Solartron 7081 #128 Repair

[Hedrix]

Hello,
first post here.

I have recently aquired a Solartron 7081, SN128 on ebay. Advertised with the "NVM FAIL" message.

First impressions:
 - front/rear switch is missing
 - some screws came loose
 - both batterys are dead, the mempack is growing crystals
 - CAL key is missing, as expected.
 - some signs of bodges, like the SMD capacitor? soldered on R711 on the floating input board.
   (those may be original/from previous services)

Exactly as advertised, it boots up to the NVM FAIL error message. Communication via RS232 and GPIB seems to work, but
dumping or refreshing the calibration constants caused no response from the instrument, aside from a 'ok'. Reinitialisation via the Monitor also did not bring any sucess.
After aquiring an replacement ER3400, the Solartron finally allowed initialising the EAROM. Instead of the NVM Fail message i am now greeted by the CAL INCOMPLETE Message,
as expected, according to the service manual.

This is where im stuck now. dumping or checking the test-constants remains unanswered. Poking around in the Monitor mode, i can read individual adresses, and the values are
as expected ( 0A 05 alternating).
Also calibrating, e.g. short-circuit conditions appears to be not working, at least the contents of the ROM dont change.

I noticed, that when printing out memory contents(P aaaa,bbbb) from the floating board, the first few characters (see attachements) are recognized and echoed back correctly,
but when writing more characters, some of them apparently get lost (p 0, 3FF is reliably losing either '3' or one of the 'f's).
This might be the reason why calibrate, dump et al are not working. (i can write them a few times, getting confirmed with 'ok', but after a few commands the instrument hangs up)
This smells like there is either a problem with internal UART communication or a problem on the floating processor.

Has anyone had the same problem here, or knows where to start looking?
My experience with older microprocessor systems is unfortunately rather limited.
Also, is it safe to probe the floating processor board with the input board on top unplugged?

[Kleinstein]

There are reports of optocouplers going bad in these older instruments. This can cause communication problems.

[unseenninja]

At the bottom of your fourth picture, not so far from the EAROM is a small, blue Phillips capacitor. It is used in a circuit which generates the 30V programming voltage for the EAROM. Without a good programming voltage, you will not be able to calibrate. So replace the capacitor. There are five more on PCB 3.
If you test them, you will see low capacitance and very high ESR. I'd also advise you to check all the other electrolytics and replace as required.

The hang ups when trying to dump the calibration may indicate other supply rails not being healthy. Start with capacitor replacement, then check all the supply rails for the correct voltage and low ripple. Then, when all the rails are fine, you can try calibrating the meter.

[unseenninja]

Also, is it safe to probe the floating processor board with the input board on top unplugged?

Probably, but it won't tell you much as the boards are quite dependent on the signals which are in the flat cables that link them. That's why the upper board is designed to be folded outwards while still connected to the two red standoffs. That way, the ribbon connectors which go to the lower board will still be connected and you can get to test points on both boards.

I use a small wooden stick to hold the upper board in the folded out position. Just remove all the screws which hold the upper board in place and lift it up by the edge closest to the middle of the case, folding it towards the outside like the page of a book.

[Hedrix]

At the bottom of your fourth picture, not so far from the EAROM is a small, blue Phillips capacitor. It is used in a circuit which generates the 30V programming voltage for the EAROM. Without a good programming voltage, you will not be able to calibrate. So replace the capacitor. There are five more on PCB 3.
If you test them, you will see low capacitance and very high ESR. I'd also advise you to check all the other electrolytics and replace as required.

The hang ups when trying to dump the calibration may indicate other supply rails not being healthy. Start with capacitor replacement, then check all the supply rails for the correct voltage and low ripple. Then, when all the rails are fine, you can try calibrating the meter.

The 47µF capacitor measured fine, and the ER3400 can be programmed during initialisation.
Supply rails measure just fine, the -28V rails is a little short with 26 volts, but thats it.

There are reports of optocouplers going bad in these older instruments. This can cause communication problems.

i temporarily replaced the 2 couplers (6n136) responsible for inter-board communication with functionally equivalent but equally old HP 4503s, unfortunately no improvement.
Are there any others that might be responsible?

[perdrix]

Run, don't walk to http://www.perdrix.co.uk/Solartron7081/index.htm  At the bottom of that web page you will find among other links to download files, a link to download the "Solartron 7081 Test Spec" document which contains the full details on the factory calibration procedure.

You'll need an RS-232 port equipped system (or an good USB-R232 adapter).

There's quite a bit of other good stuff on that page too (though I say so as the author of that web page).

HtH, David

[tggzzz]

Seconded