-
#100 Reply
Posted by
t2kv
on 15 Nov, 2016 06:33
-
Is there something or is this just chit chat?
Some of the Australians have started swapping references and sharing measurements. Not sure if there's any other activity.
-
#101 Reply
Posted by
enut11
on 15 Nov, 2016 06:41
-
Can someone give a summary of what is going on, ie a status report? Is there something or is this just chit chat?
thanks
From what I can see the post has stalled somewhat. Whilst the idea of a calibration club has appealed to many, the practicalities have put a damper on it.
I feel that most of the voltnut expertise is in Europe (just an observation from the postings) and getting a design specified, approved, built and shipped (in particular internationally) has yet to inspire this group. Also, as already suggested, there is an ongoing local European club already doing something like this at local meetings.
There have been offers from eevblog forum members to circulate current worthy standards such as Geller, etc and but nothing has come of it.
I still feel that country/region based clubs is the way to go. However, until someone takes the plunge nothing will happen.
Meanwhile, in Australia, a number of junior voltnuts have banded together to initially share their mostly home-made standards. At least they will get the benefit of a 'second opinion' and gain more confidence in the accuracy of their meters. Currently all members have shipped their standards to one member who will measure all under the same conditions. All the standards are then posted to the next member, and so on. Hopefully it will grow from there to more serious standards sharing as experience is gained.
This is a call for action. If you think it is a good idea and want to be involved, tell us where you live and what you have to offer.
Then, you too may be able to form an Aussie-type club in your area or country.
The ultimate calibration club would be an international one but that will probably have to wait until another day.
enut11
-
#102 Reply
Posted by
TiN
on 15 Nov, 2016 11:36
-
Show stopper for me on this thread is lack of hard fixed specification of what are exact rules and requirements are.
Without this everyone would just do random things to random boxes, and there would be weak if any correlation between such "transfers".
Don't get me wrong, idea is good, but it's implementation is what makes it useful.
I did international transfer on my DCV (<2ppm) and Ohms (<8ppm), so the whole concept is real, I assure you.
I mean, just look at 2DWxxx low-noise zener thread. There are just 40 folks trying to get few pieces of simple zener, but there are already communication issues, someone get held up, others busy, and so forth.
Ideas like this must be run by dictatorship, with clear specs and fixed features. Otherwise everyone will just pull things sideways and apart.
Project leader has to define the direction and show the way to others. Then interested parties will jump in, if they willing to comply and commit.
-
#103 Reply
Posted by
retrolefty
on 15 Nov, 2016 13:34
-
Show stopper for me on this thread is lack of hard fixed specification of what are exact rules and requirements are.
Without this everyone would just do random things to random boxes, and there would be weak if any correlation between such "transfers".
Don't get me wrong, idea is good, but it's implementation is what makes it useful.
I did international transfer on my DCV (<2ppm) and Ohms (<8ppm), so the whole concept is real, I assure you.
I mean, just look at 2DWxxx low-noise zener thread. There are just 40 folks trying to get few pieces of simple zener, but there are already communication issues, someone get held up, others busy, and so forth.
Ideas like this must be run by dictatorship, with clear specs and fixed features. Otherwise everyone will just pull things sideways and apart.
Project leader has to define the direction and show the way to others. Then interested parties will jump in, if they willing to comply and commit.
I agree, surely there is a Alfa Volt-Nut to lead the others.
-
#104 Reply
Posted by
ez24
on 15 Nov, 2016 21:43
-
Ideas like this must be run by dictatorship, with clear specs and fixed features. Otherwise everyone will just pull things sideways and apart.
Project leader has to define the direction and show the way to others. Then interested parties will jump in, if they willing to comply and commit.
Totally agree this would be the only way - my vote goes to TiN since he has the calibrators
(just curious - did the name TiN come from the metal or a city ?)
-
#105 Reply
Posted by
t2kv
on 15 Nov, 2016 21:48
-
Ideas like this must be run by dictatorship, with clear specs and fixed features. Otherwise everyone will just pull things sideways and apart.
Project leader has to define the direction and show the way to others. Then interested parties will jump in, if they willing to comply and commit.
Agreed.
That said, this is what's just happened in the LTZ1000 thread, with TiN and a couple of others offering to calibrate others' boards against their kit. That sounds like the pragmatic start of a calibration club to me, particularly if those references then rotate within regions after cal.
-
#106 Reply
Posted by
enut11
on 16 Nov, 2016 04:11
-
Ideas like this must be run by dictatorship, with clear specs and fixed features. Otherwise everyone will just pull things sideways and apart.
Project leader has to define the direction and show the way to others. Then interested parties will jump in, if they willing to comply and commit.
Agreed. Must be a benevolent dictator
. He also needs to be a voltnut expert to create the proper specs suitable for the ongoing success of this Club.
-
#107 Reply
Posted by
TiN
on 16 Nov, 2016 04:23
-
my vote goes to TiN since he has the calibrators
Don't vote for me, I have enough backlogs to add one more fullblown endless project. And calibrator is not relevant in this project case
(just curious - did the name TiN come from the metal or a city ?)
It's simple, first letters of my passport initials
-
#108 Reply
Posted by
Theboel
on 16 Nov, 2016 06:06
-
I vote for TiN but I am very patient until.
1. TiN someday visit to Yogyakarta or Borobudur temple with his LTZ boxes
2. My LTZ1000 and LM399 old enough and have calibrated against Indonesia NMI standard.
Until then I will waiting for any standard TiN has in his mind.
-
#109 Reply
Posted by
Nuno_pt
on 16 Nov, 2016 09:45
-
I would also vote for TiN, but is backlog is big so it's very difficult for him to lead a project like this.
I think that the best we can do is agree on specs for a Traveler Standard Box, like TiN KX LTZ1k board and 10k resistor with BME280 for logging temperature.
Since the calibrator will be very difficult to ship back and forward, a small TSB would be good.
I'll build one of TiN LTZ1k board for the TSB, I'm awaiting for Edwin set of resistors all with 0.1% 3ppm/C, and also a 10k 0.1% 3ppm/C for OHM transfer, the BME280 are on the way, the box will be a Hammond not shure what model yet.
-
#110 Reply
Posted by
VK5RC
on 17 Nov, 2016 10:37
-
Re travelling voltage standards I did a bit of an experiment today to see the effect of "careless" temperature management.
It started off cool here about 15C ambient/otdoor temp (0700) then heated uo to 38C max at about 1500hrs. My 'lab' started off at 21C ended up at 28C before I switched on air conditioner at 1700. The voltage reference was quite a bit hotter at about 34C.
I was also a lot more careless about door openings. The effect is dramatic. Moving references around will probably have a similar or even worse effect . I don't think the cargo hold is pressurised and can get really cold I believe.
One of the time-nuts actually dug a deep pit to get temperature stability for his frequency reference! Something to think about.
Below Blue trace is voltage on RHS, Orange trace is internal temp of 3458A, should be 37C (last cal temp)
-
#111 Reply
Posted by
TiN
on 17 Nov, 2016 18:06
-
Sorry, it doesn't look like your graph is showing anything, really, with due respect. First, if ambient temperature went up 7C, from 21 to 28 before AC on, it should have very similar effect on TEMP? output of the meter too. But that's not as issue, as reference output. It's span 7ppm, which is not what LTZ-based module should do, even on much largest temp span. So the setup is compromised and any correlation of temperature to output is just buried down down under the interference to LTZ module (be it EMI/RFI or air drafts or mechanical stress to cables or module).
Suggestions to improve:
* I don't know what logging sw you use, but add ACAL DCV function on each 0.5C of TEMP? change. This will remove tempco from your 3458A, down to <0.05ppm/K. My 3458A has tempco on 10V DCV range about 0.35ppm/K without ACAL, to give you the idea.
* Use twisted pair short copper wire between your DUT LTZ and 3458A. Put foam around binding posts to isolate from air drafts. You can use regular LAN UTP cable, it works great for this case. No need shmick 100$/meter teflon silver-99%-of-light phoolery here, just pure fresh cut copper will do great.
* Make sure your PSU is quiet. You can test it with 3458A , just log bit of data to see pk-pk. SMU should be fine, I use my 2400 all the time for LTZ tests, but I checked it before comparing with battery powered ref, to ensure setup is good.
* Turn off all PWM-dimmed lights in the room, those LED ones and such. Often the radiate into mains and that get coupled.
* Air drafts. Your biggest enemy with LTZ. Your task to have as still air possible. It also meaning filling with foam, just putting ref in the box not enough, as thermal gradients cause air to move.
* AC cooled room can oscillate. Meaning that your AC will be turning on and off to keep temperature setpoint, and that will cause temperature oscillation. I have it happening if I close the door in the room with gear and AC on. So I always leave door open and have AC only in the gear room, so hot air from outside dampen the temperature changes. By doing this I can reduce oscillation from ~1c down to ~0.2c. And it's no basement stuff (envy people with houses), just subtropical apartment building with thin walls and regular split HVAC unit.
-
#112 Reply
Posted by
Dr. Frank
on 17 Nov, 2016 19:16
-
Re travelling voltage standards I did a bit of an experiment today to see the effect of "careless" temperature management.
It started off cool here about 15C ambient/otdoor temp (0700) then heated uo to 38C max at about 1500hrs. My 'lab' started off at 21C ended up at 28C before I switched on air conditioner at 1700. The voltage reference was quite a bit hotter at about 34C.
I was also a lot more careless about door openings. The effect is dramatic. Moving references around will probably have a similar or even worse effect . I don't think the cargo hold is pressurised and can get really cold I believe.
One of the time-nuts actually dug a deep pit to get temperature stability for his frequency reference! Something to think about.
Below Blue trace is voltage on RHS, Orange trace is internal temp of 3458A, should be 37C (last cal temp)
Horrible jumps, even 7°C of fast temperature change won't explain these although the 3458A instrument has about 0.5ppm/K T.C. in that 10V range.
There's something fishy in your setup.
Frank
-
#113 Reply
Posted by
VK5RC
on 17 Nov, 2016 23:05
-
Thanks for the comments DrFrank and TiN
I agree it doesn't show just temperature effect, I suspect it is many factors. Unfortunately my current testing room is a lined and single insulated fibro-cement small room ie not very stable temperature wise. I am using twisted pair, shielded cables with a cloth cover over the terminals on the 3458, the other end soldered on the pins that go directly into the 03458—66509 which is in a metal box inside a plastic container.
My previous results with 'kid gloves' were good, i plan to experiment to see to what part of the later 'set' up results in these fluctuations. The lights - a smpsu (240-12v)with 12v LEDs look like a potential culprit, it will be interesting to prove that or not. I have not taken one LED (Phillips Master series) apart to see if any internal switching. I have not had any RF (amateur bands) interference with them to date.
-
#114 Reply
Posted by
BradC
on 20 Nov, 2016 11:50
-
Moving references around will probably have a similar or even worse effect . I don't think the cargo hold is pressurised and can get really cold I believe.
I would have thought moving references around would only have a dramatic effect if the extreme temperatures caused permanent changes in the characteristics. Letting things stabilize for a couple of days *should* allow things to return to normal again, otherwise how would meters make it around the globe and still maintain calibration.
I certainly don't intend to put my meter and reference in the hold of a 747 and have it log as it travels around the globe (although I'd jump if given the opportunity!).
-
-
Air travel for precision standards are not recommended, the temperatures in the cargo hold can get quite cold, below -30°C and that can have a lasting effect on the standard. While Evanohm based standards are more tolerant, they can still be affected to a significant degree when demanding better than 1 or 2 PPM accuracy, you can expect hysteresis of a few tenths of a PPM if you're lucky. Manganin based standards will shift even more. Shipping during the winter is likely to be even worse than the summer for airplanes. Ground transportation should always be used if the standard must be shipped, granted it takes longer but it is relatively safer. One other bug-a-boo, gorrilla handling in extreme cases can also cause a shift in value, standards always require kid glove handling for best long term stability. This handling also goes for top level instrumentation, for the 3458A, to maintain best accuracy in high demanding applications, the instruments are 'hand' carried by ground transportation to the calibration lab, here in the states, that is in my home state of Colorado, Loveland to be specific and that service is around $1,800 from Keysight.
Since shipment produces something of an uncertainty into the equation, even for Evanohm resistors, a precise reading of the resistor(s) at manufacture may be changed slightly if poor conditions are encountered during shipping, only another precise reading at the receiving end can tell how much, if any, change occurred. Even a highly stable resistor such as the SR-104 can be affected by bad handling. Depending on the 'local' conditions, a given standard may have a limited accuracy (uncertainty) possible, with time and handling, repeated accurate readings will indicate just how stable the standard is going to be within the 'local' area of use. No standard is absolutely stable even under the best of conditions, the change may be very small but change it will.
-
#116 Reply
Posted by
Kleinstein
on 20 Nov, 2016 19:43
-
In normal plane the cargo hold is pressurized, but even than the pressure goes down somewhat, e.g. equivalent to 3000 m altitude. Also pressure changes with altitude - so this could be a problem for instruments at high altitude. Hermetic sealed parts can see some stress.
Sometimes if you pay for ground transportation, you may still end up with your parcel send by air, just at low priority.
One crude check for stress during transport is to measure the reference before you send it out and compare when it comes back.
-
-
I never thought that I would pay so much money for a single (and used!) resistor... . Nevertheless this addition to my home lab is rather nice and very welcome. Tomorrow I will get this cube to the HP3458A and see how time and various movements have affected it.
Cheers
Alex
-
-
Good score! These are nice resistors and it looks like yours was well kept. The rear decal shows no signs of tampering and the original alpha/beta values are still there.
Mine had loose binding posts and I had to open it. Someone also used some chemicals near the posts and it reacted with the plastic so beware of the urge to over clean it.
The black rubber feet all go bad eventually. They leave black residue on everything that comes in contact with them so I find myself removing them from time to time and replacing the bottom feet.
-
-
Good score! These are nice resistors and it looks like yours was well kept. The rear decal shows no signs of tampering and the original alpha/beta values are still there.
Mine had loose binding posts and I had to open it. Someone also used some chemicals near the posts and it reacted with the plastic so beware of the urge to over clean it.
The black rubber feet all go bad eventually. They leave black residue on everything that comes in contact with them so I find myself removing them from time to time and replacing the bottom feet.
Thank you, it appears that my unit is indeed in a very good condition. I don't know how old it is, possibly over 10 years old from the look of it. My measurements at 23.5C on a freshly calibrated HP3458A Opt 002 show about +5.5ppm value , so about +6ppm possible shift from the original measured value (or less than the uncertainty of calibration for HP). The deviation on the HP meter from the calibration value was only +0.3ppm though. So far I'm quite happy with my purchase.
Cheers
Alex
-
-
I never thought that I would pay so much money for a single (and used!) resistor... .
How much did you have to pay?
Was it over $ 500?
I am looking forward to your analysis with the 3458A
-
#121 Reply
Posted by
TiN
on 22 Nov, 2016 12:48
-
And since this thread derailed anyway, and if you have gear to run tempco test, I'd love to see the result on this 742A.
I wonder if it's worthy as travel transfer standard at ppm-level, or we still need oil-filled SR104. Manual and datasheet sorta vague about tempco spec.
-
-
And since this thread derailed anyway, and if you have gear to run tempco test, I'd love to see the result on this 742A.
I wonder if it's worthy as travel transfer standard at ppm-level, or we still need oil-filled SR104. Manual and datasheet sorta vague about tempco spec.
Strictly speaking, the tempco is printed on the back of each unit and according to Fluke papers, these values do not change much with time * (unlike the resistance itself which may drifts at <2ppm/year rate according to the datasheet). Unfortunately, to calibrate this 342A properly might cost me almost as much as I've paid for it! One day I will consider it, till then I have to survive with a possible error up to 7.5ppm
.
Cheers
Alex
* - it does not mean I won't measure the tempco when I'll have a chance. At the moment I don't have a proper temperature chamber close enough to the HP3458A, and the tempco of the meter itself is probably much higher than that of the 342A, so changes in the ambient temperature can not be used reliably. I didn't run ACAL ALL on the HP meter for about a week (only ACAL DCV daily), and the measurements on the 342A before ACAL ALL today were 5.5ppm higher!
-
-
I never thought that I would pay so much money for a single (and used!) resistor... .
How much did you have to pay?
Was it over $ 500?
I am looking forward to your analysis with the 3458A
It was about that figure (depending on the current exchange rate it may be just over or just under $500
).
Cheers
Alex
-
#124 Reply
Posted by
TiN
on 22 Nov, 2016 14:24
-
Tempco of 3458A can be limited pretty much to one of 40K resistor, if you add code into logging script to run ACAL DCV&OHM on each 1C of TEMP? change. That to be expected <1ppm/K.
What is printed is one thing, but it's interesting to see actual curve and it's slopes. It may or may not match the "typical" specs.