Author Topic: USA calibration club  (Read 140198 times)

0 Members and 1 Guest are viewing this topic.

Offline hwj-d

  • Frequent Contributor
  • **
  • Posts: 676
  • Country: de
  • save the children - chase the cabal
Re: USA calibration club
« Reply #650 on: April 14, 2018, 09:52:05 pm »
I told him to keep it for a week or two and play with it.  So it looks like my DMM7510 may be a few PPM off one of his 3458's, so which is right  :scared:

Buy a second 7510  ;D
 

Offline kj7e

  • Frequent Contributor
  • **
  • Posts: 911
  • Country: us
  • Damon Stewart
Re: USA calibration club
« Reply #651 on: April 14, 2018, 09:56:09 pm »
I told him to keep it for a week or two and play with it.  So it looks like my DMM7510 may be a few PPM off one of his 3458's, so which is right  :scared:

Buy a second 7510  ;D

I'm waiting for the 8510  >:D
 
The following users thanked this post: TiN, hwj-d

Offline flittle

  • Contributor
  • Posts: 17
  • Country: us
Re: USA calibration club
« Reply #652 on: April 19, 2018, 08:50:05 pm »
Whew! I read the entire thread and I am already hooked.  I have an HP3455A, a Doug Malone 5v ref, a L&N 10K resistor and lots of unfinished projects! :scared:  Can I join the USA calibration club?
DS1054Z, HP3455A, HP3457A, Agilent 34401A, HP5334B-010-030, HP204D, EX430, Agilent 6612C, (2) Sorensen XTS15-4 /M1 /M9B, WaveTek 131, WaveTek 134,PAR 110, FG-8002,FY3200S, UNI-T61E, TEK2465
 

Offline cellularmitosisTopic starter

  • Supporter
  • ****
  • Posts: 1111
  • Country: us
Re: USA calibration club
« Reply #653 on: April 19, 2018, 09:02:53 pm »
Yes!  PM me your address :)
LTZs: KX FX MX CX PX Frank A9 QX
 

Offline SirAlucard

  • Regular Contributor
  • *
  • Posts: 133
  • Country: us
Re: USA calibration club
« Reply #654 on: April 21, 2018, 12:29:53 am »
So how does shipping to people actually work? Also cellularmitosis is your offer on the Geller Labs reference still available?
 

Offline cellularmitosisTopic starter

  • Supporter
  • ****
  • Posts: 1111
  • Country: us
Re: USA calibration club
« Reply #655 on: April 21, 2018, 01:21:21 am »
The Geller labs ref is still sick, so that’s out of rotation.

My plan this time around was to ship in a star pattern (the ref returns to me on each hop).  I had in mind that if you have an instrument which was calibrated recently, I’ll pay for shipping both ways. Otherwise, I’ll pay for shipping one way.

I made a small accessory board which has an output buffer, and I think I’ll install that this weekend and try to get this thing shipped out on Monday.
LTZs: KX FX MX CX PX Frank A9 QX
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3476
  • Country: us
Re: USA calibration club (transit environmental logging?)
« Reply #656 on: April 21, 2018, 01:21:01 pm »
I read back a few pages but didn't see any references to logging temperature during transit.   After hunting for a paper by Franco Rota on the noise diode sources he built for the 13th EME conference without success last night, I'm a bit annoyed with google.  I *know* I have a copy, but no idea where.

Is that being done?  If not, the MSP430 typically has an internal temperature sensor. So it would be dead simple to build a battery powered data logger that just rode around with a reference and logged the ambient temperature.  Not much more effort to log BP and RH.  Probably more work to select the sensors.  If the container is insulated, one would need only a few measurements per hour, so battery life would be very long as in months or years.
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: USA calibration club (transit environmental logging?)
« Reply #657 on: April 21, 2018, 01:50:15 pm »
After hunting for a paper by Franco Rota on the noise diode sources he built for the 13th EME conference without success last night, I'm a bit annoyed with google.  I *know* I have a copy, but no idea where.
You mean this one?
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3476
  • Country: us
Re: USA calibration club
« Reply #658 on: April 21, 2018, 03:39:54 pm »
There was another copy in color.  Same material, but the other copy had some additional information appended to the PDF.  I'm sure that google returned a link somewhere in all the pages.  After chasing it a while I punted and just used this for now.
 

Offline cellularmitosisTopic starter

  • Supporter
  • ****
  • Posts: 1111
  • Country: us
Re: USA calibration club
« Reply #659 on: April 22, 2018, 05:50:24 am »
Ah, damnit!  I think I just threw away the ~3 months of ageing on the board which I was about to send out.

I decided to trim the board a bit to fit inside of one of the TEKO cases, and accidentally connected VCC to the 2n3904 base, rather than the collector.  The circuit drew about 45mA, which I knew was odd, but by the time I traced down what was going on, the LTZ had been running too hot for about a minute.

It looks like it has shifted about 100uV down from where it was before the accident, and a new ageing cycle has surely begun.

The good news is that this means there's no reason for me not to assemble one of the newer PX ref boards which has an output buffer built-in.
 The whole reason I wanted to bodge in an output buffer to this older board revision PX ref was that I wanted to be able to take advantage of the several months of ageing it had accumulated.
« Last Edit: April 22, 2018, 05:53:09 am by cellularmitosis »
LTZs: KX FX MX CX PX Frank A9 QX
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: USA calibration club
« Reply #660 on: April 22, 2018, 09:50:54 am »
Hello Jason,

shi(f)t happens.
You can try to restore the old value (at least for a part) by power cycling the LTZ board several times.

Power off, let the LTZ cool down and switch on again for several cycles.
I think Frank had the Idea of putting the cirquit into the freezer. (didnt try that up to now).

Of course you will not get the same voltage as before but the shift will decrease somewhat.
Up to now I had about -2 to -5 ppm after cycling.

And yes, after such a accident I usually have about double the ageing rate as before for about 6 months.

good luck

Andreas
 
The following users thanked this post: cellularmitosis

Offline hwj-d

  • Frequent Contributor
  • **
  • Posts: 676
  • Country: de
  • save the children - chase the cabal
Re: USA calibration club
« Reply #661 on: April 23, 2018, 02:02:31 am »
Quote
I decided to trim the board a bit to fit inside of one of the TEKO cases, and accidentally connected VCC to the 2n3904 base, rather than the collector.
The TEKO's are somehow dangerous things.
 ^-^
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4543
  • Country: ua
    • xDevs.com
Re: USA calibration club
« Reply #662 on: April 23, 2018, 04:12:02 am »
Time to fire up TECbox again.

.

 :popcorn:
YouTube | Metrology IRC Chat room | Let's share T&M documentation? Upload! No upload limits for firmwares, photos, files.
 
The following users thanked this post: cellularmitosis

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4543
  • Country: ua
    • xDevs.com
Re: USA calibration club
« Reply #663 on: April 24, 2018, 04:15:16 am »
cellularmitosis, can you help to post/link me to resistor test measurement results on the ones you sent me?
They currently running my tempco test, so I am curious to compare data.
YouTube | Metrology IRC Chat room | Let's share T&M documentation? Upload! No upload limits for firmwares, photos, files.
 

Offline cellularmitosisTopic starter

  • Supporter
  • ****
  • Posts: 1111
  • Country: us
Re: USA calibration club
« Reply #664 on: April 24, 2018, 06:46:12 am »
Yes!   :-+

The number of dots on each resistor indicates which number it is.

My results start here: https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg1434052/#msg1434052

Andreas results start here: https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg1441601/#msg1441601

His summary of results is here:

https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg1457580/#msg1457580

https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg1458463/#msg1458463

The epoxy package resistor wasn't tested.

----

By the way, over the past week I have been fooling around with various setups to try and squash my EMI demons.  Just this evening I seem to have succeeded.

The winning combination seems to be three turns around a large cylindrical ferrite, inside of a TEKO case (however, with the ferrites, the board is nearly immune to noise even outside of the case).  I can no longer move the last digit by moving around my hands, touching the case, touching the meter, touching the battery, etc.    :-DMM

I haven't posted about this yet because I wanted to show the video footage I was recording, but I needed to figure out how to do picture-in-picture video editing.  I've just figured out how to do that: here's a sample of testing a battery for "hand-waving" sensitivity: https://youtu.be/Ifm13AZ-QFM

(site note: this little guy is just about small enough to fit inside of a 34401A...  >:D )

edit: the ferrites: https://www.amazon.com/gp/product/B016FC6HX0
« Last Edit: April 24, 2018, 06:51:43 am by cellularmitosis »
LTZs: KX FX MX CX PX Frank A9 QX
 
The following users thanked this post: TiN, Andreas, Kean, hwj-d

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: USA calibration club
« Reply #665 on: April 24, 2018, 05:44:25 pm »
When I do ppm bridge measurements with the 1615, I do that same few turn trick. I use a tape wound core because lower frequencies are involved. Haven't tried it for DC metrology yet.
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3476
  • Country: us
Re: USA calibration club
« Reply #666 on: April 24, 2018, 08:52:13 pm »
What inductance does the 3 turn choke have?
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: USA calibration club
« Reply #667 on: April 24, 2018, 09:27:49 pm »
Hello,

surprising good results. Would be interesting what frequencies your daemons contain.

EMI-Ferrites usually have no resonant frequency like a inductor.
They are lossy and convert the EMI into heat and thus have a broader frequency range.

with best regards

Andreas

 

Offline cellularmitosisTopic starter

  • Supporter
  • ****
  • Posts: 1111
  • Country: us
Re: USA calibration club
« Reply #668 on: April 24, 2018, 10:18:12 pm »
rhb, I'll try to use my cheap LCR tester to see what the inductance is.

Another thing I was surprised by (which I'll try to post more detailed info about) was that squashing the EMI changed the DC value (I've observed this before when placing a ref inside of my dutch oven faraday cage).  This made me realize that if the EMI is different across two environments, that could translate into additional DC uncertainty.  Perhaps I should also be looking at the AC value on my multimeter as I test the various configurations of ferrites and such.
LTZs: KX FX MX CX PX Frank A9 QX
 

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3476
  • Country: us
Re: USA calibration club
« Reply #669 on: April 24, 2018, 11:42:21 pm »
Welcome to why I want long time series and went to the trouble of doing a rigorous analysis of the thermal noise amplitude spectrum.  I've still not got to the 1/f noise.  You can't do anything about 1/f and thermal noise, but EMI is aliased and can produce significant errors.  Taking a long measurement series and examining the FFT should reveal the presence of aliased EMI, though not  the frequency. AM radio stations will be the biggest source of EMI in an integrating ADC.  Unfortunately, with a 3 Hz Nyquist, the audio modulation will spread the noise all across the spectrum.  The only indication it is present is that it does not match the amplitude-frequency curve of the 1/f or thermal noise.

My lab is in complete disarray.  I got an HP 5386A yesterday and an 8648C today (w/ a 34401A on the way ;-).  In addition I pulled out all the computers so I could get a couple more machines in and on the UPS'.

In a search on a completely unrelated topic, google threw up some AD584 boards for $3.  God only knows why. Probably an ad service they offer. But I couldn't resist, so I ordered 3 for the grand sum of $8.85.  I'm going to put them in an enclosure with the LM399s you gave me and see what happens.  These are the J parts, so I assume that the better units have been graded off.  One of the things that amazed me was the AD584 costs 3x+ more from Digikey than the eBay boards.  For what I want to do, the drift and TC don't matter so long as I  can quantify it accurately.

If I can trim the output voltage using an MCU and  PWM DAC driving a JFET  wired up as a variable resistor, I should manage to annoy a few people.

Have Fun!
Reg
 

Offline cellularmitosisTopic starter

  • Supporter
  • ****
  • Posts: 1111
  • Country: us
Re: USA calibration club
« Reply #670 on: April 25, 2018, 06:40:52 am »
What inductance does the 3 turn choke have?

I took a look at this tonight, testing the inductance of just one lead of the twisted pair, using a cheap LCR meter.

Results:

Code: [Select]
No ferrite: 2.4uH
0 turns: 3.1uH
1 turn: 5.3uH @ 721kHz
2 turns: 9.2uH @ 693kHz
3 turns: 14.9uH @ 657kHz
LTZs: KX FX MX CX PX Frank A9 QX
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1527
  • Country: au
Re: USA calibration club
« Reply #671 on: April 25, 2018, 07:32:21 am »
My expectation for apiece of straight wire of that gage would be about 1nH per cm length of wire.

Calibrate the aligator clip leads by measuring 0 cm long piece of wire.

Note that even without the ferrite You have one complete turn of undefined cross sectional area.

same logic applies to the notionally one turn through the ferrite.

I seem to recall inductance varies as square of turns.so the measurement should go in increments of 1 to 4 to 9 ... something very off.

How long is that piece of twisted pair?

EDIT:  that should be 10nH not 1nH
« Last Edit: April 25, 2018, 10:36:11 pm by IconicPCB »
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: USA calibration club
« Reply #672 on: April 25, 2018, 09:36:18 am »
Hello Jason,

shi(f)t happens.
You can try to restore the old value (at least for a part) by power cycling the LTZ board several times.

Power off, let the LTZ cool down and switch on again for several cycles.
I think Frank had the Idea of putting the cirquit into the freezer. (didnt try that up to now).

Of course you will not get the same voltage as before but the shift will decrease somewhat.
Up to now I had about -2 to -5 ppm after cycling.

And yes, after such a accident I usually have about double the ageing rate as before for about 6 months.

good luck

Andreas

The LTZ1000 output shows hysteresis, when temperature excursion is greater than about 30°C off from the nominal oven set temperature.
This hysteresis / offset might be permanent, or the output might slowly creep back towards its initial value, that depends on the individual LTZ1000 chip.
Slowly means on a years time scale.

There's the patent by Pickering, which might mitigate this hysteresis by temperature cycling the LTZ1000 oven temperature.. but I doubt that it works inside the DATRON / WAVETEK 7001 device, as the oven temperature is set to about 45°C, so the lower cycling temperature is room temperature only, which is too less to have an effect.

Anyhow, I run all my seven LTZ1000 at 12k/1k, or at about 45..55°C, and also encountered such oven temperature runaways.
In one case, that led to a permanent shift of -5ppm.

So I cycled the whole circuit (not powered, inside a plastic bag) to -23°C in the freezer, then let the oven run (outside the freezer)  at 80°C, then back to -7°C, to 60°C, 10°C, 60°C, 21°C.

That set the output back to its initial value within less than 1 ppm.
Afterwards, the drift rate was again as high as -1.5ppm/ year for the first year, so it's obvious, that the initial ageing process starts over after that cycling process.

I think, I should publish the long term drift measurements for my references....
« Last Edit: April 25, 2018, 09:42:10 am by Dr. Frank »
 
The following users thanked this post: TiN, dr.diesel, Andreas, cellularmitosis, msliva, hwj-d, 2N3055

Offline rhb

  • Super Contributor
  • ***
  • Posts: 3476
  • Country: us
Re: USA calibration club
« Reply #673 on: April 25, 2018, 02:25:11 pm »
Inductance of a toroid is L = K * N**2 where K is a material specific constant. [1]

The inductance of 1" of straight #19 wire is ~28 nH [2]

For a close wound aircore inductor L = (r*N)**2/(9*r+10*l) [3]
L - microHenries
r - radius in inches
l - length in inches

[1]  "Solid State Design for the Radio Amateur", Hayward and DeMaw, ARRL 1977 p 250
[2]  "Grounding and Shielding", Morrison, Wiley 2016,  p 35
[3] "ARRL Electronics Data Book", DeMaw. ARRL 1977 p 26

Finding that information was surprisingly difficult in more recent publications.  Ultimately I went back to what I used 40 years ago.

The test frequency is rather unusual.  It's also in the AM broadcast band. You may have an EMI problem measuring inductance.
 
The following users thanked this post: cellularmitosis

Offline Conrad Hoffman

  • Super Contributor
  • ***
  • Posts: 1928
  • Country: us
    • The Messy Basement
Re: USA calibration club
« Reply #674 on: April 25, 2018, 05:53:47 pm »
Though I have a collection of known Amidon parts (great company), I also have a lot of surplus stuff. If you wind an arbitrary number of turns, say 10-50, and measure value and loss, you can often pin down the materials quite well, even to a specific manufacturer if they have a colored coating.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf