Author Topic: Repair of Solartron 7081 SN#718 (Read 26376 times)

meandeev · « **Reply #100 on:** May 18, 2020, 03:23:08 pm »

Hello branadic,

which LT posts did you use?

branadic · « **Reply #101 on:** May 18, 2020, 05:26:37 pm »

Low Thermal Model 2758 also available at wekomm for the S7081 and Pomona 3770 + 3760 (blue, guard only) for S7061.

-branadic-

Kosmic · « **Reply #102 on:** June 11, 2020, 02:16:37 am »

I did a quick print in PLA to test the new front panel. Was done really quickly, so this is not the final quality (and color ...). I will fix the horizontal lines and labels in my next version.

Look like my binding posts are similar to LT post. I had to increase the size of the hole though since the screw part on mine is bigger. Overall i'm pretty happy. Thanks again Branadic!

tggzzz · « **Reply #103 on:** June 11, 2020, 06:54:36 am »

Quote from: Kosmic on June 11, 2020, 02:16:37 am

I did a quick print in PLA to test the new front panel. Was done really quickly, so this is not the final quality (and color ...).

I presume you have a daughter, and that you are trying to help her realise that your toys and experience might be valuable to her

I must admit, for something as precious as a 7081, once I was satisfied with the design I'd probably want to get this professionally made. Nylon isn't a bad material, being slightly flexible, but a black SLA would be good too.

Are the PLA/ABS materials hygroscopic, or would gaps between layers trap gunk over the years? If so they wouldn't be good w.r.t. high impedance or high voltage.

Kosmic · « **Reply #104 on:** June 11, 2020, 02:00:57 pm »

Quote from: tggzzz on June 11, 2020, 06:54:36 am

Quote from: Kosmic on June 11, 2020, 02:16:37 am
I did a quick print in PLA to test the new front panel. Was done really quickly, so this is not the final quality (and color ...).

I presume you have a daughter, and that you are trying to help her realise that your toys and experience might be valuable to her

Yep got 2 daughters

Quote from: tggzzz on June 11, 2020, 06:54:36 am

I must admit, for something as precious as a 7081, once I was satisfied with the design I'd probably want to get this professionally made. Nylon isn't a bad material, being slightly flexible, but a black SLA would be good too.

Are the PLA/ABS materials hygroscopic, or would gaps between layers trap gunk over the years? If so they wouldn't be good w.r.t. high impedance or high voltage.

My plan is to finalise the form factor and then try different materials (PETG, ABS, PLA). I'm also planning to use paint or epoxy to seal the material. I have a high resistance meter so I should be able to do some interesting tests.

branadic · « **Reply #105 on:** June 17, 2021, 05:43:27 pm »

For some reason my unit started getting Alzheimer, throwing out "NVM Fail". The maintenance manual mentiones to replace the ER3400 by a new one: "NVM FAIL - This could indicate a failure of the NVM circuitry, or of IC804. A spare NVM would prove useful here." which I did, but that didn't change something. I wonder if someone ever had a similar problem and can share experience. I guess it's time to check some power rails.

-branadic-

branadic · « **Reply #106 on:** July 10, 2021, 10:11:06 am »

So I think I'm a step further. Turns out replacing the NVRAM requires full calibration. Actually it's a full adjustment instead. Unfortunately, I don't have any AC calibrator or signal source to complete the calibration, which is essential to at least save the calibration constants to the NVRAM.
What's even more weird is, that there is a possibility to dump the calibration constants, which I did in the past, but there seems to be no way of playing them back to the device. I wonder what this CALIBRATE, DUMP option is then for, if you need to fully adjust it in any case. So unless there is no hidden "play your dumped calibration constant back to the unit" command, I guess I need someone helping me out to perform the whole calibration process at once

-branadic-

perdrix · « **Reply #107 on:** July 10, 2021, 11:18:00 am »

If you happen to be in the UK with your meter, I have a Datron 4808 calibrator

PS does anyone have a spare earthy logic board for one of these - I have a bad one I have totally failed to fix

David

MiDi · « **Reply #108 on:** July 10, 2021, 03:45:44 pm »

The NVRAM saves the cal constants twice, if only one version is faulty, you can rewrite it by command CALIBRATE, REFRESH.
I guess you know that and the ER3400 had no correct version any more?
You have read the cal constants over the interface, is this a correct cal before the unit got Alzheimer?

If you can read the NVRAM with external programmer, it may be possible to reconstruct a correct cal and write it to a new NVRAM/EEPROM.
Maybe not that easy to find a compatible chip that is supported by programmers today or even make a custom config file for those.

You could ask Adrian, he should be able to calibrate/adjust this units.

branadic · « **Reply #109 on:** July 10, 2021, 06:08:35 pm »

@ David,

I'm sailing under the german flag

@ MiDi

unfortunately I only have a TL866 II, which seems to not support this ER3400 chippy

Already in contact with Adrian, but thanks for the hint anyway.

-branadic-

picburner · « **Reply #110 on:** July 10, 2021, 06:55:27 pm »

No commercial programmer, not even my Data I/O 3980, is able to program an EAROM ER3400.
Look what I had to do to be able to program it.

branadic · « **Reply #111 on:** August 17, 2021, 07:49:30 pm »

Just an update.
Replacing the NVR requires a full adjustment. Thankfully Dr. Frank help with that. Meter is back since yesterday. However, it turned out at Frank that NVR issue had another root cause, the electrolytic cap for the programming voltage was bad. I had checked that rail, ripple and the cap before, but couldn't find an issue there. Frank the "Prüfgeräteelektromechaniker" however spotted the problem, fixed it by replacing the cap and completely adjusted the meter. Thanks again

I found the ADC zero tc mod by MickleT. but couldn't find any description on it, even on radiokot.ru. Does someone have a link, where this mod is described in detail?

-branadic-

Mickle T. · « **Reply #112 on:** August 17, 2021, 08:05:28 pm »

It's a simple correction (balancing) of R305 TC with copper wire.

Dr. Frank · « **Reply #113 on:** August 17, 2021, 09:56:39 pm »

Well, I'm even a 1st EloPrfGerMech (w/o electro) from the German AF

. All the stuff I repaired and calibrated in 1980.. 1982, was all electric, though..

The EARAM needs a -30V programming voltage, which is generated by a charge pump on A5, Floating Logic board. C804 was a 47µF, 50V electrolytic made by Philips, which had 3nF left only, so not delivering sufficient current anymore for the erase/programming cycle. I assume, that this killed the content of the old chip, not the chip itself.
Owners should check / replace this capacitor with 100µF/50V (which I have seen in other units).

In the service manual, chapter 8, Monitor, Calibration and Self Test, there are hints how to access (read) the EARAM Nibble by Nibble, using the MONITOR via RS232, and this should be done for the whole 1024 Nibbles, (2 pages of 256 byte of cal data) to copy the calibration constants plus the parity checksum.
There seems to be the facility to write back this low level storage information also, as there is no GP-IB command to write back the real numbers to EARAM.. similar to the 3458A, where you can only dump the calibration constants, but not write anything back from outside.

The manual is not precise at all how to do this exactly, and I did not try any more.. I had a lot of trouble with the interface programming - Handshake / Arming and Triggering methods seems to be implemented quite unstably, compared to hp instruments.

Anyhow, a quite nice instrument for its time, and still very stable, electrically..

There had been several of these in our stock @ RWTH Aachen 2. Phys. Institute, and my prof. initially requested me to overhaul and use these .. but I was very happy to instead purchase the new hp3458A in 1989.. due to being a bit faster ...

Frank

meandeev · « **Reply #114 on:** August 19, 2021, 07:28:10 am »

Quote from: Mickle T. on August 17, 2021, 08:05:28 pm

It's a simple correction (balancing) of R305 TC with copper wire.

Hi Mickle T.,
do you mean this way?

Mickle T. · « **Reply #115 on:** August 19, 2021, 07:29:53 am »

Yes, it is.

meandeev · « **Reply #116 on:** August 25, 2021, 06:54:48 pm »

I made the same test and have nearly the same absolute drift value, but in the opposite direction - so should I put the correction wire in the upper part of R305?

meandeev · « **Reply #117 on:** August 27, 2021, 08:00:28 pm »

I made some tests to decrease the TC (in the upper part of R305a):
-started with 63cm copper wire (diameter 0,20mm): better, but not good
-then with 130cm copper wire (diameter 0,20mm): too much, TC switched to the opposite sign
-last with 100cm copper wire (diameter 0,20mm): I´m satisfied with that

branadic · « **Reply #118 on:** August 27, 2021, 08:48:40 pm »

Nice that it worked for you too.
I've checked some old pictures I took and found that this resistor arrangement looks different on my unit.

-branadic-

meandeev · « **Reply #119 on:** August 27, 2021, 09:34:57 pm »

looks like a Vishay hermetic resistor. It should be a double 10k/10k, isn´t it? So you can bend one pin - and wind some wire around the resistor too. Maybe you should use an thinner wire, so it can be shorter because the resistor has not so much space.

Mickle T. · « **Reply #120 on:** August 28, 2021, 05:12:10 am »

It's a Vishay 312832 - custom made divider.

branadic · « **Reply #121 on:** August 31, 2021, 07:34:53 am »

I was asked to share the CAD drawings as a STP file, so here you find the packages for both, the Pomonoa and LT Model 2758 version, including the front panel as STL and STP, but also the complete assembly with the binding posts in a STP format.

https://www.eevblog.com/forum/repair/repair-of-solartron-7081-sn718/msg3058252/#msg3058252

-branadic-

rigrunner · « **Reply #122 on:** September 14, 2021, 04:50:21 pm »

Branadic's work realised in Formlabs Tough 2000 resin.

branadic · « **Reply #123 on:** February 25, 2022, 10:49:40 pm »

Realized there never was an image of the final units, here is the 7061 with Pomona 3770, that I recently sold and an image of Solatron 7081 with Low Thermal 2758.

-branadic-

Kosmic · « **Reply #124 on:** November 17, 2023, 07:39:41 pm »

FYI, I reproduced the feet of my 7081 (to be used on a 7061). Files are in the feet thread here: https://www.eevblog.com/forum/testgear/replacement-knobs-feet-and-fittings-for-test-equipment/msg5173905/#msg5173905


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Repair of Solartron 7081 SN#718 (Read 26376 times)

Share me