Well, Here I am again with another opportunity to demonstrate to the world that I don't know what I'm doing. Here goes...

I'm working on the restoration of a BC-453 and I've run into an issue with the capacitors. I'm trying - as an exercise in futility - to approach the reconstruction as a field radioman might. I'm looking at Table 15 on page 75 of the SCR-274-N 1943 Manual which contains the Capacitor Test Data. Now the way I understand it, this table contains the capacitances expected between each pole of the capacitor and ground WITHOUT disconnecting it. The capacitor is thus a part of a circuit and won't (shouldn't) show the rated capacity, only the apparent capacitance based on its location in the circuit.

For example, the Apparent Capacitance measured at C5 should be 2.0 mF. C6-b should be 2.3mF etc.

However, when I try to test any capacitance (with my Fluke 15b+,) I get something wildly different. For example, instead of an apparent capacitance of 2.0 mF, I might get .52pF?!

Of course after seeing this for the first time, I just KNEW the capacitor was bad. I de-solder it, test it and...WHAT?!...It's within 5% of rated?! OK, no problem, just a random fluke. I move. on...and on and on. Every capacitor that tested BAD in place (based on the Apparent Capacitance Table), was actually fine when desoldered and tested! None of the apparent capacitances in the manual seemed to jibe with reality! Maybe a bad component on either side of the capacitor? Nope...resistors good, coils good... The Apparent Capacitances in Table 15 of the manual seem to be wrong!

That's right! WWII was won even though all the data given radiomen was wrong! How could they have been so blind?!

Or maybe...just maybe...I don't know what I'm doing?! Hey! That's a real option!

I'd love to hear from some of you old salts in the field. I feel like such a dolt. What am I missing? What part of my (surely flawed) understanding is wrong?

Thanks in advance for any direction you can nudge me in!

Best Wishes,

Cliff

Cliff,

The Fluke 15b did not exist during WWII.

Use the right meter.

I worked on a lot of ARC-5 gear back in the 1950s and early 1960s, and found no significant errors in the documentation.

- Leigh

Thanks Leigh!

Yeah, I know the data is right, just not sure where I'm wrong. Do I really need to find a 666 and 665 test unit to get the same results?

As a matter of course, I did drag out my vintage Heathkit capacity test unit and got substantially the same results. I'm just stumped.

Thanks again Leigh!

You don't necessarily need to use the exact specified test equipment, but you DO need to use test equipment that duplicates its characteristics.

That's particularly important when doing in-circuit testing.

- Leigh

Hi Leigh

I've found that testing capacitors in circuit is not nearly as accurate as testing resistors in circuit.

Have you ever built a "Boris Karloff" antenna tuner? Naked caps and an inductor on wood or a metal sheet with no enclosure? Depending on how you build it, you'll see the SWR change as your hands come close to the tuning knobs. I suspect that's what you're experiencing here.

I've always unsoldered one end of a cap when I've measured them on my ARC-5 stuff and never ran into a problem.

When the documentation was written the equipment was almost new.
Now almost 70 years later, one has carbon resistors that may have drifted as well as the other capacitors that may have drifted, opened or shorted.

And then there's the real issue. Old capacitors rarely measure the wrong value. Where they fail is leakage. So even if the capacitance matched what the documentation says, the capacitors could very well still be bad. Today you have to test them with a leakage tester at their rated voltage.

The original documentation may have been written to help make fast emergency repairs on radios with bullet holes in them.

Thanks All for the great responses and informative help!

I think the leakage issue may be a source of my problems. The capacitance tester I have doesn't test for leakage.

Does anyone have a schematic, or method for testing capacitance leakage? I just don't have the funds right now to make that kind of equipment purchase.

Cheers,

Cliff

You apply rated voltage and measure the current flowing through the capacitor.

A convenient way to do this is with a Sprague TO-6 capacitor analyzer or similar.
Alternatively, you may use a regulated power supply with current limiting.

The applied voltage is important. With any impedance in the voltage source (i.e. resistor or regulator), the time required to fully charge a high-mfd capacitor may be significant. If you use a current-limited power supply, this time may be quite long.

- Leigh

If he tries the power supply method it might be helpful to know what amount of current is deemed a good or faulty cap. I'd post it myself but I don't know.

Unless you use a modern digital meter. They will measure leaky capacitors high by a decent amount. It has to do with the way the meters measure capacitance, with the leakage showing as more capacitance.

I don't know of ANY gear that gives "in circuit" values for capacitors. Too
many variables.

Roger

True but it depends on how leaky they are. I have seen many that are too leaky to use but still measure within tolerance. Testing as higher than the marked value indicates a bad capacitor but showing the correct value doesn't necessarily indicate a good one.

The easy way to test capacitors for leakage is to get a voltage, neon bulb, and resistor.
The resistor should limit the current to the maximum to keep from blowing the neon bulb.
Hook them in series to the capacitor.
The lamp should blink once as the cap charges.
If it keeps blinking, it is slightly leaky and if it stays lit there is a lot of leakage,
total cost = 1 neon bulb, one resistor, and power for the bench power supply.
73,
Pat