Data Precision 8200 - Potential proccesor PSU issue?

[fenland787]

I have recently been fixing my 8200 which I got as non-working, turned out the main issue was just a couple of dead transistors in the output amplifier. Also found an intermittent offset adjust pot and one of the logic supply PSU caps was decidedly warm to the touch.

Replaced all the polarised caps and the pot and all seemed fine but I noticed that just brushing against the 5V regulator on the 'logic' card induced 3rd degree burns! It's a 7805 on a tiny heatsink and as I'm based in the UK our nominal 230VAC supply is mostly nearer 245VAC because (I think) as a nation we decided to claim we were 230V to make us the same as the rest of Europe by leaving everything at 240V but just extending the limits! End result is that the nominal 8VDC feeding the regulator was actually over 10 and the power dissipation in the linear reg was vast!

I've met this before so I replaced it with a pin-compatible switcher module and everything cooled down nicely, in so doing I discovered that my unit (ser #3600) seems to have has a slightly different logic card from that described in the only manual I've found, 1st Ed 1981, in that it has different uProc and an extra IC but more importantly the position of the voltage reg is wrong and no longer lines up with the chassis cut-out on which the PCB is mounted.

To sort this out Data Precision simply cut the soldered pins almost flush to the rear of the PCB, glued a bit of thin acrylic - not Kapton - sheet on the rear of the PCB and did the screws up anyway, yes, well the board bows a bit but what the heck!

Easily fixed it properly in 5 minutes with a small file to put a little notch in the side of the chassis cut-out to clear the pins.

I mention this as there were three clear indentations in that sheet and it isn't hard to imaging that this could push through over time and result in a dead short on the supply?

I suggest if you have one of these instruments it might be worth checking if yours is the same!

Adrian

[Lifeborne]

Interested in a pic of the "slightly different logic card" uProc and IC. Thanks.

[fenland787]

here you go (if I've attached it right!).....

[srb1954]

I have the same CPU card as yours except my was built in 1991 vs 1989 for your unit. The CPU is the same part number and date as yours so I would assume Data Precision bought a huge batch of mask-programmed chips in 1987 and were still using them 4 years later.

The positioning of the regulator still hadn't been fixed in the 1991 version so it still has the potential for shorting a regulator pin to the chassis. However, this is only for the centre pin, which is GND on a 7805 anyway,  so it doesn't seem to be a catastrophic risk albeit it is not great manufacturing practice.

I checked the heat-sink temperature on my unit and found that the temperature rise is about 40deg above ambient, which I consider not unreasonable for a linear regulator. In my case the regulator input voltage was 9.09V so with your unit the higher input voltage should give a (still acceptable) heat-sink temperature of about 50deg above ambient. If you are seeing more than 50deg rise on your heat-sink them you may want to consider checking if you have excessive current drain in some part of your circuit.