Zener Diode aging in a Kepco MPS 620M?

[MauriceS]

This is a power supply from about 1977. Solidly made, but Kepco still wants 200 USD for a service manual... I found a 90% complete (minus schematics, but with theory of operation to component level) manual online, but I am wondering what I find...

The issue is that the 6V / 5A side doesn't work when the crowbar circuit is enabled (i was able to disable it by removing one OpAmp...)

I suspect a production date of 1980, as some components have yyww where yy=80. Except for the opamps (some plain 741's have date codes (vertical) like 027 and 031 etc.

Now what really confuses me is that the zener diodes are HORRIBLY out of whack. Of all the 12V 5% zeners I found I find voltages in the range of 9.7/9.8V, (w/ 5% tolerance I'd expect 11.4V) a 6.2V zener 5.3V, a 6.8 zener 4.8V.

Is this expect able after 38 years? Is it worth fixing this beast?

I don't have a zener tester (yet) so i don't know if these diodes are so bad out of whack...

Thanks a bunch...

Maurice

[Seekonk]

I've been working on an old HP of that vintage off and on for a while.   I've found two zeners quite out spec so far. I wouldn't have expected that. All you need to test them is a resistor and a supply.  In circuit I've seen some pretty convoluted methods to bias them so in circuit may not tell you much.

[KubaSO]

I've reverse-engineered the PCB assemblies on that supply, and thus have made the schematics public. I have no access to the official schematics other than the simplified ones available online, and the replacement parts list.

Thus, the KEPCO MPS 620M Schematics:

https://github.com/KubaO/T-M-Reverse-Engineering

Treats especiales:

• it's drawn in Elektor style in KiCad
• my hand-drawn draft is included for audience's amusement

And yes, in mine all Zeners were way outside of the tolerance. One small signal transistor had two open pins (per capacitance meter).

Is it worth fixing this beast?

IMHO: Yes. This supply has a very conservative thermal design. You can run it hard all day and it won't even notice. It's also quite moddable - even with existing pass transistors, it can be made to respond faster by using better op-amps and changing the compensation a bit. For 6V A1 supply, it helps to use better transistors in the pre-driver to ensure it has enough bandwidth to make faster op-amps worthwhile. All active parts are socketed.

I've been running it into a dead short for a day on both supplies at the same time - nice room heater, it took it without a problem.

[coromonadalix]

Seems a good design with switched xformer secondaries

i would try to repair it

[Seekonk]

Interesting, I worked on an old HP supply and a couple of zeners were also way off in voltage compared to what the factory manual said. I thought it very unusual.

[KubaSO]

This one doesn’t have switched secondaries but a huge enclosure that’s mostly holes. Has no problems shedding heat with nothing but natural convection. As long as you keep the dust bunnies out

I’ve added an SCR pre-regulator to the 20V supply and no problem running it at 2A