Hi all!
I want to build a capacitor leakage tester/ checker instead of buying something expensive.
I was thinking of using my variac, followed by a voltage doubler to reach about 700V DC.
I'm not sure how to do the metering. I could either use a 50uA moving meter in series with a 10k resistor followed by the cap under test, or I could measure the voltage across the cap instead of the uA meter keeping the series 10K resistor.
I'm open for ideas and if any one has a better layout, please share it with me.
Thank you!
I want to build a capacitor leakage tester/ checker instead of buying something expensive.
I was thinking of using my variac, followed by a voltage doubler to reach about 700V DC.
I'm not sure how to do the metering. I could either use a 50uA moving meter in series with a 10k resistor followed by the cap under test, or I could measure the voltage across the cap instead of the uA meter keeping the series 10K resistor.
I'm open for ideas and if any one has a better layout, please share it with me.
Thank you!
1) do not use a variac with direct connection to the line. Its just too dangerous.
2) Use a transformer and a rectifier with a small cap to keep the energy storage low.
3) You will still have a lot of energy stored in the cap under test. make a charge and discharge circuit that will safely charge and discharge the cap. Most of these caps with "self charge" after discharge and can be quite dangerous to lethal.
This would not be a casual sport. Make sure you have read and understood all you can about high voltage caps. You don't get second chances if you make a real mistake.
A series meter will probably get fried pretty quick. Sprague's version used a magic eye tube. You will be looking at microamps or less with decent caps. Crummy caps will probably have high esr as well. Here is more Tube-era capacitor analyzers - Sprague Tel-Ohmike
2) Use a transformer and a rectifier with a small cap to keep the energy storage low.
3) You will still have a lot of energy stored in the cap under test. make a charge and discharge circuit that will safely charge and discharge the cap. Most of these caps with "self charge" after discharge and can be quite dangerous to lethal.
This would not be a casual sport. Make sure you have read and understood all you can about high voltage caps. You don't get second chances if you make a real mistake.
A series meter will probably get fried pretty quick. Sprague's version used a magic eye tube. You will be looking at microamps or less with decent caps. Crummy caps will probably have high esr as well. Here is more Tube-era capacitor analyzers - Sprague Tel-Ohmike
Hi DIY Didi
Did you ever build the leakage tester? I'm interested as I would like to build one myself.
Peter B
Did you ever build the leakage tester? I'm interested as I would like to build one myself.
Peter B
As a preamble, I endorse Demian's cautions above.
I occasionally measure leakage using nothing more than a suitable DC source and a DVM. You put the cap under test in series with the DVM and connect to the source, with the meter set to measure volts. Leakage current is simply the measured voltage times the meter's input resistance, typically 11 Meg Ohm. e.g. if you observe 1.0mV, leakage is ~91 pA. You can always shunt the meter with a resistor (say 100 K Ohm) if the above proves too sensitive for electrolytic, tantalum, etc.
I occasionally measure leakage using nothing more than a suitable DC source and a DVM. You put the cap under test in series with the DVM and connect to the source, with the meter set to measure volts. Leakage current is simply the measured voltage times the meter's input resistance, typically 11 Meg Ohm. e.g. if you observe 1.0mV, leakage is ~91 pA. You can always shunt the meter with a resistor (say 100 K Ohm) if the above proves too sensitive for electrolytic, tantalum, etc.
I very, very rarely find the need to test a capacitor for leakage. If it's suspect, just replace it. Better not chance something valuable for something cheap (like a cap). That said, I understand that sometimes you need to choose a cap with near-zero leakage, or perhaps it's a very costly capacitor like a huge mica or a pulse-discharge cap. Or maybe you have an amp with about 100 identical caps that may or may not be a poly-film type, and you want to test this to see if they all need to be replaced or not.
The way you're going to have to do this is to charge the cap in series with a huge resistor, maybe 2M (Some of this depends on the cap type- if this is a 0.5uF coupling cap, 2M gives an RC time constant of 1), then measure the voltage drop across it using something with a high input impedance. Your typical DMM will probably work, you'll have to try it. Otherwise you might have to come up with your own high impedance buffer. All of this will depend on the values of the caps.
Generating the HV is easy. You need almost no current. Transformer and a great big long voltage multiplier string. This gives you multiple voltage taps as well. Caps don't need to be big for this- a couple microfarads is fine.
If you want to measure tiny leakage values in things like mica caps or new poly caps, you're going to need something seriously sensitive. If you want to measure something in the 500pA leakage range, you're going to need something like a 50M resistor and perhaps some sort of tube-based buffer to take the voltage. Old cap testers used a magic eye tube for a display, but you could use an analog meter too (with suitable driver circuits) if you wanted.
Could be a very fun project to design.
The way you're going to have to do this is to charge the cap in series with a huge resistor, maybe 2M (Some of this depends on the cap type- if this is a 0.5uF coupling cap, 2M gives an RC time constant of 1), then measure the voltage drop across it using something with a high input impedance. Your typical DMM will probably work, you'll have to try it. Otherwise you might have to come up with your own high impedance buffer. All of this will depend on the values of the caps.
Generating the HV is easy. You need almost no current. Transformer and a great big long voltage multiplier string. This gives you multiple voltage taps as well. Caps don't need to be big for this- a couple microfarads is fine.
If you want to measure tiny leakage values in things like mica caps or new poly caps, you're going to need something seriously sensitive. If you want to measure something in the 500pA leakage range, you're going to need something like a 50M resistor and perhaps some sort of tube-based buffer to take the voltage. Old cap testers used a magic eye tube for a display, but you could use an analog meter too (with suitable driver circuits) if you wanted.
Could be a very fun project to design.
Last edited:
Here's one plan you could try:
Electrolytic Capacitor Reformer & Tester - August 2010 - Silicon Chip Online
Electrolytic Capacitor Reformer & Tester - August 2010 - Silicon Chip Online