Heathkit IP-32 Restoration

I wanted a high voltage power supply for experimenting with tube circuits, so I decided to find an old one on Ebay to restore. I found a listing from the California Historical Radio Society for a Heathkit IP-32 power supply described as “repaired and tested”. More on that later, but first a bit of information about the supply:

• B+ output 0 to +400V at 100 mA
• C- output 0 to -100V at 1 mA
• Filament output 6.3V AC at 4A

The filament output is an isolated winding and the B+ / C- outputs are floating. This means they can either have one end tied to the chassis or they can be combined in series or parallel with another power supply to increase the available voltage or current.

Repaired and Tested

When I purchased the device on Ebay, I was expecting to do some electrical work on it, but I figured my job would be easier given that the device was “repaired and tested”. Having a working device would mean that I wouldn’t need to track down any wiring errors. Since this was originally a kit, it’s a roll of the dice whether or not whoever put it together took any care in what they were doing.

It arrived, and externally it looked very nice.

I immediately clipped off the line cord and opened it up. (Warning: tube devices can kill you. Don’t open up or plug in a device like this if you don’t understand the risks.)

This certainly doesn’t look like it was “repaired”! I spoke with Philip from CHRS about the erroneous listing, and he was very apologetic. Apparently they are implementing a new QC process to avoid similar issues in the future.

You can see in this photo all the original electrolytic capacitors as well as the original paper and foil capacitors. These will all be electrically leaky (i.e. developed parallel resistance), so trying to operate the device in this state would be very bad for the tubes.

As for “working”, you can see if you look closely that a solder joint is completely snapped off from one of the selenium diodes! The C- output would have absolutely zero chance of working in this condition.

As if that wasn’t bad enough, check out how close this lead from the filament transformer is to the chassis! Whoever was in this thing last did a “J-hook” to connect the line he cut, and didn’t bother to insulate the connection. Since this is an isolated 6.3V winding, it’s unlikely contact with the chassis would have caused a lethal shock, but it’s still concerning - seeing this made me decide to go over every connection in the device to make sure there weren’t any other safety issues.

To make a long story short, I ended up re-routing most of the AC lines to try and reduce hum, I replaced every capacitor with high quality modern components, and I replaced the ancient silicon diodes and broken selenium diodes with modern replacements.

The selenium diodes are pretty cool, I hadn’t seen something like this before. They look like capacitors on a signature tracer. I’m not sure what happened inside to turn them into capacitors.

Here’s the finial result of the chassis restoration, before I attached a new line cord. I’m very pleased with the results.

I had some minor issues with the power switch - the standby mode didn’t work properly. I was able to open it up with some pliers and clean / bend the contacts, and got it working again.

Here’s the obligatory tube shot (I replaced the metal 6L6 with some modern glass 6L6EH ones just for aesthetics - the metals ones still work fine).

The power supply works extremely well - I get only a 1V drop when placing a 100 mA load at 400V. The analog voltmeter is a little off - it reads 380V when the actual voltage is 400V - but I use a DMM anyways so this doesn’t bother me. The milliammeter works perfectly.

This was an extremely fun restoration, and I’m looking forward to using it while experimenting with tube circuits on the breadboard.

If you want to learn how to do this kind of stuff yourself, I can’t recommend Mr. Carlson’s Lab highly enough. He also has a video of his restoration of an IP-32 on YouTube.