This simple capacitor leakage and short tester can be quickly assembled on ad-hoc basis. It is especially useful for testing electrolytic capacitors. In my use it proved so useful that it gained its own case. It tests all capacitors for shorts and opens, and it is useful for testing electrolytic and larger value regular capacitors (from about 100 nF - tens of thousands of microfarads) for leakage.
Capacitor is connected in series with a power supply (I used 12 V) and a voltmeter (I used analog but digital is fine - with caveat - read along). When power is applied, initially voltmeter shows full supply voltage = 12 V in my case. As capacitor is charging, voltage across it increases, while voltmeter voltage decreases (series connection with power supply). After some time (T=R*C), depending on the internal resistance of the voltmeter and the capacity of the capacitor, voltage on the capacitor will reach 63% of power supply voltage. For example, if meter resistance is 10 KiloOhms, and capacitor has 100 microfarads, it will take 1 second to charge the cap to 63%. Charging will continue further, until almost 100% is reached (for a good capacitor).
After several T periods, voltmeter must show very low voltage (0-5% of supply voltage). Otherwise, capacitor is too leaky. The lower the voltage, the better the capacitor is (lower leakage).
Digital voltmeters typically have internal resistances in the 10 MegaOhm range. To prevent an hour long wait with large value electrolytics, you should connect an additional 10 KiloOhm resistor in parallel with the voltmeter. However, very high voltmeter resistance might be useful if you are checking the leakage of small value capacitors (say, less than 10-50 microfarads).
Practice on several known good capacitors to get a knack for this technique. You will be able to quickly spot the bad ones after that.
Typical normal leakage values can be deduced from this Wikipedia article: (https://en.wikipedia.org/wiki/Electro...ge_current) and from capacitor manufacturer's data sheets.
Example 1: Typical leakage for 1000 microF (low voltage, bellow 100 V) cap can be 5 microamps. Taking that as a baseline, any 1000 microF cap with more than 50 microamp leakage is definitely bad. On our 10 kiloOhm voltmeter, this will show up as 0.5 V voltage drop, where a good cap will show 0.05 V or less.
Example 2: 10 microfarad low voltage electrolytic cap will have normal leakage in the 50 nanoamp range. We can safely discard those that are above 500 nanoamps. However, our 10 kiloOhm analog voltmeter will only show that as 5 mV, which is unreadable on a 10 V range. In this case it is better to use a digital meter in parallel with 1 MegaOhm resistor, and we will get 500 mV reading.
Example 3: High voltage electrolytic caps can have normal leakages up to 30 times larger than the low voltage electrolytics. Also, ideally leakage should be tested when high voltage is applied. If we have 100 microfarad capacitor rated at 200 V, leakage up to 150 microamps can be tolerated. On 10 kiloohm voltmeter, this will show up as 1.5 V.
Shorts and opens Shorts can be detected by voltmeter indicating full supply voltage, and not changing. Open capacitors will not charge, and voltmeter will always show zero volts.
Non-electrolytic (regular) capacitors (100 nF and up) Here leakage currents for good caps will be so small that digital voltmeter or VTVM with 10 MegaOhm input resitance is a must. Practice on some good caps to get a knack for spotting the fishy ones. Shorts and opens can still be detected.
FINAL NOTE: check a number of capacitors of the similar value and voltage rating, to determine whether the certain cap is bad.