Recapping Hickok 288x Signal Generator - Leap Day Post

Since today is leap day (29 Feb), I want to make sure I post something for posterity since this day only happens once every 4 years. Here we have it.

I plugged the Hickok into my variac and brought it up gradually to 120 volts.  What happened was it lit up, nothing smoked, and it was drawing under half an amp of current, so there were no shorts.

I also had it connected up to one of the frequency counters and got absolutely nothing.  So I'm going to recap it and see where it goes.

I wanted to at least get an idea of where it was at, e.g. did it actually work or not.

One thing I discovered almost right away was that the output pot is frozen.  The knob was turning on the shaft, but the pot itself wasn't doing anything.  So that could be an issue, you think?

Here we have the old pot off the generator - it's a garden variety 10K linear.  Fortunately I have a bunch of these on hand for pedal and geetar amp buildin' so I can just grab one and solder it in.

Isn't having a stash of parts a good thing?  You bet it is!  I hate waiting for parts to come in the mail so I tend to order a few extras now and then.

One other thing I discovered is that a previous owner connected a dial lamp to the pilot light connections and then stuck the lamp behind the frequency readout window.

Now, this is not a bad idea, and it was such a clean job I thought it was done at the factory in Cleveland.  But the plastic insulation gave it away - the rest of the generator is wired with cloth insulation.

Here's the business end of the lamp wiring.  As I say, not a bad idea in theory, but in practice, it's not so good.

The frequency dial is black plastic, and the light really can't shine through it at all.

So rather than leave it, I took it out.

But you do see whoever did it did a tidy job.

Now I replaced all the paper caps and the electrolytics.  Here are the filter caps.  I left the original can cap in place, but disconnected.  I put a vintage terminal strip on the chassis and used that as tie points for the new filters.

One of the new caps I also put it is visible in this picture - it's the yellow one.

Hickok was good to put a bunch of the paper caps in a 'bank' on one side of the chassis.  This made it very easy to replace them with modern caps.

It looks a little funny because the modern caps are so much smaller than the old wax/paper caps.

To give you an idea of the size difference, here's an old cap .025 uF rated at 400 volts compared to a modern one of the same value rated for 630 volts.

The old caps were dielectric paper rolled around foil.  Then the whole thing was dipped in wax to try and protect it from absorbing moisture from the air.  They still absorb moisture and go bad over time - the service life was probably less than 15 years.  In our case, these are 60 years old and most definitely bad.

You'll hear these referred to as "wax" or "paper" caps.  I think the latter is more accurate as it refers to the actual construction material.

Modern ones are sealed plastic and should have a longer service life.  They also tend to be made to a higher electronic tolerance than the old ones.  But they don't look anywhere near as cool.

It is possible to take the stuffing out of the old caps and insert a modern cap in them so it looks like an old cap - this is known as 'restuffing.'  I've done this on special, rarer radios, but I'm not going to bother on this 288x - it's a pretty common piece of gear.  I will save the old caps and keep them with it anyway.

As to the date of the signal generator, the date code on the output transformer gives us a clue.  The code is "352732."   We can use the list on Triode Electronics' site to decode it.

The "352" is the manufacturer's EIA number; in this case it's Essex (Transformer Division).  The "7" is the year - most likely 1947.  And the "32" is the week of the year the part was produced.  So the transformer dates from August 4 to 8th, 1947.  So we can surmise given shipping time and perhaps time in Hickok factory stock that our 288x was built in late 1947 or early 1948.

Isn't being an electronics history detective fun?  Now let's fire this puppy up and see if it works.