures/First%20Look.jpg)
Reference Shot
(https://xdevs.com/datashort/USA_Cal_Club/Round2/CalMachine/Pictures/References.jpg)
Here is a snapshot of each meter's first measurement after the start of logging and applying power to the references.
(https://xdevs.com/datashort/USA_Cal_Club/Round2/CalMachine/Pictures/First Measurement.jpg)
-
It's been quiet, just curious about the current status?
-
Hi Conrad, the kit is with edavid at the moment. You're next after him! I would think in the next week or so. I'll keep you posted as soon as I know anything.
-
Cool- if I start cleaning up today, I might have some bench space by then! Sometimes my space is more like an archaeological dig than a test bench.
-
I think I'll have it for about another 2 weeks.
-
No problem if I'm next, no problem if I'm not. Plenty to do in the meantime.
-
I think I'll have it for about another 2 weeks.
edavid, take your time. You had to wait!
Personally, I’m really happy with the turn-around times we've had. It’s a long list, but we’re really making progress!
-
I think I'll have it for about another 2 weeks.
Great! Can you tell us something about your equipment and your goals? Are you going to be posting results to xdevs as others have done?
Thanks,
Randall
-
Still a long time before my turn in the barrel. But it is never too soon to get ready. As mentioned earlier I found my GPIB controller interface broken by one of the Windows updates and got that sorted out. For the next arrival I plan to log all of my references and the two club references at the same time, using the multiplexer on my HP 3457. The attached schematic shows the plan. Each reference will be measured directly. In addition the FX reference will be measured differentially with the KMoon reference and the PX reference, allowing use of the lower scale and perhaps achieving better resolution. I will also do a differential measure with an HP 3456.
The use of the multiplexer brings up the question of what errors are introduced. A few runs have been done with several channels all connected by nominally identical wiring to the kMoon reference. The results show that the 3457 is doing as well or better can be expected, but there do seem to be persistent biases between channels. See the chart showing one data set. In many data sets collected on this setup a similar, but not identical pattern has been shown. I still haven't figured out if the differences are noise or something else. The differences are well down in the noise neighborhood.
Any comments would be appreciated.
-
Thermal voltages?
-
Thermal voltages?
Could be, though I tried to avoid that. Photo shows test setup. Everything had been on for two weeks before the data was collected. Data showed that meters stabilized in about four hours. All identical test leads to breakout panel. Breakout panel has BNCs wired to copper PWB with identical materials and similar lengths to pickoff point for cabling leading to 3457 in. Those wires are all identical materials and similar lengths, soldered to breakout board and tied to standard 44492 screw terminals on back of 3457. Room temperature was relatively stable less than 1 deg C variation, and uncorrelated with observed short term drifts seen in all channels.
Another possibility is some sort of noise synchronous with the sampling, since the channels are collected serially, not at the same time. Haven't been able to come up with anything of this sort either.
The EZGPIP script used is included for reference. Renamed to text file to allow posting to forum. If used in EZGPIB just rename back to .488 file.
Still plenty to think about before the reference arrives. Fortunately there is lots of time on the calendar, who knows how much will be available for this project.
-
I didn't think mux relays were particularly trustworthy in terms of thermal EMFs?
-
I didn't think mux relays were particularly trustworthy in terms of thermal EMFs?
I agree, and there are other things to worry about in a switching system. Which is why I am trying to characterize this. The alternative is to correlate measurements made on several different days with each other. Neither is without pitfalls.
Right now my tentative conclusion is that the channel differences are repeatable to within roughly a microvolt and that I can correct the differences gaining about a half microvolt of precision. And hence an improvement of the accuracy of the comparison to the calibration source. All of this is trying to use the 3457 way beyond its spec but there seems to be reason to believe the answers.
-
I didn't think mux relays were particularly trustworthy in terms of thermal EMFs?
I agree, and there are other things to worry about in a switching system. Which is why I am trying to characterize this. The alternative is to correlate measurements made on several different days with each other. Neither is without pitfalls.
Right now my tentative conclusion is that the channel differences are repeatable to within roughly a microvolt and that I can correct the differences gaining about a half microvolt of precision. And hence an improvement of the accuracy of the comparison to the calibration source. All of this is trying to use the 3457 way beyond its spec but there seems to be reason to believe the answers.
Ah, I see, needed another cup of coffee... - Is it possible to reverse some of the connections and cancel some errors out with multiple readings? If only there was a way to measure a reference voltage with AC...
-
Sounds like we need to rename USA Cal Club to USA Gpib Club.
No-one have working datalogging from the first try :-DD
After following this thread this comment almost made me spit my coffee out all over the place! :-DD I agree and decided a real Keysight 82357B was better than messing with some of the other stuff. Also... the more data the better and putting it all in one place so everyone can compare it is pretty important.
Bill
-
Right now my tentative conclusion is that the channel differences are repeatable to within roughly a microvolt and that I can correct the differences gaining about a half microvolt of precision. And hence an improvement of the accuracy of the comparison to the calibration source. All of this is trying to use the 3457 way beyond its spec but there seems to be reason to believe the answers.
I found a similar (but larger) issue with the HP 34970A and a 34901A relay card. I left the same reference connected to some of the inputs for some time and the differences appeared quite stable, so I left it to investigate later.
Inside a meter is probably not the best place for a relay card to be. I usually leave a cold card at least on hour before starting to use it since there are quite large errors when it's first plugged into a warm meter. I have a card to repair, so I'll try to look at it more when doing that.
There is certainly extra work involved when trying to squeeze the last drop of good data out of equipment.
-
I've given up on getting my UGPlus GPIB adapter working. So I'll try the AR488 with the cal kit.
What software do I need? EZGPIB? Are there any scripts available for use with a 3456A? Or better yet, a pair of them.
Is there any value to anyone else of a log file on 3456A's? One tracks my LM399 within 4 PPM over 10 deg/C, the other 12 PPM. Not real meaningful without knowing the characteristics of my LM399. If no value, then I'll then I'll just take an occasional reading over a day, and adjust the 3456A's and LM399.
Just about to add a LM35 temperature sensor IC to my LM399. Can then easily read by switching to the rear terminals of the 3456A.
-
Yes, you can use EZGPIB. There is a very nice manual available for the AR488: https://github.com/Twilight-Logic/AR488/blob/master/AR488-manual.pdf (https://github.com/Twilight-Logic/AR488/blob/master/AR488-manual.pdf)
To get familiar with the club kit (and links to more AR488 info) also see the guide at https://www.eevblog.com/forum/metrology/usa-cal-club-getting-started-user-guide/ (https://www.eevblog.com/forum/metrology/usa-cal-club-getting-started-user-guide/)
-
Now since Cal Clubs don't come to me, it's time I come to Cal Club :-DD.
Count me in for 2020 :=\
-
I'd love to be in it too :)
-
Now you’re both on the list!
-
Has anyone made calibrated measurements of the FX or TX references below nominal room temperature? My lab has been in the 12-15C range lately.
Thanks...
-
The designer did. (only down to 20 C) shown. Maybe Ilya will add some data?
-
Are we still going with the same round in the cal club? Why is there a new list being generated if the current list hasn't gone through?
-
Are we still going with the same round in the cal club? Why is there a new list being generated if the current list hasn't gone through?
It's rather a “rolling admission” type of affair. If you would like, I can add you to the list.
-
Well yea, I would hope we ended up keeping the order in which we already started.
-
If someone happening across this thread was to consider adding their name to the list, it might be useful to know an "estimated wait time" based on (current list count x average cycle time). I don't know either of those numbers myself.
-
+1, Would be great to see current queue and list of members :)
-
Vindoline is the authority on this, but going back through the thread it appears that the average dwell for the reference has been about 50 days, with longest at 148 and shortest at 9. There seem to be sixteen people still to be serviced in the current cue. Some of the longer delays came when Cellularmitosis, the club starter encountered a series of health, work and other issues, but there have been long recent delays also.
I think the goal was for the dwell to be about 14 days.
-
Vindoline is the authority, but as a guess this is who is left based on an older list:
edavid
CalMachine
Conrad Hoffman
VintageNut
vindoline (CEO of the club)
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
JBeale
Grandchuck
t1d
-
Please correct me if I'm wrong but old list did not stop after t1d.
-
TiN has been added and one other person ... NOT OFFICIALLY. I am not the keeper of the list.
-
Do we have an up to date list of participants?
Here is the current membership list:
Vgkid
kj7e
martinr33
nikonoid
Svgeesus
bitseeker
Vacuuminded
hwj-d
rhb
RandallMcRee
orin
ArthurDent
edavid
CalMachine
Conrad Hoffman
VintageNut
vindoline
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
JBeale
Grandchuck
t1d
VNUTDENYER
KK6IL
Mark03
maxwell3e10
dr.diesel
GEOelectronics
flittle
dr.diesel, flittle and Geoelectronics asked to be skipped this go-around. rhb is up next.
I am not sure how far down the list we are at the moment
-
Vindoline is the authority on this, but going back through the thread it appears that the average dwell for the reference has been about 50 days, with longest at 148 and shortest at 9. There seem to be sixteen people still to be serviced in the current cue. Some of the longer delays came when Cellularmitosis, the club starter encountered a series of health, work and other issues, but there have been long recent delays also.
I think the goal was for the dwell to be about 14 days.
In which case I did not do as badly as I thought. A few days over target IIRC. Sad thing is now I have a 3457A/44492A. With it I could have hooked up all my references and compared them to the PX & FX at the same time. That will have to wait for round 3, but by then I'll be moved into a fully shielded 12' x 16' lab with DIY RF absorbent on the walls.
The walls will be soldered 26 gauge galvanized steel, covered with sheetrock and sprayed with mixtures of sheetrock mud, charcoal and powdered iron if I can find it cheap. I plan to start with a minimum mud mix and gradually thin it to very little charcoal and iron for the top layer, then paint bright flat white. Power through a 2.5 kVA isolation transformer. I'm hoping it will be very quiet when I am done.
Reg
-
A rack would not work. I could not use the gear if the door was closed. And I'd need at least four 6 ft racks. The 16' room dimension is dictated by the need to have the displays at a comfortable height. The 4284/5As could be above comfortable eye level, but my Tek 7104 and 11801 would not be much fun at an awkward height.
I'm an old man, but I'm not a car nut. I've bought a staggeringly expensive and lavish mid 90's lab for less than a neighbor paid for a 70's Roadrunner in so-so condition. I've never spent a fraction of that on a car. Which is why I can afford the lab gear.
Reg
-
Sorry I've been out of touch for a while people! Currently, the kit is on its way to Conrad Hoffman. After him the list is:
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
JBeale
Grandchuck
t1d
orin
VNUTDENYER
KK6IL
Mark03
maxwell3e10
TiN
niner_007
If anyone isn't on the list and you think you should be, please contact me! Thanks!
-
The kit arrived safely today; what a nice case! I'll be cleaning the test bench tonight.
-
Conrad,
I do hope you will report in some detail on your adventures. I've been looking forward to hearing from you about your experience with the kit.
Some testing with your DIY metrology lab kit would be very useful information.
Have Fun!
Reg
-
charcoal and powdered iron if I can find it cheap.
the cheapest i have seen from browsing around is $1/ 25kg graphite, $10-ish /1kg FeO4. there are also ferrite powders about $16-ish / kg iirc. as usual, the source i browse from is china.
have you started? if not i thought i would persuade you not to do it. instead make a active-climatized equipment rack, far smaller and could be moved to other locations. :-//
I agree with 3roomlab, I have done shielding not one layer but two layer (steel and copper) the improvement with noise etc under 500Khz is not worthy compare to cost.
-
Hopefully I can get some data this weekend. Not sure how to apply my DIY stuff as the null detector was intended as the front end of a battery powered DVM for maximum isolation, so not practical for data recording. I think I have several of the old voltage references and can compare those to their original calibrations. The few I've looked at have remained remarkably stable. Unfortunately TI/LT hasn't made that reference (LT1027BCH) in a metal can for a long time and that was the secret to how good it was. The plastic version was subject to stress and environmental issues. I notice there's now what looks like a ceramic SMT package available that might warrant testing.
My main interest is in how well my three Fluke 731s are doing. I've also got an Analogic 6 digit calibrator that's LM399-based that could use a check-up. Not sure if my equipment is good enough but I'm also interested in measuring the noise level of the reference. I've never considered the 731s as quiet as one might desire and expect the traveling standard to do much better.
-
Unfortunately TI/LT hasn't made that reference (LT1027BCH) in a metal can for a long time and that was the secret to how good it was. I notice there's now what looks like a ceramic SMT package available that might warrant testing.
Hello,
I fear the SMT ceramic package is not better than the plastic package.
Unfortunately I never had a LT1027 in metal can package.
I would rather put my money on some AD586LQ:
https://www.eevblog.com/forum/metrology/t-c-hysteresis-measurements-on-brand-new-lt1027dcls8-5-voltage-reference/msg1332775/#msg1332775 (https://www.eevblog.com/forum/metrology/t-c-hysteresis-measurements-on-brand-new-lt1027dcls8-5-voltage-reference/msg1332775/#msg1332775)
with best regards
Andreas
-
charcoal and powdered iron if I can find it cheap.
the cheapest i have seen from browsing around is $1/ 25kg graphite, $10-ish /1kg FeO4. there are also ferrite powders about $16-ish / kg iirc. as usual, the source i browse from is china.
have you started? if not i thought i would persuade you not to do it. instead make a active-climatized equipment rack, far smaller and could be moved to other locations. :-//
I agree with 3roomlab, I have done shielding not one layer but two layer (steel and copper) the improvement with noise etc under 500Khz is not worthy compare to cost.
My primary concern for the room is blocking AM and FM broadcast signals. The stuff below 500 kHz is primarily local devices and hopefully any I have not squelched will be too far away to matter.
For a variety of reasons, I'm going to do the initial build out of the room with just the steel studs, 26 gauge steel covered with 1/4" dry wall and paint and ESD vinyl flooring. I'm also going to cover the bench top with the ESD vinyl flooring. The bench will be a pair of 7' x7' x 4' sections on casters, each with 3 toggle supports that will lift the bench off the casters and level it.
Magnetic fields below 500 kHz really require mu metal or thick, low carbon iron. That's only practical for a small enclosure because of cost. I have one which is still a WIP. At present it's a pair of steel chassis with piano hinge, EMI gaskets and a pull down hasp. Still to be added is the Peltier heating and cooling, the PCB mount, thermal insulation, feed thrus and such. The steel box is 4" x 7" x 11". Once my new lab is set up I'll be able to test various shielding options fairly easily. For example, I can get 1/2"- 3/4" low carbon steel plate to cover the surfaces and measure the performance. Then toss the plate in the steel stock pile and substitute the equivalent amount of mu metal if it proves beneficial.
With respect to internal room reflections, I plan to do TDFD modeling to evaluate room noise with various levels of attenuation on the walls and also to test the effect of building the room out of square.
I found powdered iron for $7/lb which is 1/3 of what I found before. But it's still expensive enough to merit testing the actual performance before spending the money. While I can't get accurate measurements of dielectric and permeability constants with a DIY fixture, I can make relative measurements which will suffice.
Henry Ott's EMC book has lots of data from experiments he made on the effectiveness of various shielding materials. His writing is atrocious, but there is a lot of good information.
Reg
-
Well, nothing happens as quickly on the bench as it does in my mind. Ten minute jobs turn into hours! I finished the last thing that was on the bench and shoved everything aside to make some space. Yes, it's still a mess. I had forgotten the LTZ reference needs bipolar supplies. My normal supplies are the big fan cooled Kepcos in the center of the photo, brought out to the Fahnestock clips. I don't like leaving those on unattended, though no doubt they can handle 24/7 operation. Instead, I modified an L-frame supply from 15 down to 12 volts; quiet and unobtrusive. That's the rack mount thing on the left.
You can see my three Fluke 731s at the upper right. The best one is actually a 731A that I tweaked for minimum TC. The other two are the B version. My 845 is a bit noisy so I'm using an HP3478A for a null meter as the lowest scale is extremely sensitive. With quick lousy wires and immediately after power-up, the Fluke was about 24 uV below the LTZ. IMO, pretty amazing. To the right of the LTZ is the original reference from the Mini-metrology article, and another one found at a hamfest on top of the far 731. I'll check those after they've warmed up a bit. Now that I know things are working, I'll improve the wiring and try to set up some data-logging. I'll also try the other reference.
I heat with wood in the next room and control the lab temperature by how far the door is open. Things are much better in warmer weather, but with some attention I can keep it around 23C. The room is dehumidified but this time of year it will sit at about 40-45% on its own.
Hopefully more to follow over the next week. (edit- replaced the wires with clean copper and matched bananas at the meter. Difference dropped to 7 uV!)
[attachimg=1]
-
Mid-week progress report. Everything has stabilized and I've collected some data, but I need to add code to my data logger software to get the temp and humidity info from the very nice sensor unit. My Fluke reference, though I did my best to lower the TC some years ago, is no match for an ovenized unit. I also did a comparison with the original Mini-Metrology unit from about 15 years ago. Remember, that used common 1% metal films with some selection. At the time I had it checked by the local cal lab, since defunct, and it was near dead on to 10V. It's drifted about 75 uV since then, which isn't too bad for such an inexpensive circuit. Hopefully I can get my data logging software fixed up this week and do some runs over a wider temperature range when the weekend comes. I was hoping to get the unit back on the road within a week, but a few more days would be useful. How anxious is the next user? ;D
-
I can't speak for the next guy in line, but I am downstream and say that if you are actively working to get data go for it.
-
I well understand how Conrad feels. I had my turn already, but I think that Conrad's prior work in the field merits his taking as much time as needed for anything he chooses to do.
And I'd love to see an eval of the DIY kit from the papers.
Go For It!
Reg
I've been thinking more about it. And I can not think of a more valuable effort than to take the time to characterize the mini metrology lab to the extent that the PX and FX references provide additional accuracy.
-
[..]
My 845 is a bit noisy so I'm using an HP3478A for a null meter as the lowest scale is extremely sensitive.
[..]
Hmmh, my 3478A is pretty noisy in the 30mV range, the least two digits are fairly random. I thought that was to be expected, as its specifications state that it'll be accurate to within 250ppm + 40 digits 24h after calibration. I realize now, that the manufacturer's specification tell very little about the noise of an individual unit (some 30 to 40 years after production). Never bothered to actually measure the distribution, e.g. using a short as in https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/. (https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/.)
Perhaps you can add test results from yours to that list? I don't have access to mine at this time (and I'm afraid I need to replace some aging caps).
-
I can and should do that! The question that always comes to mind is, do you get the same noise numbers for a short, as for a "perfect" noiseless voltage? I can think of many reasons why they might be different. It might be possible to watch the slow discharge of a big film cap to find out. Other ideas?
I'm still working on my data-logging software but should have that done soon, then I can try a long run over a large temperature range.
-
I can and should do that! The question that always comes to mind is, do you get the same noise numbers for a short, as for a "perfect" noiseless voltage? I can think of many reasons why they might be different. It might be possible to watch the slow discharge of a big film cap to find out. Other ideas?
[..]
Yes, auto-zero might have an influence for one, but at least everyone has access to a short. I once had my Datron 1271 connected to a alkaline battery -- there was little noise, but also little stability: with a long scale DMM, the self-discharge rate gets in the way; that might not be such a problem on a 5.5digit multimeter. Wished I had a saturated standard cell, but that might not do any good for the low ranges either. I'm afraid 0V from a short is the best I can come up with.
Well, how about two unlike metal junctions, one in ice water, the other at room temperature (protected from draft)? While not all that well defined, at least over short time (seconds) the resulting thermal EMF should be fairly noise-less.
-
Just finished the mods to my primitive data logging software so it gets temperature from the Cal Club unit, plus fixed some bugs. Hopefully it will still work when I put it on the lab PC in the basement, then I'll start collecting data. My HP3478A seems pretty good. I don't see any downside to leaving auto zero on, and since the goal is to measure long term stability, that seems reasonable. I'll also plug in the Mini-M article null front end and see if that improves things or makes them worse. :phew:
[attachimg=1]
-
Getting near the end here. Hopefully will ship the standard kit later this week, but will still need some time to ponder the data. Most of my data logging just reflects my aging equipment, not anything to do with the standards, so may be of limited interest. I'll start with the ancient Mini-Metrology project voltage reference. It's hard to believe I did that 24 years ago, but I have the original reference that appeared in the photos, plus one found at a hamfest from the same time period. Both are now owned by a local friend.
The first was originally measured in 1996 by a local cal lab and tagged at 9.999948 VDC. 24 years later it measures 10.000125 VDC. Maybe the chip drifted. Maybe the inexpensive metal film resistors drifted, but IMO, that's not bad. The second one was measured by me, so it would have been against an ovenized standard cell bank; it was tagged at 10 VDC even. It now measures 10.000334 so also not too terrible at 2+ decades. I didn't build the second one so no idea if the resistors were thermally matched or how good the parts were. Interesting that both went up; maybe chance or maybe that's how the LT1027 ages?
My three Fluke 731 references proved to be interesting. My "golden" 731A turned out to have a hum issue. I don't know if it's normal or not, but the other two 731B units are much better. It would be interesting to know why Fluke decided to do design changes and come out with the B version. All have a significant amount of high frequency noise and I'll post some scope photos soon. The MML references are similar. The FX reference is much quieter than anything I own without filtering, though a 10 kHz filter puts everything on a more even footing.
The FX reference is 9.9999585 at my lab temp, so I set my three Flukes at 41 uV higher. I'm thrilled if things are within about 2 ppm and noise prevents me from doing much better. Even my HP3455A was still within about 5 ppm from its last cal, which was over a year ago using the old SV traveling standard.
More soon!
-
Hum on the 731A is not surprising - it depends entirely on the power supply rejection of the op-amp. It sort of uses the battery as a smoothing capacitor by supplying it through a 510 ohm resistor while at the same time pulse-charging the battery with un-smoothed rectified AC.
If the battery has been removed from a 731A, it needs a capacitor in its place. I guess an old tired battery with high internal resistance could be problematic too!
They improved things with the 731B, but not completely. They still pulse charge the battery and when it is fully charged, depending on the component tolerances, the pulses can get through the diode switching they use from battery/AC supply.
-
I replaced my batteries with a capacitor and zener combination that imitates the NiCd it originally used. Seems to work well but I never noticed the difference in hum with the B version. In theory the hum averages out, but I wonder how perfect that is? Now I'm thinking some power supply mods are in order.
-
I didn't had the FX by me. So, i'm interested to the PX and the resistors, to compare them with my setup. Thanks Conrad
-
Now that I have several references calibrated, I'll make a quick check of the PX against the FX. Not as good as people with 7.5 or 8.5 digit meters, but just the 6.5 digit.
Looked at the schematics for both the 731A and B and wow. I just assumed they were the same, but far from it. Little time to post as work is busy; have patience!
-
Now that I have several references calibrated, I'll make a quick check of the PX against the FX. Not as good as people with 7.5 or 8.5 digit meters, but just the 6.5 digit.
Looked at the schematics for both the 731A and B and wow. I just assumed they were the same, but far from it. Little time to post as work is busy; have patience!
Wow to be sure!
Do you still have NiCads in the 731B?
Orin.
-
No NiCads in any of them. They don't seem to care much about power-up/down, so I replaced the NiCads with capacitors in parallel with zeners, to imitate how a NiCad works. I'm thinking the A version needs a proper regulator. Seems an easy mod, but that's what I always say prior to actually doing it.
-
Bad road conditions today so hopefully the cal kit will see the post office tomorrow. Couldn't go to work, so I'm building a voltage regulator for my 731A. Actually, I could have gone to work, but the driveway was drifting in so fast I wouldn't be able to get back in.
-
OK, ref is on the way to *technogeeky. Now I just need time to post some results.
-
There isn't much of anything I can contribute concerning the FX reference accuracy or stability. It's an order of magnitude better than anything I own. Something that does interest me is noise. I know that that's been investigated too, and it's darn good. Still, it might be interesting to do a visual comparison with some other references. For my audio work I have a Tek 1A7A plug-in that goes down to 10 uV/div. Also have a similar one for my 7000 mainframe, but the 1A7A is my "daily driver." Warning- all my photos are hand-held and you'll see some intensity changes from the trace speed and shutter speed.
First, the 1A7A with a shorted input:
[ATTACH=5]
Next, the FX reference:
[ATTACH=6]
Next, my Fluke 731B reference:
[ATTACH=1]
The 1A7A is a 1 MHz plug-in and the 731B has some HF noise. Here it is with a 10 kHz lo-pass filter:
[ATTACH=2]
The real surprise is how bad my 731A is. That version had no voltage regulator and no internal floating guard:
[ATTACH=3]
Finally, the Mini-Metrology LT1027 based reference:
[ATTACH=4]
Pay attention to the vertical sensitivity labels, as I've used different scales as needed.
I've just completed a voltage regulator for the 731A. That helps a lot, but it still suffers from no internal guard. More on that later. I'll post some voltage comparisons shortly.
-
I did the Mini-Metrology articles 24 years ago! The original reference hasn't been touched since then AFAIK, and was 16.8 ppm higher than the FX. FX is about 4 ppm low, so we'll call it 12.8 ppm over ten volts. A friend of mine found an MM reference at a hamfest and bought it. It was in a plastic box and had my cal sticker from 24 years ago. I got the sign backwards on my previous posting of these. It was actually 39 ppm lower than the FX, so call it 43 ppm below ten volts. I readjusted both #1 and #2 and did some brief data logging. This was before I got temperature added, but I know my lab will typically change by 5 degrees F over a day or night- I heat with wood so control this time of year is iffy at best. FWIW, here's a screen grab of the first ref in Excel:
-
I didn't spend much time with the PX reference, other than to warm it up and compare it with the FX using my now-calibrated HP3455A antique steam powered voltmeter. I put the PX ref at 7.04565 volts.
-
Here's my 731A with its new voltage regulator. Note the vertical scale is 10 uV/div and 1 kHz bandwidth:
-
Here's my 731A with its new voltage regulator. Note the vertical scale is 10 uV/div and 1 kHz bandwidth:
Is it almost beating the 731B now?
Stunning improvement...
-
One of my B units is noisier than the other and one has just detectable popcorn noise. I need to see if it's the same unit, then decide if the problem is the reference or the opamp. You can guess which one I'm hoping for. The A unit is now excellent, save for the lack of an internal shield/guard to prevent hum pickup. It doesn't seem to affect accuracy but I can't believe it's a good thing. The "guard" connection on the front simply connects to a winding screen in the power transformer. I've no idea what that's supposed to accomplish as connecting it never seems to affect anything at all. I extended the transformer secondary wires and used my own bridge on the regulator board. That insures that the AC loop has no conductors in common with the DC loop. The reg is just an LM317 but most people don't realize how quiet they can be if you use a cap on the adjustment pin. Here's what the mod looks like- it occupies the space where the batteries went.
-
I didn't spend much time with the PX reference, other than to warm it up and compare it with the FX using my now-calibrated HP3455A antique steam powered voltmeter. I put the PX ref at 7.04565 volts.
At first, thanks Conrad.
But, hmm, thats a pitty. The PX reference is probably no longer taken very seriously besides the FX. But it would have been interesting to compare the noise behaviour. Also the resistors, no measurement?
-
I would have liked to do more but winter mostly has me carrying firewood and working. Time is short, and I didn't want to keep the kit for weeks on end. Data logging proved more troublesome than I thought it would be, so I lost a lot of time to that. Also lost time figuring out the best comparison setup as my 845 is noisier than I remember it being. Note to self- test and fix everything well before getting the kit. Note to others- see note to self! You're right, the FX has overshadowed the PX and I should have done more. In hindsight, there are also some different comparisons I should have made and more scope photos that would have been useful. I have a bit more data to ponder and maybe post, but it's more related to using the HP3478A as a null meter and the Mini-M null meter.
-
I am in possession of the USA Cal Club kit!
I won't have it long, my best multimeters are only 5.5 digits but it affords me the opportunity to test the calibration at all, so for that I'm very appreciative.
-
...Time is short, and I didn't want to keep the kit for weeks on end. Data logging proved more troublesome than I thought it would be...Note to self- test and fix everything well before getting the kit...
Trying to do all the right things in a timely fashion is quite a challenge and we all learn a lot in the process. I'm certainly thankful we have the opportunity to try it. :-+
-
> The "guard" connection on the front simply connects to a winding screen in the power transformer. I've no idea what that's
> supposed to accomplish as connecting it never seems to affect anything at all.
When we were trying to build a differential multimeter at Wavetek back around 1970, we had the transformer secondary for the voltmeter inside a foil shield. This was to prevent capacitive coupling from the primary, which could have put a common mode voltage between the supply and earth.
Maybe look at your 731 output with respect to earth with your 1A7 and see if the guard makes any difference.
Guess I should look at my LM399 standard for noise. Love the 7A22 plug-in for low level signals.
-
I understand the use of a guard in AC bridge work, as coupling to the primary can be an error source, but I don't see how it matters for DC metrology. I've got a couple 7A22 plug-ins and they're very good, though one of mine fell apart! Several of the many trimmer capacitors popped apart during storage in my unheated garage. I had to carefully desolder them and fix the retainer. Can't remember, but I think I soldered it to the post.
While using the FX standard and various null methods, I was always careful to note if third terminals were floating guards or grounded shields. With several pieces of equipment, some of which might not be floating (scope plug-in), it would be possible to short out the reference.
I'm not comfortable with it, but it seems that large amounts of hum still average to zero and don't affect the results much. IMO, everything should have floating guard systems.
-
hwj-d
I'll put PX thru paces, when kit make it's way to me. :box:
-
Low noise is clearly good, but how good is good enough is slippery. A low pass filter does not harm the measurments of something that in principal does not change. The two principal constraints on the filter are settling time (you don't want to wait days for a stable reading) and drift ( you don't want the filter to mask real variations). Neither constraint seems to prevent filters with fractional Hz bandwidth. Now it is only component constraints.
-
hwj-d
I'll put PX thru paces, when kit make it's way to me. :box:
Yes, I expect so. You'll probably be the first with your elaborateness. But let the px still live, in your rating ;) ;D
-
After a bit of a wait, I am approaching the top of the cal-club queue. I thought I would start to plan for its arrival.
Relevant available equipment:
- HPAK 34401A 6.5 digit meter
- HPAK 34970A with 34901A MUX, 6.5 digit meter, and thermocouple
- Datron 1081 7.5 digit meter
- HPAK E3631A Power supply
- HPAK 85357B GPID adapter
- Old laptop for logging.
- Homebrew environmental logger (temperature, humidity, and barometric pressure).
- I also have a 3499B and various switch cards but do not plan to use them as I think they will degrade accuracy.
Pre-arrival plan:
- Rearrange gear to put the necessary equipment together and far from the every-day bench
- Wire up gear
- Review and update logging programs
- Set up and practice logging with 0-ohms/0-volts and stand-in (house) standards.
Upon arrival:
- Post a message to this list to keep other members informed
- Allow cal standards to warm up to room temperature
- Turn on test equipment for 12 hours
- Re-read standards kit contents and wiring documents
Data acquisition:
- Post periodic updates to this list
- 4 wire measurements of each resistor using the 3 meters sequentially (i.e., resistor hooked to only 1 meter at a time) for 2 hours
- Log Px for 24 hours with standard wired to 3 meters in parallel
- Log Fx for 7 days with standard wired to 3 meters in parallel
Bonus:
- Measure standards using various less precise meters
- Experiment with the Datron guarding signals
- Attempt some noise measurements
After acquisition:
- Package and return the standards.
- Try to produce some preliminary pretty graphs
- Post a message to this list
- Write a report and upload it and the raw data to the Xdevs site.
-
No battle plan survives first contact with the enemy, but that sounds pretty good!
-
No battle plan survives first contact with the enemy, but that sounds pretty good!
;D
-
No battle plan survives first contact with the enemy, but that sounds pretty good!
"Everybody has a plan... until they get punched in the face." -- Mike Tyson
-
- 4 wire measurements of each resistor using the 3 meters sequentially (i.e., resistor hooked to only 1 meter at a time) for 2 hours
- Log Px for 24 hours with standard wired to 3 meters in parallel
- Log Fx for 7 days with standard wired to 3 meters in parallel
Resistance measurements for 2 hours barely enough to see any correlation. I'd suggest interleaving resistance tests with voltage tests. I always try to run resistance tests for days, to battle hysteresis effecte. Using PTFE wiring and robust connections is also important for last bit of accuracy. TBH connections for DUT resistors in kit are inadequate for even 5.5-digit DMM, but it's what it is.
Logging PX or FX in absolute direct measurement also hinders your measurements a lot, as both references are much more stable than any 6.5 digit DMM.
Measuring difference between PX and FX on 100mV DMM range will give you much better data for stability test. Then you can use calibrated value of FX to estimate drift of your DMMs alone over time period. After you know both errors from reference (e.g. settling after transport/power on/your ambient temperature variations), and from DMM (own tempco, noise, short-term stability) you can use obtained data to get rough calibration accuracy of each DMM.[/list]
-
Just for notation, the Fluke 732C had been made with Vishay resistor networks, for good reasons, they have been replaced with precision wire wound resistors again.....
-
TiN: thank you very much for the suggestions.
I can interleave resistance measurements with voltage for the week by employing just the 34970a without degrading precision. I could arrange for all the meters to measure it but it would require an additional matrix switch.
I'll go ahead and order some PTFE wire. I was planning on using cat5 wire.
My understanding is that the PX is nominally 7V and the FX is 10V. On the FX schematic the 7V is available but unbuffered and I certainly don't want to open the case. I'm not certain how to wire them up to provide a difference of less the 100mV.
-
Ah, my bad, I forgot that PX is direct unbuffered output. 100mV range is useless then. Cat5 cable is usually fine for 10V measurement, if your environment around references is quiet (no PWM LEDs, no soldering irons, no other RF/EMI generating stuff nearby). For 10k resistance at 6.5 digits PTFE is not really important, but always good to have.
I can interleave resistance measurements with voltage for the week by employing just the 34970a without degrading precision. I could arrange for all the meters to measure it but it would require an additional matrix switch.
It's just few resistors and DC refs, I'd just do swaps manually, no need to worry about matrices.
-
I haven't found that PTFE makes much difference as most wires are supported in air anyway. PTFE requires that the wire be silver plated. I get my best results with plain untinned solid copper wire- stripped out of phone pairs or what they sell as "bell wire". Wire directly screwed in the jacks if possible. I've gotten my worst possible results with shielded cable as the center conductor and shield are often different materials- thermal emf city!
-
Here is the current setup using stunt doubles for the FX, PX and resistors.
[attach=1]
I’m glad that I started the setup and practice logging in advance of receiving the reference standards. Despite having GPIB logging running on another computer it still took some fussing to get it running on a different laptop.
I also learned some lessons that were not obvious (at least to me) in advance:
- Turn off energy saving on the computer so that it doesn’t stop logging.
- Don’t close the lid on a laptop
- Don’t turn off the master lab power (this is a safety habit; I taped over the switch so I would stop turning it off).
My current plan is to log 3 different setups. All 3 setups will log:
- Time
- Ambient Temperature
- Humidity
- Barometric pressure
- HPAK 34970A measuring
- Temperature near the references
- Voltage of PX
- Voltage of FX
Setup 1:
- HPAK 34401A measuring 1k resistor (all resistance measurements will be 4-wire)
- Datron 1081 measuring 10k resistor
Setup 2:
- HPAK 34401A measuring 10k resistor
- Datron 1081 measuring 1k resistor
Setup 3:
- HPAK 34401A measuring FX
- Datron 1081 measuring FX
- HPAK 34970A measuring (in addition to the above):
I plan to run setup 1 for 24 hours, setup 2 for 24 hours, and setup 3 for 7 days. There are, of course, much more that could be measured but I want to keep the standards for about 10 days and get it moving to the next member.
-
Haven't seen any new msgs here for awhile.The standard still
floating around or did COVID halt progress?
-
Haven't seen any new msgs here for awhile.The standard still
floating around or did COVID halt progress?
I can't imagine that the virus crisis is doing us any good...however the club is still going strong! Currently the kit is with technogeeky.
Here is the current lineup:
technogeeky
TWMIV
Muxr
SirAlucard
MaxFrister
CatalinaWOW
JBeale
Grandchuck
t1d
orin
VNUTDENYER
KK6IL
Mark03
maxwell3e10
TiN
niner_007
If anyone isn't on the list and you think you should be, please contact me! Thanks!
-
Hi Vindoline,
what is the status of measurements. Why no postings about here in this group. Some calclub members posts comprehensive announcements, and then, nothing. Wasn't it the real point to at least post summaries here?
Thanks for your work. :-+
-
Maybe post when getting or sending kit would allow planning by upcoming members (borrowing gear, Covid sheltering, lab upgrades, etc.).
-
Has anybody been able to download all of the data from the folders on the Cal Club FTP site established by TiN? I can download some of the folders but not all. :-//
-
I think life keeps getting in the way of everyone, for sure COVID has something to do with it. However I do like people that get the kit to post either their findings or how they are testing.
-
It takes a lot more to do than you expect until it's sitting in front of you. Then it's pretty intimidating with a long list of people waiting on you to finish.
I think there is huge value in how different people approach the project and the results they get.
Reg
-
I know a lot of you guys are in the camp of "I want to measure this over a long period of time and verify my findings with everyone else". I'm more in the camp of "I want to verify if my equipment is measuring close enough to everyone else and just send it along".
-
Everybody has different needs. Those of us with elderly 6.5 digit meters mostly want to get everything close to calibrated and probably can't contribute much to the stability studies. I could have done that long ago when I had ovenized standard cells, but got rid of them in an effort to reduce the amount of mercury in the lab. The cells themselves were trouble-free, but I spent a lot of time and money on the battery backup system. Never met a gel-cell that lasted more than a few years.
-
I'm more in the camp of "In between a short statement about the state of things, doesn't hurt."
-
During the brief period Wavetek was trying to produce a 0.005% differential multimeter (DC/AC V, Ohms Hz, Ratio), we had a bank of 10 standard cells in a lead bathtub inside a polystyrene insulated container. Depended on that to slow down and reduce temperature shifts. Meter references were a pair of ~6.2V zeners with a +TC one matched with equal -TC. What a PITA those were to make.
My needs are modest - 10V to calibrate my pair of 3456's and LM399 standard, and a 10K to verify my 10K L&N standard is still close.
-
Indeed, it would be good to post status updates at least, what, once a week, while you possess the kit. Performing a long study is OK. At least let everyone know you're still OK (and what you've experienced thus far would be interesting). 8)
-
Indeed, it would be good to post status updates at least, what, once a week, while you possess the kit. Performing a long study is OK. At least let everyone know you're still OK (and what you've experienced thus far would be interesting). 8)
Seconded.
-
As for comparing the PX and FX outputs, I had planned on adding selected 3 of 8 standard cells in series to get difference below 60mv (unless consensus is that would be harmful to PX OR standard cells). Hope to add more gear before time at plate. :-\
-
Maybe should have been a question. Is it ok to series up by best 3 standard cells with PX and monitor difference to FX? This would only be after logging noise/stability of cells, E1412A zero, etc. Thanks for any tips.
-
So @technogeeky has had the hardware for 3 months now?
-
I don't see any issue with putting standard cells in series with an LTZ1k reference, since the standard cells are floating. Of course the usual caveat of limiting current draw from standard cells applies, so you may want to check the bias current and impedance of your DMM.
-
I am in possession of the USA Cal Club kit!
I won't have it long, my best multimeters are only 5.5 digits but it affords me the opportunity to test the calibration at all, so for that I'm very appreciative.
I can't imagine that the virus crisis is doing us any good...however the club is still going strong! Currently the kit is with technogeeky.
This is quite disturbing. Since the 10 February post technogeeky has posted once on 11 April. Based on his 10 February post he should have sent it back long before then.
I sent a PM to vindoline asking if he had heard anything from technogeeky before I looked into technogeeky's posting history. Hopefully he will respond here.
If someone misaddresses the package when returning to vindoline it can be lost forever. And packages do disappear in transit with all carriers even when correctly addressed.
Hopefully the kit will turn up, but I'd like to suggest some procedural changes:
Preprinted shipping labels. Use the USPS service to print a return label and an outbound label and email the outbound label to the person getting the kit for verification before shipping. Put the return label in the kit. I think that will assign a tracking number for each leg so it can be tracked even if mischance befalls the person getting the kit.
Place a label on the kit describing it and where it should be sent in the event the person who has it passes away. Covid-19 makes that a much greater risk than before. At present no one dealing with the estate of the person who has the kit would have any clue what to do with it.
Require insurance when shipping.
It might be a good idea to require a security deposit via PayPal to be refunded upon return of the kit. I don't think anyone would sign up to the list to steal the kit, but it's possible. There could also be a penalty factor for delays in turn around not justified by work presented as a means to deter sloth.
Reg
-
I am in possession of the USA Cal Club kit!
I won't have it long, my best multimeters are only 5.5 digits but it affords me the opportunity to test the calibration at all, so for that I'm very appreciative.
I can't imagine that the virus crisis is doing us any good...however the club is still going strong! Currently the kit is with technogeeky.
This is quite disturbing. Since the 10 February post technogeeky has posted once on 11 April. Based on his 10 February post he should have sent it back long before then.
I sent a PM to vindoline asking if he had heard anything from technogeeky before I looked into technogeeky's posting history. Hopefully he will respond here.
If someone misaddresses the package when returning to vindoline it can be lost forever. And packages do disappear in transit with all carriers even when correctly addressed.
Hopefully the kit will turn up, but I'd like to suggest some procedural changes:
Preprinted shipping labels. Use the USPS service to print a return label and an outbound label and email the outbound label to the person getting the kit for verification before shipping. Put the return label in the kit. I think that will assign a tracking number for each leg so it can be tracked even if mischance befalls the person getting the kit.
Place a label on the kit describing it and where it should be sent in the event the person who has it passes away. Covid-19 makes that a much greater risk than before. At present no one dealing with the estate of the person who has the kit would have any clue what to do with it.
Require insurance when shipping.
It might be a good idea to require a security deposit via PayPal to be refunded upon return of the kit. I don't think anyone would sign up to the list to steal the kit, but it's possible. There could also be a penalty factor for delays in turn around not justified by work presented as a means to deter sloth.
Reg
Hello folks;
I didn't use the kit at all for the first two weeks I had it due to work constraints; I used it to calibrate (or rather check the calibration) on a few meters, used the resistors to check calibration too. Unfortunately my meters are only 5.5 digits so I mostly just saw a completely stable reading.
On March 31, I replied to a PM saying how it was unfortunate this COVID craziness happened during this; I suggested we wait 5 days on either side of shipping (after boxing but before shipping; after receiving but before use) for safety reasons. I said I would send the device out if I was given an address and was told to "sit tight".
Unfortunately all of this distress was caused by me talking via PM and not in the public thread. Sorry about that.
Please let me know where to ship the device and I'll get it in the post this weekend.
-
technogeeky! Glad to see you're alright. Thanks for the update. Hope vindoline is around to get the address to you. I thought he included a sheet with it printed on.
-
Quick question, are there any of those GPIB kits left? I need one. Or know a good place for one?
-
Here's the Uno version I made using IDC connectors. All from ebay. The case was around $9 for 10.
Soldering the wires was a pain which is why I asked artag to design the Uno shield version.
https://oshpark.com/profiles/artag
Have Fun!
Reg
-
technogeeky! Glad to see you're alright. Thanks for the update. Hope vindoline is around to get the address to you. I thought he included a sheet with it printed on.
Nope, the box included only a printed copy of my own address! I will send it out once I have a place to send it to. Hopefully it won't be long!
-
technogeeky! Glad to see you're alright. Thanks for the update. Hope vindoline is around to get the address to you. I thought he included a sheet with it printed on.
Nope, the box included only a printed copy of my own address! I will send it out once I have a place to send it to. Hopefully it won't be long!
PM sent
-
The cal kit is on its way to vindoline. Thanks a bunch! I checked the calibration on almost 40 year old multimeters and they are only 1 to 2 counts out on the ranges I could check. Pretty amazing!
-
In relation to " ATmega 328-based USB-to-GPIB adapter" in kit, any links or advice on SW for logging Arduino temp and DMM data? Assume still best in CSV format for crunching. Thanks.
-
What OS are you using? What meter?
I wrote Unix scripts to do it using as pair of 34401As. Don't even recall which system I used, so a bit of a chore to find it.
It's just sending a series of strings to a serial port and logging the replies. Pluse changing the device address.
Reg
-
In relation to " ATmega 328-based USB-to-GPIB adapter" in kit, any links or advice on SW for logging Arduino temp and DMM data? Assume still best in CSV format for crunching. Thanks.
If you haven't yet seen it, check out the cal club's getting started guide (https://www.eevblog.com/forum/metrology/usa-cal-club-getting-started-user-guide/). It includes links to the AR488 project and the manual, which has info on installation, drivers, client software, usage, etc.
https://www.eevblog.com/forum/metrology/usa-cal-club-getting-started-user-guide/ (https://www.eevblog.com/forum/metrology/usa-cal-club-getting-started-user-guide/)
-
Just the DMMs and DC sources in sig. Figured on WIN7pro but could us VB or Linux. Is AR488 what is in kit? I need to build up AR488 or similar and get everything working before kit arrival (guessing Sept).
-
Is AR488 what is in kit? I need to build up AR488 or similar and get everything working before kit arrival (guessing Sept).
Hi, there is a GPIB to USB adapter in the CalClub kit that is running the AR488 firmware. For a single DMM you should be able to just plug it into the GPIB port in the back. No GPIB cables are included for daisy chaining a series of instruments. On your end you will need a PC with software to control and read the data from the AR488. I think most people use a Python script to do the data logging. I have a dedicated raspberry pi that I use for this.
-
Thank you rhb, bitseeker, and vindoline for helpful info. I used the long name for GPIB adapter in kit before rereading kit contents. Just trying to figure cycle time so can decide whether to function test what I have or get more (looking at Fluke 883A tomorrow - powers up. Later.
-
Seems the general consensus is that if you're doing interesting things with the kit and updating here, a little extra time to be able to accomplish those things is OK. If you're unsure, feel free to post a prioritized list of tasks and rough timeline for group input. We're a pretty laid-back bunch, which helps keep the club resilient.
-
Alrighty! Giving an update to let everyone know the kit has arrived safely, I don't have much to test, so it won't likely be with me for long. As I'm mostly looking to just calibrate what I have and sent it along.
EDIT: Quick note, I noticed the arduino with the temp sensor has a resistor that snapped off? I'm assuming this isn't suppose to be the case. Please correct if I'm wrong, tho I wouldn't see why lol.
Then again it does looks like it's this way on the getting started page, I must have missed why this was.
-
Oh, wow. Can you tell what that resistor might be for? The LED?
-
Oh, wow. Can you tell what that resistor might be for? The LED?
Yea it's just connected to that LED that's right next to it. Doesn't seem to be used tho.
-
Quick update, I won't be shipping the kit back quite yet, picked up a 3478A for under $100 from fleabay. So once it shows up I'll actually have something with some decent resolution to check with and share. Granted it's in any sort of calibration. Slight cosmetic damage, but otherwise looks alright. So will look forward to seeing what kind of condition it's in.
-
Alright, so the 3478A is in my hands, and here come the first few numbers! Feel free to give input if you think they aren't accurate enough. Cause personally I have a problem with wanting things to be exact.
First up resistance tests.
HP 3478A
1k Measurements
2-Wire: .99975 kOhms
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1046560;image)
4-Wire: 0.99951 kOhms
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1046564;image)
10k Measurements
2-Wire: 9.9951 kOhms
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1046568;image)
4-Wire: 9.9949 kOhms
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1046572;image)
Fluke 8012A
1k Measurements
1.000 kOhms*
10k Measurements
9.99 kOhms*
AMProbe AM-520
1k Measurements
1.000 kOhms*
10k Measurements
10.01 kOhms*
I'm going to start running the 7.0456 voltage reference and will report back my findings. So far I think I got a good deal.
*Additional Photos on Server
-
Looking good! :D
-
Alright PX Reference numbers are in! After 12+ hours of measuring on the 3478A, again I think the numbers look really good. Tho I'm biased cause I don't want to think I got a lemon. You guys tell me! This is with Auto Zero turned off. It's been a little while, but it seems the meter has settled, and with Auto-Zero turned on it sits at 7.0459
First off the Temp/RH of my office in which I was doing the testing. It's not the most stable environment, nor the coolest. However at night when I left the office it did cool down to about 25c
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1047278;image)
Then the difference over the measured reference of 7.0456. Mine averaged out to be 7.0462 which makes it only out by .0006, I assume that's good enough to keep?
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1047282;image)
-
What is the rate at which the arduino reports the temperature? It's not quite every second, and it's not every two seconds. Wondering if we can add a rtc to the arduino and have it report back the current time, as well as have it trigger every second. Unless there is a better way to log readings from an HP3478A and an Arduino at the same time into the same file. I'm just currently using HP 3478A Control to log my readings, and putty to log readings from the Arduino. I don't get the same samples, so it doesn't really tell you what the current temperature was when the reading was taken, just the general temperature of the room over time, as much as I try to keep them logging in the same amount of time.
Also with help from TiN I've got the FX Up and running. Short story I don't have a bi-polar linear PSU. So with the help of a battery I got it running, and it looks really good. A tad high for my liking, but I have an obsession with things being exact and perfect. I'm getting anywhere between 10.0000 to 10.0010. Is it recommended to take readings with Auto Zero on or off? I'm guessing off since the readings would be faster.
-
Posting to let everyone know the Cal Kit is on it's way back. I've uploaded all my logs of what my 3478a was giving. Definitely needs to be calibrated to be a bit more accurate. Was able to calibrate at least the 30v range using the FX as a reference, as well as the 1k and 10k Ohm ranges. Which is perfectly fine for me. As those are my most used ranges anyways. Eventually I'll get the whole thing calibrated. For now close enough is good enough for my use case.
-
So after all that, I ended up getting the part I needed to make my own GPIB connector. Personally I like things really tight and tidy, so this is for the person that complained about all the wires. :) Works perfectly!
-
That's tight and tidy all right. Nicely done!
-
Thanks! I really enjoy tight and tidy wiring and soldering jobs. After going through this all I ended up purchasing some parts to make a relatively accurate 1/4w decade resistance box.
After looking back I probably should have gotten the Omron switches instead of the Adafruit ones because the Adafruit ones add 1Ω once it's all together. Each switch was measuring about .17Ω between common and 0. I also would have been ok with 1/10w resistors or 1/8th.
I purchased extras on the resistors so I could match my resistors going up the decade so there wasn't a significant drop between each. So for instance if one was 9.9972k the next would be 10.0013k to round it out as near to 10k as I can. For my 1Ω I used 1% 50ppm, 10Ω .1%15ppm, 100Ω .1% 25ppm, 1kΩ .1% 15ppm, 10k .1%25ppm 100k .1% 50ppm, 1M .1% 50ppm. On another look I should have gone with lower PPM on the 100k and 1M, and definitely a lower wattage cause they were huge, but I made them fit.
Anyways here are some pictures of that build.
(https://i.imgur.com/pSQkwiw.jpg)
Each module ready to be connected together.
(https://i.imgur.com/BgtadnC.jpg)
After wiring each one to each other I had realized what a tight job it was.
(https://i.imgur.com/7PaEttF.jpg)
Quick check on the meter. Showing .007% accuracy.
(https://i.imgur.com/WfShaCX.jpg)
Fit into it's little box.
(https://i.imgur.com/b6nJ9ol.jpg)
And done.
So after doing this project it made me wonder, we have the kit going around, what are the chances of getting more ranges sent around? Would be great to be able to calibrate the entire range of my meter to match that of other's. There should be a full calibration kit, in decades 1, 10, 100, 1k etc with a verified reading on someone's fancy highly expensive already fully calibrated machine. :) At least for resistances. Just a thought.
-
Superb piece of work!
-
That's quite tight and tidy, SirAlucard. Nicely done!
-
Posting to let everyone know the Cal Kit is on it's way back. I've uploaded all my logs of what my 3478a was giving. Definitely needs to be calibrated to be a bit more accurate. Was able to calibrate at least the 30v range using the FX as a reference, as well as the 1k and 10k Ohm ranges. Which is perfectly fine for me. As those are my most used ranges anyways. Eventually I'll get the whole thing calibrated. For now close enough is good enough for my use case.
It has been a while ... any recent information about the Cal Kit?
-
That's quite tight and tidy, SirAlucard. Nicely done!
Yea unfortunately I'm not happy with the 0Ω setting resistance. Each thumbwheels 0 is about ~0.2 ohms over the corse of all 7 thumbwheel switches it adds up to about 1.3ohms giving the whole unit an offset and never having a true 0 ohm end digit. I ended up bypassing the 1 ohm so at least 2 would = 2 instead of 3. I'm thinking about taking it apart and either A) getting some boards made with thicker traces and gold plating to keep the resistance lower and more precise, or B) reflowing a somewhat decent layer of solder over the traces which seems to have improved the low end resistance a bit. I mean I kind of expected to have lower quality switches considering I paid $8 for a pack of 5 adafruit switches, vs $18-$19 for 1 when buying an omron thumbwheel switch which would have taken the price and almost multipled it by 4. Tho I didn't think the low end resistance would have been so bad.
-
I know it would add to the complexity but the "good" boxes I know of, like ESI, were set up as 4-terminal resistances. I have an ESI box that's 2-terminal, but it's not for very low values. I also have an ancient L&N terminal and plunger box for low values that uses very heavy wiring and assumes you'll make a 4-terminal connection to the giant brass binding posts.
-
I know it would add to the complexity but the "good" boxes I know of, like ESI, were set up as 4-terminal resistances. I have an ESI box that's 2-terminal, but it's not for very low values. I also have an ancient L&N terminal and plunger box for low values that uses very heavy wiring and assumes you'll make a 4-terminal connection to the giant brass binding posts.
I am checking resistances with a kelvin clip up until the binding posts, and I have a few extra switches that I've tried modifying and playing with. I was able to take the one and bring it's resistance from .22Ω to .126Ω just by adding a layer of solder to the board, but I still don't think that's the way to go as it's not a super great connection.
-
It's hard to believe any switch could have a resistance that high, if it's clean and if it uses metal on metal contacts. Have you ever disassembled one to see how it's made?
-
Does anybody know anything about the Cal Kit? I have sent a couple of PMs but have not been able to learn anything :'(
-
I haven't heard anything. I assume vindoline received the kit back, but haven't seen confirmation of that.
-
Vindoline has contacted me and the kit is headed my way. His schedule has apparently been very full lately.
-
Very good. Thanks for the update!
-
The reference arrived today. I had my meters on and stabilizing for two weeks prior to a five hour power outage yesterday. They will have had at least a day to stabilize before I start for real. This evening will be initial hookup and test runs, with the real data gathering starting tomorrow. I hope to finish and have it on its way within a week.
-
No plan survives contact with reality. I had planned to use the HP 6626 power supply to power everything, but when I quickly realized that a lot of harnessing work would be required. Switched to a Power Designs supply. Allowed using existing harnesses for the club references and simplified the others. Not totally happy with the noise, there are several millivolt spikes at around 1500 Hz running on the millivolt or so of white noise. Will probably make a battery powered run later if all else goes well. Also ran into problems using the Hp3456 in differential mode between the FX and my KMoon reference. So I just paralleled the HP 3456 with the Hp 3457.
Rather than figuring out the software interface for the club temp sensor I compared the on screen measurements of the club sensor with my own. About 0.1 deg C difference on temperature and 1% RH difference so carried on with my own. I will continue to compare them from time to time, though my lab area is fairly stable in temp this time of year so will probably not get enough information to make a slope correction, just more confirmation on offset.
System is up and gathering data. I'll run 24 hours in this configuration and look at the data before deciding what to do next. Looking at the resistors and initial voltage data it appears my old and lame instruments have held calibration fairly well, but will reserve judgement for a closer look at the data.
I will post some brief data samples here, but most data will be posted in TiN's data site.
Document shows final test configuration and photo shows data logging plugging along.
-
I'm done and the reference will be on its way today. I did do a battery powered run, but haven't analyzed the data to see if it made a difference. All data posted on TiN's server, but ended up marked private. I don't know how to change that, or if that will affect visibility to others.
-
All data is available here as usual (https://xdevs.com/datashort/USA_Cal_Club/Round2/). If it's uploaded, then there is no such thing as private data.
-
The CalKit arrived with a not working FX voltage reference. It is on its way back to the manufacturer (TiN). As can be seen in the photo, the FX is showing only 10 mA and presents a very low output voltage. The PX is good and its output is as expected. More to follow.
-
It will be interesting to find what TiN discovers. I went back and reviewed data and the FX was operating properly until I stopped logging. The final run was made with the reference operating on two 12V lead acid batteries.
1. Logging program was stopped. Shutdown was normal. This means that the HP3457 input mux was connected to the PX reference. The FX mux input on the HP3457 was open. The HP 3456 was continually connected to the FX reference.
2. Power was removed from the references. This occurred within a couple of minutes of the last logging event. There is no information on power sequencing, and I don't remember whether I removed +V or -V first, but both were removed within an interval of a couple of seconds (probably less than a second, but as it involved pulling banana plugs the timing is not documented.) The ground connections were then removed.
3. The FX power cable was removed and packed.
4. The FX output connections were removed and the shorting bar installed.
5. The FX reference was placed in the shipping container.
While none of this was performed at an anti-static workstation, I observed my normal static precautions which involved grounding myself to equipment ground before touching any hardware. Static conditions were not extreme as seen by the RH measurements.
While I don't see anything I did which should have killed the reference, I am looking forward to the diagnosis and will take responsibility for anything which I did wrong.
Hopefully this will lead to another note in the Start-Up guide which will help prevent any future problems.
-
Probably just mechanical issues, as internally there are wires from power BNC inputs and reference output to posts.
Will see, also I'll get an answer on my whole reason donating FX to calclub - what happens to non-rugged LTZ1000A ref when its shipped around and used by random people in random ways for 1 year. Could not ask for more real-life test that this. :-DD
-
Probably just mechanical issues, as internally there are wires from power BNC inputs and reference output to posts.
Will see, also I'll get an answer on my whole reason donating FX to calclub - what happens to non-rugged LTZ1000A ref when its shipped around and used by random people in random ways for 1 year. Could not ask for more real-life test that this. :-DD
Probably nothing a dose of hot snot can't cure! :D
-
Have finished a first run of analysis on my data and came up with a few happy for me conclusions.
1. My HP 3457 still meets its 24 Hr specs for DCV and 4 Wire resistance on ranges checked. Since it is many years since its last calibration this bodes very well for continued long term stability and for accuracy on the unchecked functionality.
2. My HP 3456 wasn't quite as good, but still meets 90 day spec.
3. Both instruments showed small temperature drift over the range seen in my tests. This data must be taken with a grain of salt because the temperature monitor placement was selected to get best approximation of the reference temperature and wasn't well coupled to the instruments. But the overall temperature environment was very slowly varying and relatively stable so there is some good here.
4. The best estimate for my KMoon reference value turned out to be 10.00279 volts, compared to the 10.00276 volts inked on the reference in 2015. Apparently mine was one of the ones that was actually measured, not just written down, and also it appears to be relatively stable over time.
Analysis spreadsheet will be loaded on TiNs site, though I was unsuccessful this AM. Probably a problem with my internet connection - I am in the boonies and I get lots of "special" experiences.
-
Got the box. Autopsy revealed that patient died due to autopsy.
Jokes aside, FX reference still works just fine. I've connected it to powersupply for +12/-12V , hooked meter to it and on power on its drawing expected 42mA and 5mA on rails. Output is about +1.5ppm so far wrt assigned value in May 17, 2019. Live data: https://xdevs.com/usac_fx_as_received_k2002x_nov2020/
Why it didn't work before? It was an operator error, now that I looked closer on the posted photo. I think CatalinaWOW will feel relief that it's not borked on his watch :) I see connected +12V to PSU ch2 output HI and -12V to ch2 LO. Common return wires from either power rail are not connected to anything. So there is no ground provided to reference and no workie as result.
Perhaps instructions on connecting were not clear enough to stress that FX require bipolar power, it cannot work with single +12 supply and was not designed for single supply.
Since I have it here anyway, i will replace bad cracked pomona 3770 posts (very common issue with these low cost posts, sadly) with something better and fully recalibrate the unit. It may take few weeks. Will also donate hard-case for it and make users manual with proper connections shown with photos.
-
Excellent news. I don't know what your expectations for stability were, but 1.5 ppm sounds pretty good to me, so excellent news on that side too.
-
Oh, what a bonehead! :palm:
Thanks a bunch TiN! Very impressive long-term stability!
-
TiN, That's great news! I can see how it may be a bit confusing to correctly power the FX, but this is the first time it has happened. Great news about the stability. Regarding the hard case, the USA cal club kit is always shipped in a custom padded hard case to protect all the gear. No problem if you want to send a "special" one just for the FX though!
-
Reference as received, connected to output spade lug cable. Hard to see from photo, but crappy 3770 posts don't hold connectors well.
(https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_module_1.jpg) (https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_module.jpg)(https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_posts_1.jpg) (https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_posts.jpg)
Connection at PSU shown and DMMs.
(https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_psu_1.jpg) (https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_psu.jpg)(https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_test_nov2020_1.jpg) (https://xdevs.com/datashort/USA_Cal_Club/Round2/TiN_nov2020/usac_test_nov2020.jpg)
Right now sitting in that poor looking styrofoam thermal chamber with watercooled TEC module to sweep thru 18 - 45 °C, being logged by triple 3458A.
-
logged by triple 3458A
Are you sure you have enough 3458A ? What if all of them fail ? Are you sure you don't need a few more to be sure ? ;D
-
logged by triple 3458A
Are you sure you have enough 3458A ? What if all of them fail ? Are you sure you don't need a few more to be sure ? ;D
A man with one watch knows what time it is, a man with two watches is never sure. A man with three watches can pick the best two and calculate a confidence interval.
-
If you assume independence, which is probably a fairly good assumption, particularly if tempco of the instrument is compensated you can average the readings for an improved estimate of the value. And the confidence interval reduces by the square root of the number of 3458s. All you need is 100 to add a whole digit of precision.
-
If you assume independence, which is probably a fairly good assumption, particularly if tempco of the instrument is compensated you can average the readings for an improved estimate of the value. And the confidence interval reduces by the square root of the number of 3458s. All you need is 100 to add a whole digit of precision.
On the bright side, you only need 4 to double the performance!
-
Okay. I saw 3 3458A for sale. Should I buy them all ? >:D ::) ;D
PS: Just kidding. This stuff is way above my metrology skills. But I enjoy seeing it. And yes there are 3 3458A for sale in China if anyone is interested
-
Got the box. Autopsy revealed that patient died due to autopsy.
Jokes aside, FX reference still works just fine. I've connected it to powersupply for +12/-12V , hooked meter to it and on power on its drawing expected 42mA and 5mA on rails. Output is about +1.5ppm so far wrt assigned value in May 17, 2019. Live data: https://xdevs.com/usac_fx_as_received_k2002x_nov2020/
Why it didn't work before? It was an operator error, now that I looked closer on the posted photo. I think CatalinaWOW will feel relief that it's not borked on his watch :) I see connected +12V to PSU ch2 output HI and -12V to ch2 LO. Common return wires from either power rail are not connected to anything. So there is no ground provided to reference and no workie as result.
Perhaps instructions on connecting were not clear enough to stress that FX require bipolar power, it cannot work with single +12 supply and was not designed for single supply.
Since I have it here anyway, i will replace bad cracked pomona 3770 posts (very common issue with these low cost posts, sadly) with something better and fully recalibrate the unit. It may take few weeks. Will also donate hard-case for it and make users manual with proper connections shown with photos.
That's a relief, TiN. Glad the patient is fine after all.
If you have hookup photos, I'll add them to the guide along with more explicit info about power requirements.
-
USA Club revive. ^-^
Reference is happy and ready to go. I have recently revalidated the xDevs Volt, so uncertainty is real nice at the moment :)
According to my measurements this FX reference survived just fine and after all the shipping and labs it drifted up +0.800 ppm since my last calibration 572 days ago. Measurements done against my primary +10 VDC using my trusty golden Keysight 3458A units. Alternatively measurements also verified with Keysight 34420A nanovolt meter and low-thermal DP 160 scanner with support from ManateeMafia. :clap:
Today's data, temperature sweep 20-30 °C:
(https://xdevs.com/doc/xDevs.com/FX/usac/recal/usac_recal_dec_tempco_1.png)
Original data, wider temperature sweep 20-50 °C:
(https://xdevs.com/doc/xDevs.com/FX/usac/final_spec_1.png)
If we make huge assumption that LTZ1000A module was drifting linearly during this time frame, we get 0.800 ppm / 572 days = 1.399 ppb/day = 0.51 ppm/year stability. I'd say that is amazeballs for a reference shipped all around hobby labs.
(https://xdevs.com/static/images/pdf-24.png)USAC FX PDF calibration report (https://xdevs.com/doc/xDevs.com/FX/usac/recal/XVR_USAC_FINAL_DEC_10_2020_AS_SHIPPED.pdf)
Also cooked powering note for better clarity, hopefully it will make stuff easier for new members.
(https://xdevs.com/static/images/pdf-24.png)Power up procedure guide (https://xdevs.com/doc/xDevs.com/FX/FX_Power_directions.pdf)
I have replaced output crappy 3770 binding posts that all go bad fast with nanovolt-grade pure-copper spade lug only terminals. Banana people will surely hate me now, but I don't mind :popcorn: Maybe somebody in USA club can make adapter spade lug -> banana if anyone still wants that. :P
Updated module appearance front side :
(https://xdevs.com/doc/xDevs.com/FX/usac/recal/usac_top_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/recal/usac_top.jpg)
Updated module appearance back side :
(https://xdevs.com/doc/xDevs.com/FX/usac/recal/usac_bot_1.jpg) (https://xdevs.com/doc/xDevs.com/FX/usac/recal/usac_bot.jpg)
:-+
-
@TiN
Hi TiN,
is the PX with you as well? Because i've had the PX only at the beginning, can you please measure the PX with one of your precise 3458A and take a foto from it? I still don't know how to estimate my dmm's.
Thanks a lot.
-
No sir, I got only FX without the packaging/case or anything but cable.
-
@TiN
Hi TiN,
is the PX with you as well? Because i've had the PX only at the beginning, can you please measure the PX with one of your precise 3458A and take a foto from it? I still don't know how to estimate my dmm's.
Thanks a lot.
I will compare the two references when the FX arrives and report my 3458A readings here.
-
@TiN
Hi TiN,
is the PX with you as well? Because i've had the PX only at the beginning, can you please measure the PX with one of your precise 3458A and take a foto from it? I still don't know how to estimate my dmm's.
Thanks a lot.
I will compare the two references when the FX arrives and report my 3458A readings here.
Thanks Grandchuck, that's nice.
-
Something happening here?
-
It is scheduled for pickup from TiN. I will post here when it arrives.
-
It was also recalibrated and relabelled few days ago for fresh volt. I've included shorting dual banana plug back with reference, but obviously now it has no any use. New copper low-thermal posts are not banana compatible :-+
Callout for everyone participated in round 2 already. Please, upload all data to server, so we can plot a measurement agreement between everyone and guesstimate deviation of each person, as now trend of FX reference over 600+ days is guesstimated :)
(https://xdevs.com/doc/xDevs.com/FX/usac/usac_recal_jan_tempco_1.png)
-
The FX arrived safely and works as expected. TiN included a 4 page calibration document. The first page of this document is attached with this post. Pages 2 and 3 show uncertainty error in ppm (not attached). Page 4 has probability histograms: one for UUT- and one for UUT+ (not attached). Of course, the entire 4 page calibration document will be included with the cal club kit when I ship it to the next member in line.
Hats off to TiN and many thanks to him for taking the time for providing this valuable reference to the US CAL CLUB! Not only did he design and build this particular reference, he put a lot of time and effort into verifying its performance … again! :-+
Note to hwj-d: The PX reference is showing 7.045639 volts on my 3458A (recently calibrated to the International TiN Standard [ITS]).
-
Here is some data for the FX reference. 2,692 readings (NPLC = 100) with a 3458A that was calibrated using the FX reference itself (I guess it is what one calls a bootstrap operation). The XL file is also attached. Later, wondered if the data were normal so I prepared a histogram plot.
-
Here is some data for the FX reference. 2,692 readings (NPLC = 100) with a 3458A that was calibrated using the FX reference itself (I guess it is what one calls a bootstrap operation). The XL file is also attached.
Very neat. Is the X axis in Seconds?
-
The X axis is the number of the reading ... from one (the first) to 2,692 (the last).
-
The X axis is the number of the reading ... from one (the first) to 2,692 (the last).
Ok, so one reading every 1.667 seconds (60Hz) with 100nplc? 1 hour and 15 minutes total test time?
I'm trying to understand how stable it is over time...
-
The data collection runtime was about 4 hours. You can see it in the data file (there is a column for time).
-
It's hard to believe any switch could have a resistance that high, if it's clean and if it uses metal on metal contacts. Have you ever disassembled one to see how it's made?
Yea, it's how I was able to throw some fresh solder down to kinda beef up the pad on a test unit, I think my plan is to copy and redesign the PCBs a little bit for these and get them made and just replace all the PCBs in the switches, would be cheaper than getting the nice push button switches.
-
Not able to upload to the club repository?
FTP server: ftp.xdevs.com
Login: usac_run2
Password: same as login, case-sensitive
Anybody?
-
Not able to upload to the club repository?
FTP server: ftp.xdevs.com
Login: usac_run2
Password: same as login, case-sensitive
Anybody?
I had a similar problem. Went and got an FTP server program and the problem was solved. Never took the time to debug what was wrong.
-
Some ISPs block FTP port connections, so you can try alternative port 9876 for connections.
-
OK, the files have been uploaded using WinSCP (a free download). Thanks everyone for the help. :-+
-
OK, the files have been uploaded using WinSCP (a free download). Thanks everyone for the help. :-+
Ah, WinSCP. I use it all the time at work to get logs from servers I maintain. Just note that the remote file view doesn't automatically update as files are added/updated, so use the refresh button often.
-
Thanks,
so the PX is also very good!
As in july 2018:
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1643525/#msg1643525 (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1643525/#msg1643525)
the PX today is bang on!
(and my dmm6500 too :-DMM
:-+
-
Just curious; any recent news on this topic? I've lost track of where I was on the list.
-
Just curious; any recent news on this topic? I've lost track of where I was on the list.
PM sent.
The club is alive and active! Currently KK6IL has the kit. You're next!
-
A box-full of calibration has landed at my doorstep. With my 6.5 digit Keithley 196 DMM, I am feeling unworthy! So it won't stay here too long. So far, I have only hooked up the little Arduino Temp-Humidity monitor, and had a quick look at the resistance standards. So far, so good. Apparently I have not been paying attention, because there are more widgets in here than I anticipated.
temp_c,humidity,crc16
18.996,55.012,b216
18.991,54.854,dafc
18.988,54.687,35e8
18.981,54.532,b9ee
18.988,54.382,8ebd
18.988,54.208,916e
18.979,54.050,5e74
18.981,53.892,73a0
18.982,53.746,a13c
18.979,53.605,2450
18.978,53.492,c79c
18.964,53.381,bbca
-
It's quite the cornucopia of calibration contraptions. And you've got fresh calibration data from TiN. :-+
-
A box-full of calibration has landed at my doorstep. With my 6.5 digit Keithley 196 DMM, I am feeling unworthy! So it won't stay here too long. So far, I have only hooked up the little Arduino Temp-Humidity monitor, and had a quick look at the resistance standards. So far, so good. Apparently I have not been paying attention, because there are more widgets in here than I anticipated.
temp_c,humidity,crc16
18.996,55.012,b216
18.991,54.854,dafc
18.988,54.687,35e8
18.981,54.532,b9ee
18.988,54.382,8ebd
18.988,54.208,916e
18.979,54.050,5e74
18.981,53.892,73a0
18.982,53.746,a13c
18.979,53.605,2450
18.978,53.492,c79c
18.964,53.381,bbca
You, unworthy? Modest I'd accept, but "unworthy" would be grounds for a cage fight conflict. That said, it's pretty intimidating when you open the box.
I'm hoping I'll be able to participate with proper HVAC control in the next round, but it's been going far too slowly to suit me.
I wandered away from cyberspace for quite a while, so I have a lot of catch up to do.
Have Fun!
Reg
-
Yeah, I heat with wood and cool with... nothing, so the only good times of the year for me are spring and fall. I can keep the lab at a very consistent 20C for a few weeks. The lab dehumidifier died, so I need to buy a new one.
-
Yeah, I heat with wood and cool with... nothing, so the only good times of the year for me are spring and fall. I can keep the lab at a very consistent 20C for a few weeks. The lab dehumidifier died, so I need to buy a new one.
Sounds like New England. There are several hundred milliseconds a year where the temperature is 'just right'! :D
-
Here's a ~ 4 day measurement on my DMM-196, looking at the PX-REF#1 powered by a random 12V wall-wart. I realize that's sort of casual, so I just switched it to +12.00 V from a Power Designs 2005A which is a little bit more stable as a supply.
So far, my DMM-196 (last cal on 6-16-2008) reads 7.04545 V, sigma = 20.5 uV, on 35k samples.
PX-REF#1 box label reads 7.0456 V
According to the provided Arduino/SHT21 sensor, the temperature in the room is 17.5 C +/- 1 C and the relative humidity RH is 57 % +/- 2 % over the time period shown.
The temp/RH sensor reports readings at about 1.5 second intervals. To make it easier to line up the measurements I standardized on reading at the top of every 10 second interval (or the closest reading just after that point).
Here is a python program that does this with the provided Arduino board T/RH sensor:
https://github.com/jbeale1/DataAcq/blob/master/SerialLog-Even10s.py
-
Very nice. Great to see another 196 in action. :-+
-
I first found EEVblog forum when I was trying to fix a Keithley 2001 DMM. And now, 600 posts later, this is the meter collection:(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1212572;image)
Most of these meters have never been adjusted after manufacturing. The only one with current calibration is DMM6500. To calculate the deviations I used resistor values as reported by Arthur Dent and verified by CalMachine and voltage reported by TiN and also confirmed by CalMachine. Then I used 10 kOhm resistor (9.99979 kOhm) and 10V FX reference (9.999966V) to calibrate my HP3458. In general, it seems all meters drift by comparable amounts.
Another interesting aspect is how much does the calibration change after AutoCal on recent DMMs
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1613443).
In particular on 34465 and 34470 meters, it can change a lot. Well, the last time the AutoCal was run on those meters was 2 years ago. Still ~100 ppm is a big change considering that the temperature has not changed by much and the meters have already been somewhat aged. So, it seems Keysight uses some low-quality components and relying on AutoCal to cancel the drift. This may explain funny drift of 34470 after Autocal. It seems they should force the user to run AutoCal at regular intervals, otherwise the meter can drift a lot.
I also looked at the voltage stability of the FX reference measured by several meters.
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1212576;image) Here we see the famous noise bump of Keithley DMM7510 that is prominent not only at 0V but also at 10V.
I also looked at the noise spectral density of the voltage references. For PX reference I compared it with an HP LTZ1000 voltage reference board 03458-66509 that I had. The difference between their voltages was a little more than 100 mV, but I could monitor it using HP3457 on 300 mV range. Another way to measure the noise spectrum is with a very low cut-off (0.03 Hz) high pass filter (100 kOhm and 50 uF) and a low noise amplifier (LNA10 from Alpha lab). (https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1212578;image) The two measurements agree well and confirm the noise density of <200 nV/Hz^{1/2} at 0.1 Hz. It should be noted that I divided HP3457 spectrum by a factor of 2, one factor of sqrt(2) comes from taking difference of two LTZ references and another factor of sqrt(2) comes from the fact that DMM with autozero only measures the signal approximately 50% of the time.
I also looked at the noise spectrum of FX reference and found that it is a little higher than one might expect naively from voltage scaling and also has a pronounced oscillation at around 3 kHz.(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1212580;image) One can see it in the time domain on the oscilloscope with an amplitude of about 20 uV using a gain 1000 of LNA10 with 100 kHz bandwidth.(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1212582;image)
Probably putting a capacitor somewhere in the op-amp circuit will get rid of the oscillations.
So this is my report, I am ready to send on the kit unless someone comes up with a quick test idea. It has been very helpful to get access to this calibration gear, thank you everyone for contributing to it and organizing the exchange.
-
I also looked at the noise spectrum of FX reference and found that it is a little higher than one might expect naively from voltage scaling and also has a pronounced oscillation at around 3 kHz.
Probably putting a capacitor somewhere in the op-amp circuit will get rid of the oscillations.
So this is my report, I am ready to send on the kit unless someone comes up with a quick test idea.
Mhm,
did you check the 3 kHz oscillation also for the positive and the 0V output independantly?
The capacitors are already there. (C39 + C40)
https://xdevs.com/article/792x/
I guess that C40 together with R32+R33 is somewhat low against the product (time constant) of R31+C48.
Whereas C39 together with R1-R3 should be about right. (ok somewhat at the limit).
So it should be the 0V output which is more likely to oscillate.
with best regards
Andreas
-
In particular on 34465 and 34470 meters, it can change a lot. Well, the last time the AutoCal was run on those meters was 2 years ago. Still ~100 ppm is a big change considering that the temperature has not changed by much and the meters have already been somewhat aged. So, it seems Keysight uses some low-quality components and relying on AutoCal to cancel the drift. This may explain funny drift of 34470 after Autocal. It seems they should force the user to run AutoCal at regular intervals, otherwise the meter can drift a lot.
Nice to have another confirmation on this.
In my experience it is only the 34470A that drifts and NOT the 34465A.
I had two of each and the two 34465A were always spot on after an ACAL, exactly as expected by the book.
The two 34470A started drifting after a few months and it only got worse over time, even if you would use ACAL every day. I still have one 34470A but sold the second one, because it drifted far too much.
I started a thread somewhere here on the forum.
-
I first found EEVblog forum when I was trying to fix a Keithley 2001 DMM. And now, 600 posts later, this is the meter collection:(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1212572;image)
Most of these meters have never been adjusted after manufacturing. The only one with current calibration is DMM6500. To calculate the deviations I used resistor values as reported by Arthur Dent and verified by CalMachine and voltage reported by TiN and also confirmed by CalMachine. Then I used 10 kOhm resistor (9.99979 kOhm) and 10V FX reference (9.999966V) to calibrate my HP3458. In general, it seems all meters drift by comparable amounts.
Thanks for sharing. In general (above only prompted to write) I'd like to see error bars on measurements, i.e. estimates of the consistency of the readings. One way to do that to measure a given artifact with a giving meter multiple times (also reversing polarity to minimize effects of thermal EMF at the artifact<->probe junction) and report the standard deviation.
We can look up the specifications of those meters to get a hint of how close those measurements could be, but without results from procedures like the one described above, we have no way of knowing how well those specific meters actually performed then.
-
The two measurements agree well and confirm the noise density of <200 nV/Hz^{1/2} at 0.1 Hz. It should be noted that I divided HP3457 spectrum by a factor of 2, one factor of sqrt(2) comes from taking difference of two LTZ references and another factor of sqrt(2) comes from the fact that DMM with autozero only measures the signal approximately 50% of the time.
I also looked at the noise spectrum of FX reference and found that it is a little higher than one might expect naively from voltage scaling and also has a pronounced oscillation at around 3 kHz.
The extra factor sqrt(2) for the DMM only sampling the input half the time only applies to white noise, but not the higher 1/f noise part. So the comparison is not that easy.
The broad peak at around 3 kHz does not look like oscillation - it is much to wide for this. It may be just noise amplification from not so ideal compensation, close to instability, but still well on the stable side. The amplitude is also quite small. The frequency is a bit low to be a chopper amplifier residual.
-
The two 34470A started drifting after a few months and it only got worse over time, even if you would use ACAL every day. I still have one 34470A but sold the second one, because it drifted far too much.
I started a thread somewhere here on the forum.
I am wondering if doing ACAL frequently is what is contributing to long-term drift. After ACAL there is clearly additional drift, likely from some change in heating of internal components. It maybe that during ACAL some resistor is assumed to be the same for several measurements, but in fact drifts a little and always in the same direction, accumulating the errors.
-
The extra factor sqrt(2) for the DMM only sampling the input half the time only applies to white noise, but not the higher 1/f noise part. So the comparison is not that easy.
Yes, good point, I will have to think more about how to do it properly.
The broad peak at around 3 kHz does not look like oscillation - it is much to wide for this. It may be just noise amplification from not so ideal compensation, close to instability, but still well on the stable side. The amplitude is also quite small. The frequency is a bit low to be a chopper amplifier residual.
Yes, its just noise gain peaking, slightly larger compensation capacitor should be able to kill it.
-
The kit is on the way back to vindoline, but here is a comparison of voltage drift from several LTZ references (of course measured with another LTZ reference inside the HP3458).
(https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/?action=dlattach;attach=1213483)
-
Can we get a teardown of the SRS DC205 pretty please?
-
Can we get a teardown of the SRS DC205 pretty please?
Yea, I was going to do a teardown, dig into the circuit and see if it can even be improved, but there is a long list of other projects ahead of it now. Here is the picture of the circuit. The op-amps that have guarded traces going to them are OPA192.
-
I'm kind of looking forward to round3, As I kind of used round2 as a calibration round to self calibrate my 3478A, I'd be curious as to see if there was any drift with my DMM since last I had the unit. Verifying if my ebay purchase was indeed a success or not.
-
Hoping that there will be a 3rd round.
Your post gave me an idea. Perhaps there could also be another type of club ... one with a quick turn-around time for a fast check on our instruments. It could contain two standards: an FX voltage reference and a 10 k resistor.
-
Plus a standard capacitor, maybe a bank of NP0 types, a dissipation factor standard and an inductance standard! Shouldn't take up much extra room.
-
I'm very much with Conrad on adding some other reference standards.
I'd also like to point out that Doug Malone's DMMCheck Plus from voltagestandards.com is an excellent piece of affordable kit. I've been using mine to check an ebay 3478A that came a few days ago.
Have Fun!
Reg
-
I'm very much with Conrad on adding some other reference standards.
I'd also like to point out that Doug Malone's DMMCheck Plus from voltagestandards.com is an excellent piece of affordable kit. I've been using mine to check an ebay 3478A that came a few days ago.
Have Fun!
Reg
+1 for voltagestandard.com
They also have a 0.002% 10V reference which works very well:
https://voltagestandard.com/002%25-10v-reference (https://voltagestandard.com/002%25-10v-reference)
-
[..]
I'd also like to point out that Doug Malone's DMMCheck Plus from voltagestandards.com is an excellent piece of affordable kit. [..]
It does what it promises (verifying multimeters to moderate accuracy) and for its intended purpose it's a good value, but it isn't ideally suited for application in metrology. I wished he'd have used terminals with low thermal EMF. The resistors used don't instill confidence in long term stability either.
-
I also have the 10 V reference.
Neither is metrology grade, but they *are* affordable references for most practical purposes.
A budget metrology grade version of the DMMCheck Plus would run well over $1000 and probably cost 10-15% of that per year for cals. Those last few decimal places get expensive quickly.
Have Fun!
Reg
-
Having a lower quality reference is one approach to getting a faster cycle time, which in my opinion would be the largest improvement that our club could have. None of us who don't pay for formal calibrations has had two measurements with the same reference since the club started. So we can only guess about stability of our equipment. As far as I know TiN is the only one to have multiple observations on the same reference so our window into the stability of the reference is limited. Fortunately due to the quality of his measurements we can have fairly high confidence in that small window.
While a couple of the stops in the current and previous cycle have tried to look at longer interval (days or weeks) stability, it doesn't appear that lengthy observation periods are the main reason for the long cycle times.
Perhaps the next cycle through will be quicker just because of experience, but anything we can think of to help the process is worth considering.
-
I also think shortening the cycle time would be very helpful.
I suggested having a 2 tier set of kits for voltage and an additional kit for RF for round 3, but no one seemed to be interested in the RF and not much enthusiasm for a 2 tier voltage set.
I bought an HP RF power meter calibrator specifically for round 3 and have some capacitors and inductors which should be pretty good if properly packaged. But I seem to be the only RF oriented person.
For lower tier DMM cals, I think a DMMCheck Plus and/or 10 V reference that circulated quickly with stops at Doug Malone's shop every 12 months would be useful for many people who are not volt-nuts, but simply want a cheap annual cal verification.
That said, both of my 34401As were well within spec despite age and heavy use and I'm not trying to wring the last bits of accuracy out of them.
Have Fun!
Reg
-
Perhaps the next cycle through will be quicker just because of experience, but anything we can think of to help the process is worth considering.
Having the familiarity of at least one round under our belt should help speed up the process, although one could argue that the long cycle time contributes to that familiarity fading by the next round.
Refreshing one's recollection of the kit, prepping one's testing setup, and having a defined process to execute (and/or automation) prior to the kit's arrival can help with turnaround as well.
I'm behind on updating the docs for the kit, so I need to address that as well.
Apart from having more than one kit, any other ideas for improving our efficiency?
-
Perhaps someone with web-programming skills could write a simple page, that would show queue of people in loop, and where the kit currently is?
Or even step further - add data fields so members can upload their measurement results and data, so there is one nice combined page with all measurement results and time slots. This can be used to finally plot members results equivalence plots and deviation between each lab. Yea, I know, dreams... :=\
-
Perhaps the next cycle through will be quicker just because of experience, but anything we can think of to help the process is worth considering.
Having the familiarity of at least one round under our belt should help speed up the process, although one could argue that the long cycle time contributes to that familiarity fading by the next round.
Refreshing one's recollection of the kit, prepping one's testing setup, and having a defined process to execute (and/or automation) prior to the kit's arrival can help with turnaround as well.
I'm behind on updating the docs for the kit, so I need to address that as well.
Apart from having more than one kit, any other ideas for improving our efficiency?
I think you have identified the most important speed up points.
Prepping your equipment and having a plan of action. There is always a few weeks warning of when it will get to you. Setting up and doing dry runs on local sources (can be as simple as a battery) saves lots of dwell time. In both arrivals at my place the prep work uncovered surprises in my instrumentation that needed significant time to rectify. Even with this prep work there were last minute adjustments, fortunately taking only hours, not days. Things like lead lengths and places to put the DUTs are the kind of simple things that can take a fair amount of time to deal with. I am sure others have found surprises too.
-
The best laid schemes o' mice an' men...
Yes, both times I thought I was well prepared and all manner of issues still came up, even having to rebuild failed equipment.
-
Another way to organize it in the next round might be to do away with the list and instead have each person put up a post when they finish with the kit: "Kit available, who wants it next? First to post proof wins." Proof would constitute a fresh dry run on an alkaline AA or something.
If competition gets too heated, we could chill it with a "dibs" system -- accept the top two entrants from every other post or the top three from every third, etc. If we turn this knob all the way to N we transform the competition system back into the reservation list system.
Just a thought :)
-
I like the "I'm ready" concept. Sometimes you may not even want it when it's officially your turn. I think that's happened a time or two.
-
I think there is merit in the I'm ready idea, but needs some more thought on implementation. Not that what has been proposed so far wouldn't work, but maybe we can come up with something even better.
Here is one possibility. Use the ordered list but give the next half dozen people on the list a day (or some other fixed, short time) to respond to the I'm ready message with proof. Kit goes next to the person closest to the front of the list to respond. List order would be unchanged other than moving the selected party to the next round list.
If the kit moves as fast as I would hope it does this could be modified by allowing the I'm ready proof to be valid for a month or two. This would avoid people having to generate repeated proof runs. Other modifications might apply to those with better equipment (calibrated, 8 digit, traceable standards or whatever we feel would be appropriate).
-
Good ideas. Wouldn't want the process to become cumbersome.
Well, the best way to find out if it works is to give it shot and refine any rough spots.
-
I would like to endorse the "I'm ready." concept. A $3-5 ebay reference would be sufficient to let someone get things set up. I was quite stressed trying to do a reasonable test run and get it back out the door. I'd never done this before and didn't grasp how much work it entailed.
Even though I have done it once already, I'd set up and do a trial run before getting the package the next time.
Have Fun!
Reg
-
Perhaps it might be worth posting some guidance on how to make best use of the shipped gear (somewhere with an easily recognizable URL, i.e. not buried here among dozens of other posts).
But then, even the pros mess up, despite clear instructions: https://www.bipm.org/documents/20126/47750471/EURAMET.EM-S41.pdf/39f8462f-4bb8-09ef-67d1-661535764a36 (https://www.bipm.org/documents/20126/47750471/EURAMET.EM-S41.pdf/39f8462f-4bb8-09ef-67d1-661535764a36) :o
-
Perhaps it might be worth posting some guidance on how to make best use of the shipped gear (somewhere with an easily recognizable URL, i.e. not buried here among dozens of other posts).
But then, even the pros mess up, despite clear instructions: https://www.bipm.org/documents/20126/47750471/EURAMET.EM-S41.pdf/39f8462f-4bb8-09ef-67d1-661535764a36 (https://www.bipm.org/documents/20126/47750471/EURAMET.EM-S41.pdf/39f8462f-4bb8-09ef-67d1-661535764a36) :o
I agree with having some instructions and with TiN's suggestions for a web page of some sort, although I'm unable to meaningfully help in that area. Perhaps just something in Google Docs? I would like to participate, but I'm not sure I'm worthy nor how much I could contribute. Perhaps I could build a shipping case?
I read that document and I'm still thoroughly confused. Unless I missed something, they indicate that it was just a math error, but I don't see how that could cause the one device reading to be so far off.
-
Thanks for the interesting paper. Indonesia has a Josephson array voltage standard, but they turn it on only every three years or so. I was just wondering whether there is a Josephson Array in Kabul.
Regards, Dieter
-
Hello Cal Club members, at this point I believe that everyone on the list has had the opportunity to use the club's cal kit. If anyone has been missed, please send me a PM. Also, there have been several members who asked to be “put on hold”, or “moved to the end of the list.” If any of you are ready, please send me a PM
As it is, mostly, the USA Cal Club, Happy Thanksgiving everyone!
Vindoline
-
Thanks for the update, Vindoline. Happy Thanksgiving, everyone!
-
I've started converting the old USA Cal Club: Getting Started & User Guide that used to be in a forum post into a PDF file. This will make it easier and nicer to print out as some have requested.
This is a work in progress. Thus far the existing content has been migrated and laid out so it looks presentable for printing. Next phase is to add the new information that TiN has provided during Round 2.
Since the file is too large to attach to a forum post, you can view and download it at
https://drive.google.com/file/d/1AoXDJ-ybEtL40Hh-SGyr5oJMXcGVgf-l/view?usp=sharing (https://drive.google.com/file/d/1AoXDJ-ybEtL40Hh-SGyr5oJMXcGVgf-l/view?usp=sharing)
-
I've started converting the old USA Cal Club: Getting Started & User Guide that used to be in a forum post into a PDF file. This will make it easier and nicer to print out as some have requested.
This is a work in progress. Thus far the existing content has been migrated and laid out so it looks presentable for printing. Next phase is to add the new information that TiN has provided during Round 2.
Since the file is too large to attach to a forum post, you can view and download it at
https://drive.google.com/file/d/1AoXDJ-ybEtL40Hh-SGyr5oJMXcGVgf-l/view?usp=sharing (https://drive.google.com/file/d/1AoXDJ-ybEtL40Hh-SGyr5oJMXcGVgf-l/view?usp=sharing)
That sounds great Bitseeker! Let me know when you're done, and I'll put it on a USB drive and include it with the kit.
-
You can include the current version, if someone is already ready to use the kit. Since it's a PDF now, it's easy to overwrite with new versions.
-
Hi, Newbie CalClub reader. I have not read the 30 or so pages, just the first and last.
Can someone summarize where this project is at this point, a few questions off the top of my head:
0) What is the objective?
1) Do you need more participants?
2) I don't have a 7.5 or 8.5 DMM, just a 3x 34401 and 2 x 8846A none calibrated recently, however they track each other quite well, is there a minimum qualification requirements for DMM?
3) How long do we capture data before sending to next participant.
4) I might want to build an equivalent board for use here, are the schematic and PCB files readily available?
5) Is there a published format for the data; time interval, temperature etc?
Thanks,
J
-
Welcome, Jester. Since you've read the first and last pages, you have some basic info, where to get and put data files, can check out the log files that have been submitted thus far, and a link to the latest manual about the club's calibration equipment, getting started, how to hook everything up, etc.
Reading this thread at least from about page 6 or 7 will get you to the point that folks started talking about their experience with the kit. It's recommended to read from there forward. There are tangential posts, so you can effectively skim over stuff that's not relevant (i.e., you won't actually read every post in the thread).
Between the manual and this thread, you'll find most of what you seek. Info on the reference devices might be either in this thread or the thread for Round 1. I don't recall.
Your DMMs will be fine and it's good that yours all agree. We don't all have uber-resolution equipment. If you can log for a few days at least, that'd be good. If you're able to go longer, that can be accommodated.
Contact vindoline after you've gone through the manual and the thread, if you'd like to join the rotation. Note that the list is long, so it can take quite some time for the kit to get to everyone. But, don't dismay. It eventually gets to everyone.
-
I will add just a bit to bitseekers comments. Many of us in the club have older equipment, not recently calibrated and are looking for some indication of where our equipment stands without sending in for an expensive calibration. Your equipment list sounds like it will fit in quite well with others in the club.
For those with the knowledge and interest the ensemble of measurements can provide even more information about performance. But none of the exotic analysis is required.
The primary reference sent around is a standard design and it is possible to build your own copy. It is not inexpensive or easy to fully calibrate the result. TiN has graciously done a lot of work to build and characterize this reference, and access to it is a major benefit of the club. You may wish to look into any number of lower cost and lower performance references for your local use. When you are not operating in the 8-10 digit precision world you can get by very well with lesser devices.
As you will pick up as you read the thread, patience will be required. It has taken a year or more for a circuit around the club membership and while there is reason to hope for somewhat faster cycling, growing club membership will like counter any gains in that area. Fortunately most of our older instruments have stabilized enough that the long interval isn't critical.
-
I just speed read through the 30 or so pages of round 2, so I now have a feel for progress so far.
First off special thanks to: cellularmitosis, TiN and vindoline for their extraordinary efforts. Also to everyone else that contributed along the way.
I would like to participate to either the end of round 2 or get in line for round 3. I will PM vindoline.
What I plan to do:
1) Prepare while waiting:
A) get a GPIB logging system up an running well in advance * complete
B) log data using my Data Precision calibrator, even though it appears to drift more than all of my best 5 DMM’s in progress
C) previous step will also give me some idea of temperature swings
D) fabricate a four sided metal box to allow me to easily slip out Data Precision** (presently on bottom of heap) complete
Objectives:
a) begin logging drift history of my DMM’s relative to each other and occasionally with the cal club references.
b) calibrate at minimum the 10V range of my 5 DMM’s
c) explore the possibility of upgrading my calibrator ** with improved: reference, resistors and op amps, I will start a new thread asking for advice.
*Question:
Regarding data logging, I’m left with the impression that different participants used different data logging programs and windows created a fair bit of problems related to GPIB drivers and updates. I have had multiple GPIB instruments successfully communicating in the past with a clone HP to USB adapter however I also experienced numerous problems in the past as well, in fact the last time I tried to use some excel macros that were working fine they no longer worked. I have wasted too much time on this in the past and prefer to learn from that and move on. To that end I’m inclined to setup a Raspberry Pie as a dedicated reliable data logger. I have never used a RP so if this idea holds merit if someone can suggest what model I should order, suitable operating system (or do they already come with an OS?) and data logger software that would allow logging 2 or 3 DMM’s at the same time for the sake of time efficiency
Thanks,
J
EDIT--> UPDATE
My Data Precision 8200 calibrator has about 10 pieces of equipment on top of it, making removal a real PITA, so I decided to weld up a 12 gauge steel box that houses the calibrator making it easy to slip out without moving everything above. Mission accomplished....
I was able to get three instances of the Nx-1997 HP34401A Standalone software running and logging 3 DMM's on a Win 10 tablet, when I tried to log a 4th channel for temperature Win10 had a conniption so I put together a quick and dirty temperature logger that logs to a microSD card. I will let it run for a day or two and see how the data looks.
-
You have way too many meters! We have the same situation- my 8200 is also under pretty much everything else and is very difficult to get out for calibration. BTW, IMHO the 8200 is a really excellent device. Even though it only uses a lowly 399 reference, mine stays within a couple ppm for years at a time. The only negative is mine gets dirty intermittent switches and I have to clean them every now and then. Even with that they're never perfect for long. It would be great to have the GPIB board, but I don't.
-
You have way too many meters! We have the same situation- my 8200 is also under pretty much everything else and is very difficult to get out for calibration. BTW, IMHO the 8200 is a really excellent device. Even though it only uses a lowly 399 reference, mine stays within a couple ppm for years at a time. The only negative is mine gets dirty intermittent switches and I have to clean them every now and then. Even with that they're never perfect for long. It would be great to have the GPIB board, but I don't.
My 8200 seems to drift very slowly (I think). with my 3x 34401A's connected in parallel the difference between the three meters is only about 30ppm, however a few years ago after a fresh calibration of the 8200 the difference between the 8200 an all three meters was quite small, now it is significant, I need to set the 8200 to 10.00129 to get the meters displaying 10.00000V
It's possible that all three meters have drifted, but that seems highly unlikely. So my assumption at this point is that the 8200 is drifting down. I'm getting ready to start logging the 8200's reference relative to the 34401A. I would like to eventually get the 8200 to drift less than the 34401A's possibly by upgrading the: reference, resistors and opamps. Looks like this will take quite some time. I managed to get data logging of my 3x 34401A's going yesterday, so a small step in the right direction.
When you have a dirty switch, is it obvious?
Cycling the switches on mine seems to be okay, nothing I have noticed. I have noticed that on rare occasion mine gets in a snit and outputs a value that is off by a bit, however cycling the +0- switch seems to reset the problem, seems fine after that.
-
Dirty switch can be the occasional "off by a bit" and does no harm. Mine is far worse and just cycling the switches doesn't fix it. Once I had a resistor go bad and you want to be very careful of what you use for replacements in the whole reference/amplifier/summer system, especially at the top end. I think I used a Vishay metal foil part. If you have a lot of drift you might want to inspect for previous repairs and cleanliness.
-
You have way too many meters!
How dare you say such a thing! Oh wait this isn't TEA.
-
Been quiet here for a while. What is the status? Any ideas about a 3rd round?
-
Has been quiet here, Lately I've been thinking about setting up my 3478a with it's own raspberry pi for data logging, however I'm not really super verbose with the need to know information. My idea is sound and theoretically it should work without issue. My issue is knowing the actual coding that needs to be done to get it to work. Currently I have a 3478a connected up with an arduino that connects to the GPIB port and I'm capable of talking to it and reading from it through usb no problem. So the idea is to connect that USB into a raspberry pi that'll run a script for grabbing the data from the meter. Then send that data over to a graphana server that I have running for temperature logging for the server, in which I just followed a simple guide I don't actually know how to use Graphana. Allowing me to log anything I'd wish to log. With the idea of logging temperature humidity and voltage/amperage over a long period of time. I'll likely post this elsewhere trying to find help with it, as the hardware side of things I can do no problem, it's the software that I'm clueless with, but I don't really know where to post it.
So hopefully if we do get around to a round 3 I can keep a better log over time.
-
I remember there was a thread about Pi logging, but I haven't kept up on the latest.
Ah, yes. Here it is: https://www.eevblog.com/forum/metrology/raspberry-pi23-logging-platform-for-voltnuts/ (https://www.eevblog.com/forum/metrology/raspberry-pi23-logging-platform-for-voltnuts/)
There were some issues regarding Linux drivers for GPIB, but I think that was resolved quite some time ago. Anyway, have a look if there's anything useful there.
A common setup of open source tools might be good to define for general use.
-
Will there be a round 3?
Any hint about when it will start?
-
Is round 2 over?
Will there be a round 3?
-
I recently talked to Vindoline.
He is currently too busy to take on another round (so that must mean that the previous round is over).
I agreed to be the next "moderator of rounds" if no one better able and qualified is up for it.
So please PM me if you would like to be the "king of rounds" or "A-number-one-calibrator" or ?
Thanks,
Randall McRee
Davis, CA
-
I recently talked to Vindoline.
He is currently too busy to take on another round (so that must mean that the previous round is over).
I agreed to be the next "moderator of rounds" if no one better able and qualified is up for it.
So please PM me if you would like to be the "king of rounds" or "A-number-one-calibrator" or ?
Thanks,
Randall McRee
Davis, CA
TiN has graciously re-certified the club FX voltage reference and donated a characterized 10k ohm reference resistor:
https://xdevs.com/article/usac_2023/
https://xdevs.com/review/ln4040b/
-
I find those L&N resistors perfectly decent so long as you know the temperature. Mine tend to be a bit leaky, so probably should be double bagged or used dry at low currents. Definitely interested in a 3rd go at some point.
-
I find those L&N resistors perfectly decent so long as you know the temperature. Mine tend to be a bit leaky, so probably should be double bagged or used dry at low currents. Definitely interested in a 3rd go at some point.
I also have some leaky ones, but I'd prefer not to go dry, because the oil at least adds to the heat capacity and slows down any temperature changes. So removing the oil would probably exacerbate the relatively high tempco and make the coupling of the thermometer to the resistive element much worse.
Removing the oil would cut down on shipping weight if that helps, though.
-
Is this going forward? I have some spare time and can help with organization and shipping if others are busy.
-
It is going forward. Randall McRee has agreed to be in charge.
-
Going through the thread and getting up to speed, but if still possible, I'd like to take a turn with this.
-
You have way too many meters! We have the same situation- my 8200 is also under pretty much everything else and is very difficult to get out for calibration. BTW, IMHO the 8200 is a really excellent device. Even though it only uses a lowly 399 reference, mine stays within a couple ppm for years at a time. The only negative is mine gets dirty intermittent switches and I have to clean them every now and then. Even with that they're never perfect for long. It would be great to have the GPIB board, but I don't.
My 8200 seems to drift very slowly (I think). with my 3x 34401A's connected in parallel the difference between the three meters is only about 30ppm, however a few years ago after a fresh calibration of the 8200 the difference between the 8200 an all three meters was quite small, now it is significant, I need to set the 8200 to 10.00129 to get the meters displaying 10.00000V
It's possible that all three meters have drifted, but that seems highly unlikely. So my assumption at this point is that the 8200 is drifting down. I'm getting ready to start logging the 8200's reference relative to the 34401A. I would like to eventually get the 8200 to drift less than the 34401A's possibly by upgrading the: reference, resistors and opamps. Looks like this will take quite some time. I managed to get data logging of my 3x 34401A's going yesterday, so a small step in the right direction.
When you have a dirty switch, is it obvious?
Cycling the switches on mine seems to be okay, nothing I have noticed. I have noticed that on rare occasion mine gets in a snit and outputs a value that is off by a bit, however cycling the +0- switch seems to reset the problem, seems fine after that.
Jester - check my thread on fixing the switches on the 8200 ("Data Precision 8200 calibrator - repair and other fun stuff"). It's not hard and it fixes them 100% - my conclusion is this happens due to the hardening and degradation of the grease used in the switch - though I wish I'd have put some fresh grease myself when reassembling them. The only conditioning I did inside is deoxit on all contacts, which hopefully lubes the moving parts a bit.
But well worth the effort.
The drift you're seeing seems very high to me, maybe the unit needs actual repair.
Apologies to everyone for posting this here, but hopefully we're merging it into the thread where it belongs.
-
A heads up for those who don't lurk the forum regularly. There is a new thread for Round 3.
https://www.eevblog.com/forum/metrology/usa-cal-club-round-3/ (https://www.eevblog.com/forum/metrology/usa-cal-club-round-3/)