Solartron 7150Plus restoration

[masterx81]

Hi!
Finally, i've decided myself to buy a "precision" DMM, as i was searching the best bung for the bucks i've seen only few options, like Fluke 8840a, HP 3478A, Keithley 199, etc. The Keithley 196 was really nice, but can't be found, also "don't working", cheap.
At the end i've bought a "not working" Solartron 7150 Plus, EX NATO, sold as failed. Changed the "smoke maker" input filter and it's working back as normal (with some bad smell, but a good clean must do the job).
This unit has quite nice characteristics (to my eyes, better than the keithley 199).
Now i'me doing the mods suggested here:
https://holzleitner.com/el/solartron-7150/index-en.html
(thanks to the aurthor!)
So a good clean (near the exploded input filter), new white leds, security input sockets, fix the front panel that was a little bit damaged applying/removing the calibration stickers, etc.
Now i want to backup the eprom/earom contents. As my unit have the "combo" mcu/eprom instead of the OTP mcu, i can back it up easily (so, n°2 eprom 27c** to backup), the problem came with the earom units (2 of them), that save the calibration constants...
I find only really few informations about this devices, i can't neither find the package pinout (that eventually can be reverse engineered). And also can't be found any replacement part.
The best thing that i've found is this:
https://ballyalley.com/documentation/nitron/Nitron%20Catalog%20(1981)/Nitron%20Catalog%20(Revised%20Nov%201981).pdf
They are DIP8 package, serial addressable EAROM code NC7033 (orc NCR7033), 21x16 memory organization.
I can't find the protocol that can be used for reading/programming them. Would be nice to find a modern substitute to use with an add-on board (for signal shifting, for sure at least for the vpp), but by now i'm only interested to "refresh" the earom content, only to be sure that the calibration values doesn't disappear (and maybe have a valid backup). The last calibration seem in 2001, this earoms seem that can mantain the charge AT LEAST for 10 years, so isn't a bad idea to do something...
Maybe someone know a way at least to read and write them via gpib?

Thanks!

[masterx81]

Maybe in the user manual ( http://www.ko4bb.com/manuals/2.34.22.213/Solartron_Schlumberger_7150_Plus_User_Manual.pdf ) at paragraph 5.3 i've found a way to refresh the earom content, and eventually read via the display the values. I think that's enough for a backup. Hope that this memories not fail soon, as they seem impossible to find...

[masterx81]

And here a photo of the unit...

One question. It's normal that at poweron (also after some time that it's powered on) there is a small offset?
As seen in the photo there is a 34/35 uV offset.
For sure can be eliminated with the NULL function, but i not know if it's ok.

[Andy Watson]

35\$\mu V\$ is of the order of thermal effects due to junctions between dissimilar metals - i.e. a thremocouple. Your photo shows the test leads disappering somewhere (?). Measuring such low levels requires careful management of the temperature(s) along the measurement path. I wouldn't be too worried about it.

[masterx81]

I've tried to short the 2 plugs with different leads, and have always that little offset. Change a bit from cold to hot, but never disappear (until i null it).
The banana cable plugs are nickel plated, as also the female plugs (i've replaced them with the same plating that had the original plugs).
I've compared it with a chinese ad584 reference checked with a 34401 (the one on the meter) and when hot i have near 10uv difference on all the 4 voltage outputs (2.5, 5, 7.5 and 10v). I think that is quite good. My bad i discovered that my fluke 87-iv is reading a bit low...
I've bought this solartron with the intention to use it for calibrate the other measuring instruments...
As soon that i'm sure that it's electrically ok i hope to find someone that can check it with a calibrated 6.5 meter (i not need a certificate, only home use). Need to play with it more time and check all ranges.
In the calibration menu i've found how to refresh the earom content, so for now i'm sure that it not loose the calibration values. I've also seen that maybe i can only calibrate the "0" of the ranges, maybe i can simply fix that little offset.
I'm tempted to wash a bit the boards with some IPA alcohol (for example the smoke caused by the failed input filter can have leaved some deposits  around) , but i think that i can do more damage than benefits. During the operations with the device open i had care to not touch the analog board (only on the edges, far away from traces).
First time that i handle a so sensitive device so i'm a bit worried to do some mess 

[masterx81]

I've had to fix also a broken wire on the meter, as the back inputs were not working.
Then i've cleaned the analog board with ipa alcohol, not changed much, offset still present
It's the case to check the electrolytics? Or maybe other things?
I've not noticed anything out of ordinary, the measurements seem "good", and seem to work as it should.

This thing takes ages to stabilize at warmup. After 1h of power on, still changes the readings. After 2h seem to be quite good, and the full warmup seem after 3h (but the same offset is always here). Need to leave this thing always powered on
And in any case, the last digit in 6 1/2 mode is not so useful as it jumps a bit.

Now, time of some comparison.
Korad kel103 load, lb02 process calibrator, fluke 87-4, deree de5000 (for resistance) and as "precision" voltage source a few months old chinese ad584kh, that seem checked against a 34401. Also a korad ka3305p as psu.
The kel103 is brand new, the 87-4 never had a calibration, and the lb02 process calibrator it's 1y old but never read different from my "cheap standards". As said the last calibration of the solartron is in 2001.
The de5000 have proper kelvin clamps, the fluke use brymen leads with theyr clamps (short circuit res of 0.16ohm measured from the same meter), for the solartron have bought proper kelvin clamps but aren't arrived in time, so in 4w i've used 4 normal banana wires, with a clip at the end, so no shielding. Before every measure i've null'ed the meter.

First, resistance mode, against few 1/8w 1% resistors.

1.8ohm
De5000 rp 1.807
De5000 dcr 1.80
Solartron 2w 1.75
Solartron 4w 1.821
Fluke 1.88 (- 0.16ohm of inputs shorted)

1.8k
De5000 rp 1.969k
De5000 dcr 1.7977k
Solartron 2w 1.797060k
Solartron 4w 1.797076k
Fluke 1.7985k (- 0.16ohm)

R 18k 1%
De5000 rp 17.959k
De5000 dcr 17.965k
Solartron 2w 17.95757k
Solartron 4w 17.95757k
Fluke 17.972k

R 180k 1%
De5000 rp 181.14k
De5000 dcr 181.19k
Solartron 2w 181.1560k
Solartron 4w 181.1570k
Fluke 181.32k (- 0.16 ohm)

The 180k was difficult to measure as i think that the measuring current was self heating a bit the resistor.

For the Dcv range, using the ad584kh ref.

Source (declared): 2.50017
Solartron: jump a bit 2.50015 to 2.50020
Korad kel103: 2.5004
Fluke 87-4: 2.4991
Lb02 calibrator: 2.498

Source (declared): 5.00098
Solartron: jump from 2.500096 to 2.500100
Korad kel103: 5.0010
Fluke 87-4: 4.9989
Lb02 calibrator: 4.997

Source (declared): 7.50179
Solartron: jump from 7.50183 to 7.50188
Korad kel103: 7.5015
Fluke 87-4: 7.499
Lb02 calibrator: 7.496

Source (declared): 10.00215
Solartron: jump from 10.00218 to 10.00225
Korad kel103: 10.001
Fluke 87-4: 9.997
Lb02 calibrator: 9.995

Interesting thing is that the kel, the fluke and the lb02 never change the measure also after 1h that are powerer on, no jumps. For sure going after the 4 1/2 digit need a different level of stability.

Impressedby the precision of the voltage measurement of the kel103 load. Until today, as fluke and lb02 was telling almost the same voltage, i tought that the chinese ref and the kel wasn't correct, as i was trusting a bit more the fluke, but i think that i was wrong...

Now the ma range, as source the ka3305p, cc limited by korad kel103.

Korad kel 103 (setting):0.00000a
Ka3305p: 0.000a
Lb02: jump from 14ua to 20ua
Fluke: from 15 to 20ua
Solartron: 20 to 22ua

Korad kel 103 (setting):0.0005a
Ka3305p: 0.000a
Lb02: jump from 490ua to 500ua
Fluke: from 480 to 490ua
Solartron: 483 to 485ua

Korad kel 103 (setting):0.0010a
Ka3305p: 0.000a
Lb02: jump from 1.240ma to 1.250ma
Fluke: from 1242 to 1251ua
Solartron: 1.248 to 1.253ma

Korad kel 103 (setting):0.0100a
Ka3305p: 0.007a
Lb02: jump from 9.648ma to 9.663ma
Fluke: from 9.653 to 9.657ma
Solartron: 9.663 to 9.668 ma

Korad kel 103 (setting):0.0200a
Ka3305p: 0.018a
Lb02: jump from 19.930ma to 19.956ma
Fluke: from 19.923 to 19.928ma
Solartron: 19.937 to 19.939 ma

Korad kel 103 (setting):0.1000a
Ka3305p: 0.098a
Lb02: excluded, max 22ma
Fluke: from 99.72ma
Solartron: 99.798 ma

Korad kel 103 (setting):0.1000a
Ka3305p: 0.098a
Lb02: excluded, max 22ma
Fluke: from 99.72ma
Solartron: 99.800 ma

Korad kel 103 (setting):1.000a
Ka3305p: 0.098a
Lb02: excluded, max 22ma
Fluke: from 0.9999a
Solartron: 1000.251ma

Then i've tried excluding the load (for have a more stable reading (as the load limit current with switching pulses, but in any case also at low currents it's not that bad). By now i not know how measure current with it without limiting, so i simply bypass it. I've tried using the current limiting of the psu, but the reading are not stable as jumps in the order of 10/20ua (suspect that the psu current limiting is also not so stable).
So, always bypassing the kel load, i use the psu in cv mode, via a 2w 0.56ohm resistor. This time reading are way more stable.

Ka3305p (setting): 0.01v
Lb02: 0.066-0.067ma
Fluke: 66.83ua
Solartron: 0.066ma

Ka3305p (setting): 0.05v
Lb02: 0.330ma
Fluke: 329.80ua
Solartron: 0.333ma

Ka3305p (setting): 0.30v
Lb02: 1.901ma
Fluke: 1901ua
Solartron: 1.907ma

Ka3305p (setting): 1v
Lb02: 16.865ma
Fluke: 16.852ma
Solartron: 16.866ma

The solartron seem read few ua more, but nothing that bad.

I've learned some things... past the 4 1/2 digits, things are getting a bit difficult.
For resistance values, i think that i can trust for precision both the solartron and the de5000, and i think that with proper kelvin clips the solartron can be much better. The fluke is ok for fast checks.
For voltage values, seem that after the 3h of warmup the solartron is quite ok. Need to check in higher voltages, but i not have any good reference, and at higher than 10v most of the times i not need absolute precision. For now seem that is more urgent to calibrate the fluke than the solartron.
For ma ranges, all seem quite good.
Now i need to find in europe someone that can check the calibration of the fluke, or better, i can calibrate the solartron and then calibrate by myself the fluke...