Print

[KC2RLY]

With the abundance of capacitors in the market mostly from USSR surplus, I always want to use them but shy away because I can't usually find reliable freq/current ratings online. Since their cost is so much lower than domestically available options, if they are useful, I wouldn't mind sacrificing a few to the point of destruction to truly discover their current capabilities.
If I wanted to test capacitors for CURRENT handling capabilities, could we take a 100-500v ac source, ideally current limited?, connect an ac ammeter in series with a capacitor and load resistor?
Or is that test worthless without a high frequency ac source?

[G3RZP]

Or is that test worthless without a high frequency ac source?

Basically, yes. Plus you need to measure the temperature of the capacitor to decide if it's getting too hot.

The way i have done it - and it's not too precise, but as they say, close enough for government work - needs an RF source  an amplifier and a dummy load. Then make  a tuned circuit partially loaded by the dummy load and feed and feed some power into it. Start at low power and measure the volts across the capacitor under test. Then you can gradually increase the power and knowing where you started from, you can calculate the volts and the current in the capacitor and check the temperature rise. Many ceramic capacitors, especially if bigger in capacity than about 300pF, have a poor temperature co-efficient. I prefer to limit the temperature rise to 50C but some people use 70C.

I wrote an article for Vintage and Military Amateur Radio Society magazine 'Signal' on transmitting capacitors. Send me your email - I am g8on@btinternet.com   and I'll send you a copy.

[W1VT]

A power attenuator can protect the rig if you have concerns about.
A temperature rise is easy enough to measure these days with an inexpensive IR cooking thermometer.

[K4SAV]

There is a very large difference between what a capacitor can handle in a circuit where it looks like a short circuit for RF and in a circuit where it is supplying reactive power.  That is a circuit where the capacitor is charging and discharging on every cycle, such as it will do when used in a tuned circuit, or maybe in a matching network for an antenna.

There are 3 specifications for a cap delivering reactive power, a voltage limit, a current limit, and a power limit (a KVA rating).  The KVA rating is a product of current in phase with the voltage on the cap. Current in the cap is a combination of current in phase with the voltage and also a current 90 degrees out of phase with the voltage.

I'm not sure how the manufacturers do this test.  Just guessing, you could set up a circuit, maybe something like a resonant LC network with taps on the inductor to feed  some kind of dummy load and measure the heat dissipated in the cap.   I don't know what limit you should apply for heat rise.  Measuring the reactive current will be tricky.  It won't be the same as the driving current from the generator.  The 90 degree out of phase current is also not the current that causes the heat in the cap.  You need the in phase current.  I'm not sure how to measure that.  Maybe you could do that with an oscilloscope.  You will also have to make measurements at different frequencies.  The allowable current limit will be a lot less at lower frequencies. 

If you can measure the temp rise, set a limit for that, and measure the in phase cap current, then calculate max voltage times in phase current, you could get your own limit for KVA.

There will also be a voltage limit for the cap.  You probably won't be able to test for that without destroying the cap.  Voltage on the cap in a circuit where it is supplying reactive power won't be the same as the driving source.  That can be measured, but be careful.  You will be playing with very high voltages.

If you just set up a circuit and only measure to see if the cap gets hot, you won't know what the cap will do in a different circuit without having the numbers for limts.

Read this pdf to learn more about how caps work:
https://www.highenergycorp.com/High%20Energy%20Corp.%20Ceramic%20Capacitors.pdf

Jerry, K4SAV

[G4AON]

I have used surplus 10KV rated doorknob caps in traps and not had any issues with them. The caps in question are 22pF to 47pF and are occasionally listed on eBay from Eastern Europe. Make sure they include the screws.

I don’t know what values and power levels are needed, but in parallel traps at typical power levels of our 400 Watt limit and some EU countries running 1 KW, they are fine. Not sure about 1.5 KW.

There are a pair of 27pF, 10KV, 8KVA rated caps on eBay at the moment, item number 284112567377

73 Dave

[G3RZP]

The biggest problems can arise where  capacitors of greater than about 250pF are needed, such as padding capacitors for pi and pi-L tank circuits on 80 and 160m, because such capacitors either need to be greatly under run so they don't get hot and thus have insignificant drift or used in limited duty cycle applications - so not FT8 or SSTV or RTTY. I saw a 50 kW LF tx that used something like 100  capacitors of 150 pF in parallel to tune the PA tank to about 20 kHz: being of that value, they could all be nominally COG types.