Identifying specs of vintage caps - "Sprague Compulytic"

[ManuelMcLure]

I just got a new-to-me Ambitrol/Power Designs TW-4005 power supply. Internally and externally it looks in fantastic shape. The main filter caps do not show any physical signs of problems, but in case I need to replace them, I need to identify what replacement caps to use.

I'd start with the four main filter caps. There are two pairs. One pair is labelled

Sprague
Compulytic
CE 72-.25
1000-25DC
RED POS

and the other is

Sprague
Compulytic
CE 47-85
350-85DC
RED POS

The "RED POS" is obvious - the red terminal is the positive (the other terminal is white). I'm guessing that the 4th line is the capacitance in uF and the voltage - i.e. 1000uF at 25V and 350uF at 85V, but I would like to be sure. I have no idea what the "CE XX-XX" lines mean. Does anyone have experience with these caps and can confirm my assumptions?

If I'm right, it's astounding how much smaller equivalent modern electrolytics are . It would be very easy to fit several in the space of the old cans.

[JLNY]

Often times those old Sprague caps are quite well sealed, and a lot of them still perform fine after 30-40 years. It probably wouldn't be too difficult to check them with an ESR tester and see if they need replacing yet.

I think you are correct to assume that 1000-25DC would mean 1000uF, 25VDC.

maybe the CE is a case size or tolerance code?

According to the modern datasheet for compulytics (apparently they are still making caps under that name), a "CE" case code would mean the can size is 2.5"Dx5.125"H, but who knows if the modern ones have the same marking codes as these. Are the two values the same size physically?

Also, a bit of advice about those old TW power supplies: an old mentor of mine uses one very similar to that, if not the very same model, and it has a potentially problematic glitch: if the wiper on the voltage pot goes open circuit, the output pings to the maximum voltage and could damage whatever you are working on.

Keep a careful eye on the needle when adjusting the voltage to make sure. The pots in my mentor's supply were very scratchy, so he had to disconnect the power leads any time he wanted to adjust the voltage. Also, as with any power supply that uses panel meters, be particularly mindful that the needle might be out by 10% or more from the actual value, so always double check the readings with a DMM. I recall his being particularly inaccurate; even with the needles zeroed out properly one meter was always way low and the other was usually too high.

[Gyro]

Often times those old Sprague caps are quite well sealed, and a lot of them still perform fine after 30-40 years. It probably wouldn't be too difficult to check them with an ESR tester and see if they need replacing yet.

It's probably easier to check the voltage ripple across them than ESR testing. If that's respectable then leave them alone.

[ManuelMcLure]

Often times those old Sprague caps are quite well sealed, and a lot of them still perform fine after 30-40 years. It probably wouldn't be too difficult to check them with an ESR tester and see if they need replacing yet.

I think you are correct to assume that 1000-25DC would mean 1000uF, 25VDC.

maybe the CE is a case size or tolerance code?

According to the modern datasheet for compulytics (apparently they are still making caps under that name), a "CE" case code would mean the can size is 2.5"Dx5.125"H, but who knows if the modern ones have the same marking codes as these. Are the two values the same size physically?

They are the same size, but closer to 1.5"Dx2.5"H so either the CE is not a case code or the code has changed over the years.

Also, a bit of advice about those old TW power supplies: an old mentor of mine uses one very similar to that, if not the very same model, and it has a potentially problematic glitch: if the wiper on the voltage pot goes open circuit, the output pings to the maximum voltage and could damage whatever you are working on.

Keep a careful eye on the needle when adjusting the voltage to make sure. The pots in my mentor's supply were very scratchy, so he had to disconnect the power leads any time he wanted to adjust the voltage. Also, as with any power supply that uses panel meters, be particularly mindful that the needle might be out by 10% or more from the actual value, so always double check the readings with a DMM. I recall his being particularly inaccurate; even with the needles zeroed out properly one meter was always way low and the other was usually too high.

I bit the bullet and powered the thing up. The good news is that both channels seem to output pretty close to the values on the meters, and the pots, although not perfect, do not seem to have any major issues. There's a bit of jumping around so they'd probably respond well to a cleaning but I did not notice any jumps all the way to full voltage - just a couple of places where the voltage "jiggled" a bit as I turned the pot instead of increasing monotonically.

There is a diode on the "A" side that has a case that is quite discolored. The diode is in a metal case and (if it's the same as the corresponding one on the "B" side) is labelled P R200 M 65xx (unfortunately the last two digits on the diode on the B side are not easily visible - I'll need to desolder it to be able to read them. Other diodes on the board are labeled with 6545, 6510 and such, and I haven't been able to find any reference to these in case I need to replace any of them.

[JLNY]

I bit the bullet and powered the thing up. The good news is that both channels seem to output pretty close to the values on the meters, and the pots, although not perfect, do not seem to have any major issues. There's a bit of jumping around so they'd probably respond well to a cleaning but I did not notice any jumps all the way to full voltage - just a couple of places where the voltage "jiggled" a bit as I turned the pot instead of increasing monotonically.

There is a diode on the "A" side that has a case that is quite discolored. The diode is in a metal case and (if it's the same as the corresponding one on the "B" side) is labelled P R200 M 65xx (unfortunately the last two digits on the diode on the B side are not easily visible - I'll need to desolder it to be able to read them. Other diodes on the board are labeled with 6545, 6510 and such, and I haven't been able to find any reference to these in case I need to replace any of them.

Glad to hear that it seems to be working reasonably well. Definitely keep an eye on the voltage output if you are getting scratchiness on the pots. I don't know what the rise time is on these supplies, but even momentary voltage spikes too fast for the needle might be enough to do damage.

I couldn't find any info on those diodes. It might be a good idea to look into finding a manual for this supply, which would probably have the part numbers listed as well as schematics which can give you a better sense of what everything is doing.

Speaking of rise times, if you have a storage scope it might also be a good idea to check the output transient response for any voltage overshoot on turn-on or on recovery after having the output shorted. It's simple enough to check if you have the equipment for it, and knowing the quirks of your PSU ahead of time might save you from any nasty surprises when working on sensitive projects.

[ManuelMcLure]

For anyone following this thread, I picked up a copy of the manual (actually, the manual for the 4005/4005R, but the TW-4005 is just two of those in a single case).

There seem to be four types of diode in the supply:

PT 520 - silicon rectifier, 500mA, 200PIV
F1044P - silicon rectifier, .5A, 100PIV
PS 359A - 5.5V +/- .1V zener
IN645 - silicon rectifier, 400mA, 225PIV

It looks like 1n4004 would replace all of the rectifiers if needed.

[ManuelMcLure]

The really good news is that the discolored diode (CR15) seems to just be the reverse polarity protection diode on the output, so it can definitely just be replaced with a 1N4004. The question becomes whether I want to try to maintain the appearance by somehow inserting the 1N4004 into CR15's metal case.

[tigr]

To test old capacitors, a very important test is a loss test. This test allows you to understand how the capacitor is charged and discharged. It is verified by a transistor tester.

[IanMacdonald]

Unless there's an obvious malfunction I'd leave those caps in. They were very well made components. That said they often had no safety vent so if they do blow it can be with one hell of a bang. Safety glasses not a bad idea when working over it, at least until it's had an hour or two of ontime.

[bitseeker]

Often times those old Sprague caps are quite well sealed, and a lot of them still perform fine after 30-40 years. It probably wouldn't be too difficult to check them with an ESR tester and see if they need replacing yet.

It's probably easier to check the voltage ripple across them than ESR testing. If that's respectable then leave them alone.

Yep, that's usually sufficient verification for these. I have a variety of PD supplies (TP340A, TP343B, 6050A, 2010, 2015, 3650-S) and the Sprague caps in all of them are still OK. Modern caps are significantly smaller, but you can't beat the aesthetics of the old cans. If I ever have to replace them, I'll probably restuff them.