DMM Volts/Ohm Calibration Club - anyone interested?

[enut11]

Hi @beanflying
Long way from me. I am on the east coast just south of Sydney.

Temperature is king when it comes to Vrefs. Open refs respond to air currents and other influences. Try a metal box with insulation. One of our members has been working on a Vref oven with control down to -/+0.1C
enut11

[beanflying]

If you have ever flown competitive Gliders or Electrics we would know some of the same mob 

Should have them in the little Deltron enclosures today or tomorrow to start burning in. 480mA will run over a week while I get the charging boards built.

[beanflying]

Quick question on enclosures. Did I read correctly somewhere to leave the gaskets out? Need to know how much space I don't have left 

The posts are gold (chinese  ) plated brass, brass washers and some brass knurl nuts I have sliced in two to hold the board. The posts will still be soldered to the output of the reference rather than using the brass to copper junctions. Basically removing any iron from the output path I can easily was the plan.

Breadboard hack up got up to 5.00003 and as it cools a touch sitting at 5.00002.

[BradC]

Breadboard hack up got up to 5.00003 and as it cools a touch sitting at 5.00002.

How much has your room temperature fluctuated? How much of that drift is the ref and how much is your meter?

[beanflying]

The great unknown 

Over the afternoon raised by about 8 from this morning when I first hooked up but it jumped quickly after 30 minutes settling. Top of the chip is now at 24.5 degrees and reading 4.999998. Based on bulk I am sure whatever the Meter is drifting it will lag behind the exposed IC.

The bits will be in a box in a couple of hours, still have room to cram a charging and monitoring board and an RTD next to the chip over the next week.

edit: Just dropped the second 6350 on the bread board running at 4.99982 as the temperature drops to about 17C

[Andreas]

Hello,

my 2 MAX6350 samples drift around 10 ppm/year.
So compared to many of my AD586LQ (after 6 months run in time) this is around factor 5 more.

with best regards

Andreas

[beanflying]

Mmmm I have a 3rd enclosure too 

More bits for my next Mouser order ............

[TiN]

I would avoid lithium packs due to shipping restrictions .

[beanflying]

Internal shipping on small LiPos by road post is fine https://www.master-instruments.com.au/files/product_warranty/customer_guidelines_for_lithium_batteries.pdf.pdf

Most Freight companies in Australia have a dangerous goods area or shipping option too.

LiPo's can be shipped by air but there is plenty of hoops to jump through so not worth the effort for this sort of thing. We have shipped multiple dozens of 6S 5000mAh batteries for model aircraft use internationally.

Generally LiPos get shipped by post without the declarations on the parcel anyway 

Part of the reason to maximise the LiPo in the enclosure was to allow it a 7-8 day unplugged life to allow for moving around powered up.

[beanflying]

Now 28 degrees and it has flicked up to 01.

Still some bits to add in the box. Will be adding the trimpot but it will have an external jumper to add it or not.

[beanflying]

So now I am officially broke but the owner of two 34970a's and 4 modules. 

Get in quick these are a steal thanks to Toyota pulling out of the local car industry https://www.ebay.com.au/itm/Agilent-34970a-Data-Acquisition-Switch-Plus-2x-modules/173019637414?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2055119.m1438.l2649

Drop Rohan a message tell him Tim sent you and ask about the 1 and 2m GPIB cables. There was initially 34901a, 34903a and 34907a modules available. He will pack and post items.

Also some Power Supplies listed and Sig Gens coming up but I am out of pocket money .........

[mimmus78]

Beware of those fake copper transparent binding posts ... I get a couple of them few months ago and if you cut in half you'll find they are not copper.

[beanflying]

Fairly comfortable they are Brass based on some filing and magnetic response. The Nuts, washers and soldering washers were certainly iron based on the couple of different ones I got in so I junked them in favour of plain brass and silver based solder. For their 3 and 4 1/2 digit testing job I wasn't to worried about emf precision.

[VK5RC]

I have not been completely volt-nut lazy over the past few months but doing a bit of an experiment - mainly to help me understand voltage references - in both long term drift , temperature and construction of Voltage references. The basis has been TiNs Kx board - again thanks Ilya for sharing your efforts. The reference that TiN evaluated is No 2.
I have 11 voltage references , 10 of which have been powered on continuously since the experiment started in August 2017 (UPS). All Kx have a 2 stage linear regulation power supply, 1st to 18V then to 15V.
Each reference is slightly different - by accident initially - but then expanded.  Below is a table of their differences

Ref No    LTZ Version   Kx Board version   OpAmp    Resistor type      Temp Resistor
   1                 A           early                    2057      Vishay z202s             12k       
   2                 A           early                    2057          Pettis                    12k
   3                 A         early modified         2057         Pettis                     12k
   4                 A              later                  2057     120 Vish, TE-UPW      13.3K
   5                 C              later                  1006         TE-UPW                 13.3k
   6                 C             later                   1006          TE-UPW                13.3k
   7                 A             later                   1006          TE-UPW                13.3k
   8                 A             later                   1006      120Vish, TE-UPW       13.3k        Vish S102
   9                 A             later                   1006      Pettis120, TE-UPW     13.3
 3458    A 3458s LTZ reference board
Fluke 731b
The boxing/insulation of 4 to 9 are identical.
The set up for each measurement has been similar, 3458 ( a very old one) on for 12 hrs before, room temp >17 for 4 hrs before, heaters/lights off during measurement, DC ACAL, NPLC 100, the same low EMF (Pomona) cables used. Cables left connected for 5 mins before reading taken - draft cover on.

I was a little surprised with the longer term stability  of the measurements, I thought I would see more drift, I suppose I am trying to measure "one ruler with 9 very similar rulers" - I can't get around that short of a JJA ! My findings seem to support the comments of some of the more experienced of you do seem to get some "golden" LTZ - some seem to be doing better.
 -
The graphs below are
1 Orange line; the references change in ppm compared to its average
2 Blue line; the room temperature change in degrees C compared to its average.

The experiment continues.

regards Rob

Edit ; You can see the set up in the photo in  Reply 250 above, the Refs 4-9 are on the RHS, Ref 1, 3 and 3458 are inside the Black Box, Ref 2 is outside.

[TiN]

Ref 7 is a stinker  

[enut11]

Well done VK5RC. A lot of patient work there. Have you picked a favourite yet?
Further explanation of how you generated the the graphs might be beneficial.
enut11

[VK5RC]

Hi enut11,
The data are all manually processed in Excel, the readings down to uV level are taken by hand.
The room temperature was taken from my trusty Keysight U1253B (no the display hasn't faded yet ! HiHi)
Voltage averages for each reference were calculated and then the individual readings from the 3458 subtracted then multiplied by 10e6 to get ppm variation. The orange graph is this in ppm.
The temperature (blue) curves were done the same way- an average room temp was calculated and the variation from that is shown in C for each reading set. I did it that way so I could use the same axis readings and not be too far out.

I need to get some more data points so I can get enough to do some correlation coefficients of volts vs date (drift) and Volts vs temp.

Ref 7 and 8 are looking pretty good -but I was suprised as HP only rates the 3458 around 8ppm/yr - this is looking a lot better than that - perhaps HP again underquoting.
Regards Rob

[AG7CK]

I have noticed that the US Cal Club thrives. But from recent posts or lack thereof, it seems that the Aussie club is less active. Anyone keeping anything going?

If not, I would like to see a low key, laid back and minimum stress Asia/Pacific(/World) thing. The guy(s) with the best meter(s) closest to cal come forward and offer to help. In exchange they get paid for helping others with "transferring" 10v / 7.xv and checking short time stability / drift (1h/24h) and stdev noise. Add resistance for those interested. No big packages.

IMO it should not be free. Look at the US thing. Months of waiting. Payment weeds out the "nice to be in - it's free" crowd.

Just my 2 satang.

PS I sold my 2x LTZ1000 boards a few years ago. Now I am in the market for a loaner before I finalize my refamp builds and buy a new LTZ1000 chip or board.

I pay shipping both ways, a decent rent and insurance ("real cost or more" deposit) for a ref board "close to real" 10v that I can borrow 2-4 weeks.

[BradC]

I have noticed that the US Cal Club thrives. But from recent posts or lack thereof, it seems that the Aussie club is less active. Anyone keeping anything going?

Not at the moment. We got 2 go-arounds, but we had a couple of fairly long stalls where the references sat and gathered dust. I was the last recipient this time around, but I encountered some issues with my measurements that made me reticent to post results.

In all honesty it's quite $$ to shunt references back and forward across Aus, and I dunno about the other guys but I found limited value in it for me given I had little confidence in my own measurement setup and my reference was by far the worst of the bunch. Personally, I think we'd have to come up with some more defined goals and outcomes before attempting it again.

On the other hand, if & when I want to get some references independently measured, I have people I trust now that I can send them to who have setups that are more accurate than mine.

[AG7CK]

BradC

Thank you for reporting.

I have heard about the results of sending a bunch of refs of varying goodness around without intermediate checking.

What I would like to do is similar to the US club reference going around in a star pattern (or at least frequently being checked by someone with calibrated gear). Also the LT1021 mini reference seems to be a success (on a more modest level) always returning to the owner for checking.

I just wanted to check if there was any interest. I'll guess I just wait and see. If nothing turns up, I will try to buy a "calibrated" LTZ1000 board in the Buy/Sell section. I don't have the time or inclination to build, burn in and then beg someone for "calibration" now.

If a few people are interested we could put in some USD50-100 each for a group buy of an aged board for a few hundred or so here on the forum. Having it reside with the person with the most recent calibration and paying him for the "docking" service should make it possible for everyone to check their gear to at least 5ppm or so. We'll see ...

[enut11]

I have noticed that the US Cal Club thrives. But from recent posts or lack thereof, it seems that the Aussie club is less active. Anyone keeping anything going?

If not, I would like to see a low key, laid back and minimum stress Asia/Pacific(/World) thing. The guy(s) with the best meter(s) closest to cal come forward and offer to help. In exchange they get paid for helping others with "transferring" 10v / 7.xv and checking short time stability / drift (1h/24h) and stdev noise. Add resistance for those interested. No big packages.

IMO it should not be free. Look at the US thing. Months of waiting. Payment weeds out the "nice to be in - it's free" crowd.

Just my 2 satang.

PS I sold my 2x LTZ1000 boards a few years ago. Now I am in the market for a loaner before I finalize my refamp builds and buy a new LTZ1000 chip or board.

I pay shipping both ways, a decent rent and insurance ("real cost or more" deposit) for a ref board "close to real" 10v that I can borrow 2-4 weeks.

Hi @AG7CK
Yes, the Aussie Cal Club is currently in a state of hibernation, except for perhaps @VK5RC with his LTZ experiments.
The initial 2 rounds were about establishing a degree of confidence in uncalibrated quality digital bench meters. And from my perspective it worked well.
Also, I agree with @BradC, it is expensive to ship items around Australia so the next round will have to fulfill a higher need and be confined to more stable and accurate refs.
I dare say that the members may be experimenting in secret for the time being. I know one is well into double-oven temperature stabilisation and this should improve Ref stability by an order of magnitude.
I am near the end of an 18 months programme on 'running in' a number of LM399 refs and will publish results soon. A couple of the 6 chips look promising.
enut11

[VK5RC]

Hi enut11 et al
I have added a new experiment, I have found some nice sockets for some LTZs (DigiKey PN 1-1437508-6-ND) and built a couple of Kx boards (Thanks TiN for design and open sourcing it and SvanGool for OshPark-ing it ) with TE 3-5ppm resistors and 1006 OpAmps,  1k/13.3K and 120R.
I will burn in the LTZ for a couple of months then do some short term temp co studies to see if I can find another 'good' LTZ1000A (or two), I have a total of  6 LTZs to sort out for the new Kx board which will go into a HP 3245 as per TiNs mod on one channel of the HP 3245A.
From the latest measurements of the ongoing experiment, No 7 and 8 are still behaving themselves - despite the weather having changed here quite significantly (cool wet winter 7-13c  from hot dry summer 18-34C).
The Fluke 731B also looks pretty good.
One really interesting factor is the old 3458 itself is doing well - the overall drift of all of the references measurements  looks quite small now - suggesting the 3458 itself is behaving itself. 
 Rob

[enut11]

I have posted my latest results for six LM399 chips that have been powered on now for about 18mths.
https://www.eevblog.com/forum/metrology/lm399-based-10-v-reference/msg1696106/#msg1696106

The idea is for the best one to be part of a precision temperature oven experiment. The oven and controller were designed by a friend but you could also use a diecast metal box.
enut11