I recently found that if the parts I will be mentioning are more readily available than the money to buy a good capacitor tester, a home made one can be made with a multi-meter capable of reading up to 250 volts with at least a 20,000 ohm per volt rating, two 100K ohm resistors, an octal tube socket, a 117Z6 tube or solid state rectifier (which would make the octal socket unnecessary), a wire with alligator clip, a line cord, a couple of .05 mfd capacitors at 400 volts, and a small wooden board. The tube is wired as a standard AC-DC half wave rectifier. The output of this power supply is sent through the 100k resistors. There should be one for the positive connection, and one for the negative connection, so that all current must flow through these resistors, and there will be no direct contact made with the power supply. If any shorts are made after these resistors, they will protect the power supply. Next, take the two .05 mfd capacitors and connect one across the resistors so that one lead is connected directly to the cathodes (+) of the 117Z6 tube, and the other lead is connected directly to the return line (-). The other .05 capacitor is connected across positive and negative after the resistors. These capacitors will form a mild filter, and help keep voltages high under normal circumstances. When a short is developed across this supply, though, the voltage will quite easily collapse. The positive lead of the multi-meter is then connected to the 100k resistor which is connected to the cathodes of the 117Z6 tube (+). The wire with alligator clip is soldered to the negative 100k resistor, with the alligator clip end free for tests. When this lead is touched to the negative lead of your meter, the meter should swing up to about 100 volts or better. If it swings backwards, reverse the meter connections. I'm sure I have the right connection procedure, but I could be wrong. At any rate, any capacitors of a 150 volt rating or higher you would like to test for shorts or leaks are connected between these leads (the negative power supply lead and the negative meter lead). This method of testing gives a much more magnified view of shorts than placing the capacitor across the meter as an ohm meter. Also, since a high voltage is being used, it is more likely that any arcing that may take place in the capacitor will show up. When testing a capacitor, the meter should swing up a little, and then swing back down to infinity or zero. Some condensors will cause the meter to swing up higher than others, depending on their value. Anything at about .001 mfd or higher should cause the meter to swing up and then back down. If this does not happen, then the capacitor is likely open. Smaller values may not cause the needle to move noticeably (perhaps not at all). Any value capacitor with a short will cause the needle to not drop all the way to zero. Some with an arcing short will cause the meter to jump around. Though some circuts may work with leaky capacitors, for perfect, like-new performance, the capacitor being tested should have absolutely no leakage. NONE AT ALL! ZERO!

I have an old automobile ignition capacitor tester from the 1930s made by American Motors, and it loads the capacitors with a nice 300 volts. The device will even give capacitance readings. However, it does not work, and at this time I have not been able to repair it. The idea sounded really good, though, so I came up with my own tester that will show shorts. It is much more likely that a capacitor will short out than change value to any great degree, so this should be a very good help. As always, though, since this device is AC-DC, you must insulate it and you from all sources of grounds. Don't go testing capacitors inside a radio, especially if it's plugged in. If they are still connected in the radio, this is likely to give false readings anyway (due to grid and plate bias resistors, etc.). Remove them and then test them. Don't go sitting on a concrete floor while grabbing the test leads. It is wise to touch only one lead at a time if the unit is plugged in, and wiser to either put a momentary test switch on the leads (at least one lead, or a spdt switch to turn on and off both leads at the same time....I think it's called a spdt, but I could be wrong....it should be two separate switches in one housing that operate simultaneously with one button), or unplug the unit while attaching the capacitor. If you are using a tube for a rectifier, I recommend building some sort of test switch instead, because you don't want to have to wait for the tube to warm up each time you test. If you must build a better unit, perhaps a transformer at hand and a 5Y3 or similar tube will do. You can achieve higher voltages that way, and also isolate the tester from the line. If you have the spare money to purchase a transformer (assuming you don't have an old one at hand), you're probably better off just buying a very good capacitor tester. Perhaps you still may want to build this tester, though, as most modern capacitor testers don't test with high voltage.

Thomas

P.S. An example of the performance of this tester: My DeWald BT-100 television (really an RCA 630TS in a DeWald cabinet) had jittery horizontal sync. It was all over the map. I tested all condensors with the ohm meter alone. I found some paper condensors that were obviously shorted. They usually are very obvious. Replaced them.....still same problem. Tested the mica condensors.....when do micas short?...never. Found nothing. Finally I got the idea to build the condensor tester listed above. Tested all condensors again including the little Micamold ones and found that the 82 mmfd condensor leading from the sync. amp. circuts to the horizontal discrimination and oscillation circuts had a jittering spurratic short that made the meter jump all over the place. Obviously it was an arcing short that only developed with high voltages. Replaced this condensor with a new one, and all troubles have ceased.

:I recently found that if the parts I will be mentioning are more readily available than the money to buy a good capacitor tester, a home made one can be made with a multi-meter capable of reading up to 250 volts with at least a 20,000 ohm per volt rating, two 100K ohm resistors, an octal tube socket, a 117Z6 tube or solid state rectifier (which would make the octal socket unnecessary), a wire with alligator clip, a line cord, a couple of .05 mfd capacitors at 400 volts, and a small wooden board. The tube is wired as a standard AC-DC half wave rectifier. The output of this power supply is sent through the 100k resistors. There should be one for the positive connection, and one for the negative connection, so that all current must flow through these resistors, and there will be no direct contact made with the power supply. If any shorts are made after these resistors, they will protect the power supply. Next, take the two .05 mfd capacitors and connect one across the resistors so that one lead is connected directly to the cathodes (+) of the 117Z6 tube, and the other lead is connected directly to the return line (-). The other .05 capacitor is connected across positive and negative after the resistors. These capacitors will form a mild filter, and help keep voltages high under normal circumstances. When a short is developed across this supply, though, the voltage will quite easily collapse. The positive lead of the multi-meter is then connected to the 100k resistor which is connected to the cathodes of the 117Z6 tube (+). The wire with alligator clip is soldered to the negative 100k resistor, with the alligator clip end free for tests. When this lead is touched to the negative lead of your meter, the meter should swing up to about 100 volts or better. If it swings backwards, reverse the meter connections. I'm sure I have the right connection procedure, but I could be wrong. At any rate, any capacitors of a 150 volt rating or higher you would like to test for shorts or leaks are connected between these leads (the negative power supply lead and the negative meter lead). This method of testing gives a much more magnified view of shorts than placing the capacitor across the meter as an ohm meter. Also, since a high voltage is being used, it is more likely that any arcing that may take place in the capacitor will show up. When testing a capacitor, the meter should swing up a little, and then swing back down to infinity or zero. Some condensors will cause the meter to swing up higher than others, depending on their value. Anything at about .001 mfd or higher should cause the meter to swing up and then back down. If this does not happen, then the capacitor is likely open. Smaller values may not cause the needle to move noticeably (perhaps not at all). Any value capacitor with a short will cause the needle to not drop all the way to zero. Some with an arcing short will cause the meter to jump around. Though some circuts may work with leaky capacitors, for perfect, like-new performance, the capacitor being tested should have absolutely no leakage. NONE AT ALL! ZERO!
:
: I have an old automobile ignition capacitor tester from the 1930s made by American Motors, and it loads the capacitors with a nice 300 volts. The device will even give capacitance readings. However, it does not work, and at this time I have not been able to repair it. The idea sounded really good, though, so I came up with my own tester that will show shorts. It is much more likely that a capacitor will short out than change value to any great degree, so this should be a very good help. As always, though, since this device is AC-DC, you must insulate it and you from all sources of grounds. Don't go testing capacitors inside a radio, especially if it's plugged in. If they are still connected in the radio, this is likely to give false readings anyway (due to grid and plate bias resistors, etc.). Remove them and then test them. Don't go sitting on a concrete floor while grabbing the test leads. It is wise to touch only one lead at a time if the unit is plugged in, and wiser to either put a momentary test switch on the leads (at least one lead, or a spdt switch to turn on and off both leads at the same time....I think it's called a spdt, but I could be wrong....it should be two separate switches in one housing that operate simultaneously with one button), or unplug the unit while attaching the capacitor. If you are using a tube for a rectifier, I recommend building some sort of test switch instead, because you don't want to have to wait for the tube to warm up each time you test. If you must build a better unit, perhaps a transformer at hand and a 5Y3 or similar tube will do. You can achieve higher voltages that way, and also isolate the tester from the line. If you have the spare money to purchase a transformer (assuming you don't have an old one at hand), you're probably better off just buying a very good capacitor tester. Perhaps you still may want to build this tester, though, as most modern capacitor testers don't test with high voltage.
:
:Thomas
:
:P.S. An example of the performance of this tester: My DeWald BT-100 television (really an RCA 630TS in a DeWald cabinet) had jittery horizontal sync. It was all over the map. I tested all condensors with the ohm meter alone. I found some paper condensors that were obviously shorted. They usually are very obvious. Replaced them.....still same problem. Tested the mica condensors.....when do micas short?...never. Found nothing. Finally I got the idea to build the condensor tester listed above. Tested all condensors again including the little Micamold ones and found that the 82 mmfd condensor leading from the sync. amp. circuts to the horizontal discrimination and oscillation circuts had a jittering spurratic short that made the meter jump all over the place. Obviously it was an arcing short that only developed with high voltages. Replaced this condensor with a new one, and all troubles have ceased.

I found your homemade cap tester very interesting. Have you tried it on electrolytics? I made a homebrew tester for electrolytics from a 4 watt 110V candelabra lamp (a night light works good as long as it has a clear bulb). I put a night light in series with a silicon diode (rated 1 amp/400V) and conected it to an electrolytic capacitor (again in series). With AC power applied, a good cap will cause the lamp to glow mometarily and then go out indicating a charged but not leaky cap. A shorted cap will cause the lamp to glow continuously. A 'dead' or open cap will not show any glow. Of course this test will apply about 100V half wave DC to the cap so it must be rated for 150V or more. The 4 watt lamp acts as a ballast and limits the current to about 50 ma. For lower voltage caps, a transformer and lower voltage lamp can be used. Be careful to discharge the caps once they are charged up by this method. Also be extremely careful handling 110 AC power. A push button switch or the switch on the night light can be used so only momentary power is applied. Would appreciate any comments. Nick

Well, a good electrolytic will have no more than several million ohms leakage. Though the night light will give you a fair idea of leakage, it will not show values this small. For a quick test, though, it sounds like a good idea.

The capacitor tester does work for electrolytics.

T.

:I found your homemade cap tester very interesting. Have you tried it on electrolytics? I made a homebrew tester for electrolytics from a 4 watt 110V candelabra lamp (a night light works good as long as it has a clear bulb). I put a night light in series with a silicon diode (rated 1 amp/400V) and conected it to an electrolytic capacitor (again in series). With AC power applied, a good cap will cause the lamp to glow mometarily and then go out indicating a charged but not leaky cap. A shorted cap will cause the lamp to glow continuously. A 'dead' or open cap will not show any glow. Of course this test will apply about 100V half wave DC to the cap so it must be rated for 150V or more. The 4 watt lamp acts as a ballast and limits the current to about 50 ma. For lower voltage caps, a transformer and lower voltage lamp can be used. Be careful to discharge the caps once they are charged up by this method. Also be extremely careful handling 110 AC power. A push button switch or the switch on the night light can be used so only momentary power is applied. Would appreciate any comments. Nick

The general idea of my condenser shorts tester was to test for leakage and arcing in small value condensers, which tend to be more problematic in the arcing category. Though it will work with electrolytics, electrolytics are usually satisfactorily tested on the conventional resistance checker section of your multi-meter. Leakage for them in the millions of ohms is somewhat acceptable, and even some brand new ones have this leakage. With the smaller values used in audio and RF sections, which are sensitive to very small leakage, it is much more critical. Also, since electrolytics are always wet when in good condition (the wet electrolytic, which is even wet in the so-called "dry" electrolytic, but in paste form), they are less likely to arc, though I suppose they can. Paper and mica condensers are much more prone to arcing in high voltage applications. This arcing cannot be duplicated with a conventional low voltage multi-meter, so the problem will not be revealed with one. With my condenser shorts tester, which uses high voltage, an arc would more readily be revealed, as many have for me in problematic equipment. Another tester that shows both leakage and arcs and even capacitance values, and it even uses high voltage to test the condenser, would be the Solar or Sprague capacitor tester. These fine units are readily available on eBay. They are simple in construction, and are easily repaired. Sadly the Solar units sometimes use an eye tube, which looks really cool, but does not give exact values. I bet they're pretty good, though. I wouldn't mind owning one, myself.

As for the bulb tester for electrolytics, I have never tried this, so I will have to and see how I like it. I can see it really telling a good deal about the electrolytic in question. The only drawback I can see is not being able to show a very accurate amount of leakage, and also, since no calibrated meter is involved, the amount the meter swings up for one electrolytic cannot be compared to another of same value, but in known good condition. Since electrolytics dry up, this is just as valuable of information as leakage. The human eye cannot compare brightness levels as well as meter readings. Since I have never used this method, though, I cannot say how well it works. For general filter condenser applications, I can see that in most cases it would work quite well.

Thomas

Hi, Thomas. My ears pricked up as you reminded me of the old automotic condenser checkers. Most filling stations that did mechanical work had them.

Kids, please don't try this: a common stunt among those of us hanging around the filling station was to charge one, and then toss it to somebody - if they were surprised, they would try to catch it. Zap. It was a good way to learn about how caps worked.

Which then led to adapting a Model T coil, with a switch, to shock anybody who touched your car. Inductance cannot be taught any better. Model T coils cost $1 at any junkyard.