Topic: Repairing a damaged HP 6235A bench supply
Date:  2024 MAR 28

While working on a project that needed +5V for logic and +/-15V for analog rails, I decided to buy a smaller power supply than our Lambda LPT-7202-FM triple-rail bench supplies, so that I could take the project with me without so much bulk. I decided on a HP 6235A as Connor Krukosky had one in his shop and I liked the size and features. Plus, it’s a HP supply, hard to go wrong there!

I ordered a “working” supply from a popular Internet auction site, and received a mess. The seller did a very poor job packing it, basically placing it in a just-barely-fit used single wall cardboard box with a handful of peanuts. Damage was immediately obvious from the busted knob:

The impact was hard enough that it damaged the potentiometer controlled by the knob, and it no longer turned properly. The seller agreed to a partial refund, to cover the cost of the knob and pot, but upon opening the power supply, I found more problems…

Corrosion Repairs

This supply clearly saw serious moisture. Many of the fasteners were heavily rusted, as seen above. The aluminum heatsink had a lot of white, powdery aluminum corrosion. Here’s a look at the top of the heatsink mounting bracket, where the power devices live:

I’d originally planned on at least recapping this PSU, since it’s old enough to need it by default, and replacing the broken pot, but the repair was clearly going to be more involved than that. The heatsink would have to come off for cleaning, and to ensure there was no oxidization packed up in the feed-through holes for the power devices. Heatsink compound would need replaced, too.

To remove the heatsink, all of the power devices have to be removed, power input wiring desoldered, and the linear transformer has to be unbolted. This was a pain, but necessary, and there turned out to be a bunch of corrosion on the bottom of the heatsink and against the PCB:

The heatsink was cleaned up with a wire brush, as were the power devices. All fasteners were replaced to ensure good contact – the case is one of the terminals on a TO-3, so this is important. New thermal compound was applied.

Potentiometer Repair

The impact that broke the control knob on the front of the power supply also damaged the potentiometer. It looks to have smashed the shaft into the back of the unit:

This damage caused the pot to bind, and also caused intermittent contact with the resistance element. Fortunately, this impact did not damage the wiper or resistance element in the pot. I was able to undo the crimped closures on the shell, clean and lubricate the wiper and resistance element, and reassemble the pot. Here’s a picture of the undamaged resistance element:

Obviously a very high quality pot – you can see the laser trimming marks around the inner circumference of the resistance element! The pot functioned perfectly after cleaning, lubrication, and reassembly.

Power Switch Replacement

The push on/push off power switch in this unit was apparently damaged: upon heating one of the terminals to remove the power inlet wiring (necessary to remove the heatsink), the terminal moved freely and came out of the switch! Fortunately, this is a pretty common Schadow unit, and I was able to find a replacement. The replacement used a different color mechanical indicator, but I was able to disassemble and swap the indicator from the original. Here’s a picture of the original switch, showing the damaged terminal:

It’s hard to tell, but the terminal in the foreground has moved up slightly from its correct position, and would only make intermittent contact with the switch pressed. So much for tested, eh?

Capacitor Replacement

The HP 6235A is old enough that it’s pretty much guaranteed to need at least new bulk filtering capacitors. Since I had it apart anyway, it got completely recapped. HP used high quality components throughout their equipment, but capacitors do have a limited life, and the construction techniques of the larger radial electrolytics of the time pretty much ensures they’ll be leaking or dried out. Visually, they looked OK, but on pulling them, they were found to be bulged and ready to start leaking:

The rightmost capacitor was already leaking around its main terminal. Two of the bulk filtering capacitors were dual section units, but there was enough space on the PCB to use modern capacitors without much fuss:

Remember to link the pads used for the capacitor case, if required, when replacing large vintage capacitors like these! Several positions used the can as a link, and without something joining the pads, the PSU would not have functioned correctly.

Final Repairs and Wrap-Up

This little power supply ended up needing around $50 in parts to fully repair it. A large part of that was the broken knob, which cost $22 from a surplus equipment dealer! These knobs are mid-sized, and none of the scrap HP equipment in our inventory matched.

In addition to the heavily corroded fasteners, I ended up replacing pretty much every fastener in the PSU with new 18-8 stainless steel hardware. Everything had at least a light layer of surface corrosion. Fortunately, the binding post on the heatsink was not heavily corroded, and cleaned up with a brass brush.

Once everything was buttoned up, the HP 6235A was tested with incandescent lamp dummy loads. It functioned perfectly, the panel meter didn’t even require recalibration! Quite a bit of work for an old DC power supply, but it’s quality equipment that will last a long time.

corroded PSUs rebuilt