Capacitor leakage tester design & build - Part 4 

It's been a long time since I've gone through electronics school, and this is really a great refresher! Thank you!

Definitely not too much theory. Essential not only to understand the design process and operation of the device, esp when not understanding can destroy the device itself !

You don't need two voltmeter, you can add a third section switch, in discharge position the voltmeter is connected directly to the cap in the others positions is connected before the ammeter. Thank you for your job and time spent on. Tom

Thank you for your video. I am relearning electronics after several years away from the research laboratory. The safety concerns and explanations are appreciated.

I am building this lovely tester. In making the shunts for the meter, for the 100mA shunt, I use a 50 ohm mini multiturn pot in parallel with a 2.4 ohm resistor, likewise, for the 10mA shunt, I use 100 ohm multiturn in parallel with a 39 ohm, 1mA shunt use 1000 ohm multiturn in parallel with a 470 ohm. These values are based on the multiturn pots set at midpoint to get the shunt resistance of 2.16 ; 21.8 ; 240 ohms respectively.

This was an extremely interesting series and the theory was presented in a manner that is easily understood. Great job. I'd build one myself but I'm still relying on my old school IT-11 and TO-5.

Thank you so much for these carefully laid out, properly taught insights.

Many thanks for this video. I have just built the one from Max Robinson's website before I found your series. I am actually tempted to build a copy of your design as it reads directly and consequently means the testing process of capacitors is quicker per capacitor. A very clear and concise video that makes the whole device simple to understand. Thank you once again.

Great explanation of the Shunt circuit. I have a 0 to 400vdc Power Supply. So I will be using that to feed the second half of this device. The one problem is finding the "shorting" (make before break) switch. Several switches I found do not specify if they are "shorting" or "non-shorting."

Hi, nice project, I came across it because I'm thinking of something similar, but I want to go from 0 and end at 600V. To check the discharge of the capacitor, I intend to use a circuit from a separate discharger that I once made - a diode resistance divider, an LED is used as a control element - when it goes out, it is safe. A bipolar led was used in the discharger itself, so it doesn't matter which polarity I choose when discharging electrolytes (+ on the red terminal - it lights up red, + on the black terminal it lights up green), in this case an ordinary LED would be enough because it is necessary to keep the polarity. Nice day 🙂 Tom

Great project

Great series!

I've been enjoying this presentation. I had in fact forgotten I had partly watched this in the past. 6:15 : A tiny improvement you may want to consider, if you have more than one spare positions on the rotary switch, is to add additional progressively lower value resistors on the additional unused positions adjacent to the starting position. It's not fail-safe as one could still have the switch at one of the measurement positions before plugging the unit into the mains.

You could have placed a diode across the 100 uA meter to protect it, so the meter would not burn out if you put it into a too low range by accident. In normal operation the diode would not conduct as the voltage drop across the meter is lower than the threshold voltage of the diode.

You are really good at design. And you have to get the design down on paper before you build otherwise the design becomes a problem.

Good work 👍

Hi, I am completely self-taught and continuing to learn every day. I got interested in your leakage tester for a couple of reasons: 1) I have some nice tube equipment that was in great condition when I stopped using it 10+ years ago, but it's been sitting in its packing crates in my office ever since, and I'd like to be able to test the big caps (12,000uF, 100V); and 2) I am finally getting around to doing some DIY stuff that I started thinking about 15+ years ago, and I have a box full of 15+ years old Black Gate caps that have never been used. I was thinking your tester could also be used to test and possibly "re-form" any that are salvageable. So, I did my best to build your tester, only I used a 12VDC power supply and a boost converter to get an adjustable 25-390VDC. (I already had the power supply and the boost board was cheap.) Just this evening, I think I got the tester working. I have some NOS Panasonic caps from 2002 and tried one of those first. I couldn't get it to settle down once the voltage got to about 300V (330uF, 400V cap). The leakage current seemed to shoot up around that 300V point. I assume the cap is bad and beyond re-forming. I then hooked up one of the 12,000uF, 100V caps. When I turned on the current from the discharge to 100mA position, the voltage dropped to nearly zero, and after switching ranges a little, I figured out that, as the leakage current ticked down, the voltage ticked up. I kept messing with the cap until I got the voltage to around 90V and the leakage current stopped settling--in fact started getting a little jumpy--around 20-30uA. Does the tester behavior I describe make sense? For example, would you expect a very large capacitance cap to drop the voltage significantly while its leakage current settled back. I don't have enough confidence at this point to trust what I'm seeing on the meters rather than wonder if I've wired something incorrectly. FYI, I measured the voltage across the cap with my multi-meter and it was showing the same thing as the tester--a very sow tick upward as the leakage current very slowly ticked downward. Thank you in advance for any help. Stay well. Louis

The 100 ohm inrush current limiter behind the bridge rectifier should be around 2 watts.

For the scale adjustment (100µA - 100mA) I understand the usage of paralleling a resistor with a trimpot. I assume that you set each of the ranges exactly as you calculated? Or did you set the pots with the probe wire included? I tried looking through the series but this wasn't mentioned.

Thanks for your response on my earlier question. Before I do a smoke test on what I have created, another couple of questions popped into my head. Do you have a part number for the 12v zener diode that you used or does this not matter (will any 12v zener do the trick)? Is it possible to get this exact ampmeter to read 10μA or does involve more complex circuitry to increase the signal strength to read a lower value?

Good evening, I really liked your video and I was interested in putting this device together, I would like to know which is the secondary current of the total transformer?

Great stuff! On the selection of the mosfet, what sort of power dissipation would we be looking at given 30mA max current, I notice the mosfet you went for is rated at 370W which is very high... I'm planning on building one with a step-up transformer to get up to 600V working DC, so am looking at a STW34N65M5 rated at 710V D-S, max dissipation 190W. I'm not sure what to do about the adjustment pot situation though, as 600V across feels a bit dodgy. As for not being able to use the voltmeter across the cap as itself draws current, I presume there's not enough zeroing adjustment on the uA meter to compensate or any other way around that?

Thanks for this series of video’s, your are great at explaining the working and design process for the circuit! You have motivated me to build this, once I have got my head around the only point in the schematic that I am not completely comfortable with, that is the max voltage on the control pot. I did not have a suitable transformer available, so I sourced one at a reasonable price that is 220V to 150V-0-150V. So the output voltage range would be a bit higher, and I can resize components accordingly, but also the max voltage on the pot would rise even further over the component voltage rating. I am curious, have you seen any isolation issues with a standard pot? Finally, if I may suggest a simple addition, as you point out at 15:14 in the video during discharge the voltmeter does not show the voltage on the capacitor terminals, so there is no simple way to know when the capacitor is discharged. Instead of adding a second voltmeter, one could add a few diodes and a led to provide a visual clue that there is tension on the capacitor and to indicate when it is safe to touch it. I have recently described in an instructable (called “ Safe Capacitor Discharge Tool ) the build of a simple safe capacitor discharger, and it would be very straightforward to add a diode string, a resistor, and a led between the 47k discharge resistor and ground. One could even keep the two strings/two led circuit as described in the instructable for the case that an electrolytic charged capacitor is connected to the leakage tester with inverted polarity, and use a different color led to warn of the mistake.

Nice circuit I may just build one! However, the power LED arrangement you're using is not good. An LED should never be reverse biased as it will be in your case since it's driven from an AC source. They can be damaged by that or have their lifespan significantly reduced... Add a series diode like a 1N4148 and you'll be OK. Do you find that the test voltage adjustment is non-linear because of the non-linear relationship between Vgs and Rds? I'm sure there is some, but do you find it any problem?

Would a 120->240ac travel voltage converter be an acceptable source of voltage for this, in place of a transformer?

Hi there Why did you change the MOSFET to the IRF540n Thanks for a good tutorial Ron

Hi there...my congratus about all the videos, I liked it a lot once its very well explained. About this Capacitor Leakage tester is there a Part 5? I wonder if you have the final schematic avaliable. thanks a lot!

I might actually build this one thanks to you. I think you actually need to have high voltage to test a capacitor properly. Ignition condensers are a prime example, but they need even more invasive testing due to the harsh conditions they live in. They need to be thumped hard to prove their worth. For your audio work, I was wondering, without any real experience doing so (so big grain of salt) if putting a poly fuse in series with the µA meter, IE right next to the meter on the upstream side of the meter could be used to limit the current and save the meter if the switch is on the wrong scale. A very large resistor could be placed in parallel with the polyfuse to soften the response (edit: and perhaps make up for the lower voltage rating of poly fuses). I have never used a poly fuse, but off the top of my head it seems like a possible solution. Thanks very much for your well thought out and practical solution rather than buying an 'iffy' and expensive leakage tester. Fred Winterburn

Still room for lots and lots of improvements that can improve performance, accuracy, protection, and all with less parts.

One question to this wonderful video: with the 100Kohm under the 1 Mohm pot you get a minimal test-voltage of about 30 Volts. How do you test elcos with 12 or 24 Volts, which are available?

What are y'alls thoughts on taking a 120vac input 24vac output stepdown transformer and using it in reverse to take 120vac and stepping it up to 600vac. This would give me 845vac after my rectification. Right???

I think I'm going to build the HV voltage source from freely available (battery powered) 5V->1200V converters out there.. that'll limit the current somewhat and be cheaper to build without the transformer (also using part of a circuit on edn.com to regulate 0-500V in a different way, optoisolated and avoid the HV pot issue)... I'm wondering though, what sort of power rating the shunt resistor and trimmer pots need to be in the situation where there could be a capacitor draw max current (go short) under test?

I just built this circuit (as described in my other pots). It seems like the discharge cap needs to be a lower value. It takes a long time for the caps to discharge. Instead of 47K, maybe a 4.7K?

Hi, can I leave out the 100k resistor under the potentiometer? Doesn't fet cause problems? It would be nice if I could set the voltage from zero, I have an irfp460 which is good up to 500 volts. Thanks!

At time 8:58 you state that you have calculated the voltage drop across the meter at FSD to be 0.216 V but do not state how. To calculate it you would need to know the sensitivity of the meter. Did you have the specifications for the meter or did you measure the volt drop at 100 micro A? BTW I am also refurbihing some boat anchors and intend to build the capacitor leak tester.

I was just on eBay and the good heathkit testers are $200 and up now.

I think it would be safer if you put a schottky diode across the meter. If you are monitoring leakage on a low current settingand the cap decides to partly short you will loose your meter as the circuit has no protection below 30 mA. I have known selector switches to have bad or open connections and again you will loose your meter.

hello, firstly this is a great series of video's and has inspired me to build one. i've just built my first guitar valve amp and would like to make one for testing valve amp caps. i have a power transformer which i have bridge rectified and have a B+ of 450VDC,if i use a 100UA panel meter for the leakage would i have to recalculate the value of the shunt resistors?

The problem with the pot on 340 Volt could be solved with one flip switch with double contacts. I would not keep that out, because testing on low Voltage caps is just as important as the rest. A pot of 500K between two 500K resistors, and a double switch that shorts or the top- or the bottom resistor. Adds 'some' safety too when only testing on lower Voltages.

A clearer image of the complete schematic would help. Anyway thanks. My screen shot is not clear enough.

I think if you accedently short the testleads of the capacitor tester you will blow up your ampmeter

Hi, I am completely self-taught and continuing to learn every day. I got interested in your leakage tester for a couple of reasons: 1) I have some nice tube equipment that was in great condition when I stopped using it 10+ years ago, but it's been sitting in its packing crates in my office ever since, and I'd like to be able to test the big caps (12,000uF, 100V); and 2) I am finally getting around to doing some DIY stuff that I started thinking about 15+ years ago, and I have a box full of 15+ years old Black Gate caps that have never been used. I was thinking your tester could also be used to test and possibly "re-form" any that are salvageable. So, I did my best to build your tester, only I used a 12VDC power supply and a boost converter to get an adjustable 25-390VDC. (I already had the power supply and the boost board was cheap.) Just this evening, I think I got the tester working. I have some NOS Panasonic caps from 2002 and tried one of those first. I couldn't get it to settle down once the voltage got to about 300V (330uF, 400V cap). The leakage current seemed to shoot up around that 300V point. I assume the cap is bad and beyond re-forming. I then hooked up one of the 12,000uF, 100V caps. When I turned on the current from the discharge to 100mA position, the voltage dropped to nearly zero, and after switching ranges a little, I figured out that, as the leakage current ticked down, the voltage ticked up. I kept messing with the cap until I got the voltage to around 90V and the leakage current stopped settling--in fact started getting a little jumpy--around 20-30uA. Does the tester behavior I describe make sense? For example, would you expect a very large capacitance cap to drop the voltage significantly while its leakage current settled back. I don't have enough confidence at this point to trust what I'm seeing on the meters rather than wonder if I've wired something incorrectly. FYI, I measured the voltage across the cap with my multi-meter and it was showing the same thing as the tester--a very sow tick upward as the leakage current very slowly ticked downward. Thank you in advance for any help. Stay well. Louis