What is better to use to determine the leakage of a capacitor, homemade tester, multimeter or transistor tester? It is interesting to know the opinion of experienced people. Perhaps someone has practical schemes for testing capacitors for leakage or loss.
There are a few simple capacitor leakage testing kits available for not crazy money , usually consisting of a boost switch mode high voltage generator and a milliammeter.
You can get proper test gear that does it, including tests such as dielectric absorption. Note there are many different types of tests for capacitors, so you end up with a few different testers, eg conventional value tester, ESR tester, leakage. Some do 2 of these things but not all usually all.
VK5RC
A sourcemeter works well but since I do not have one, I use a power supply and multimeter. Below is a photograph where I tested some NOS solid tantalum capacitors of uncertain heritage.
I stack up all my linear bench supplies to get the max rated voltage for the cap and slide my shonky old fluke 8050 in series. Once it’s gone a few time constants, which is usually pretty fast as the source is very low impedance, it will settle at the leakage current.
I wouldn’t use a switching boost converter as they seem to make the meter dicky for me as it’s measuring very low current.
Really though I know what caps are likely to be duff and just replace them anyway.
The difficulty of the problem is that the capacitor losses are of a different nature and they are all interconnected. By measuring poorly-functioning capacitors, it is sometimes difficult to understand the reason for its improper functioning.
There are a few simple capacitor leakage testing kits available for not crazy money , usually consisting of a boost switch mode high voltage generator and a milliammeter.
Such a homemade tester I have, but there is only 160v. It seems to me that such a tester can see a certain leak, but it does not solve all the problems.
Capacitor losses.
http://www.sayedsaad.com/fundmental/12_Capacitor%20Losses.htm
I'd look at a Heathkit IT-28 - it is on my shopping list - or something very like it. Depending what you're doing you really need up to 600V.
Then you need to do the ESR measuring - but if it leaks to much before it gets to rated voltage it's a bin job.
I spend hours on Mr Carlsons Lab (YouTube) - he has a lot of really good stuff on testing capacitors and insertion the correct way - i.e. outer foil to the shortest path to ground etc - but first you must find said outer foil - well perhaps I'm misquoting but something like that.
Just stick this in your favourite search engine "mr carlson's lab capacitors"
My design goes up to 600V. Resolution 10mA to 100uA full scale. I use it a few times a week.
http://www.pa4tim.nl/?p=1385Besides leakage I measure Capacitance and dissipation factor. These 3 thing will find almost every bad cap. If you really want to find every possible defect you also need an impedance analyser, do vibration tests and temperature cycling, and measure capacitance with a DC bias.
Then you need to do the ESR measuring - but if it leaks to much before it gets to rated voltage it's a bin job.
It is clear that in the capacitor all the losses have a certain connection with each other, but I noticed that this is not always connected with the ESR.
3300uF/6,3v.
That is not the case for the toy on your picture
A cap has a impedance that is the combination of the resistance and reactance. The resistance is the real part, the reactance the imaginary part
The real part is also know as ESR and the ratio of the real and imaginaire part is the dissipation factor (DF). You can recalculate that to the loss angle (tan d) and the tangens from that is the DF (also known as D or 1/Q)
So all are related. Leakage is given as a parameter in the datasheets.
Besides a high ESR, to low capacitance, leakage, there are more defects like cathode rupture and dielectric absorption.
Most defects are related to the oxide layer. That is the DC barrier and the source of the capacitance. Most defects are oxidelayer related.
That is not the case for the toy on your picture
For me, any toy that helps in the work has value.
Moreover, the toy is properly calibrated.
Could you make a circuit in a normal form and a printed circuit board?
This kind of had a capacitor analyzer.
The most interesting thing is that I also have components for this scheme for a long time.
Here everything is clearly visible. The capacitor has worked in the radio receiver for 40 years.
1000uF/10v.
Another good circuit is the capacitor leakage tester.
The leak at the capacitors can be different. If it is large, then an analog tester is enough. But in many cases there is a hidden leak, which manifests itself only at the working voltage. In such cases, the analog tester will no longer help because of the low test voltage. However, there are also difficult cases when it is not possible to determine the leakage, but the capacitor does not work well in the circuit. In this case, the transistor testers are very useful.
One has an additional parameter that determines how the capacitor holds the charge(Vloss). The other determines the voltage to which the capacitor is charged and to the breakdown(DY294).
That is nonsens, my analog leakage tester goes to 600V and the advantage of an analog meter is that you see if it is constant leaking or pulsing (and close to going short) Digital or analog has nothing to do with the height of output. The leakage itself is analog
There is no hidden leak, it is just the oxidelayer that leaks. If that is locally to thin it will pass current while rebuilding the oxide. If the leakage is to high it heats up the layer, electrolyte vaporizes and without that the oxide layer can not recover. The thinner the layer the lower the minimum leakage voltage becomes.
As far as I know the Vloss on your meter is no component parameter but a sort off "fault" of the meter, stated in the original design documentation. There is a huge topic about it.
We do not know what happens to the capacitor in the circuit and how it behaves. Capacitance and esr can be normal, but it does not hold the charge at working voltage and therefore the power supply of the TV can not start.
In addition to toy devices, I use adults up to 1000v.
And I'm experimenting with something that is not being written about in big topics. In-circuit testing of capacitors for a leakage.
Bad capacitor 470uF.
A meter, measure the Farads/ufarads if there 5 to 10% low but alaways will be extremely low caps bad. There is a tester out there that takes t2 leads hookup one on each side and it has a speaker, short bleep good long bad.
A meter, measure the Farads/ufarads if there 5 to 10% low but alaways will be extremely low caps bad. There is a tester out there that takes t2 leads hookup one on each side and it has a speaker, short bleep good long bad.
It's not very clear what you mean.
My capacitor leakage tester 0-160v.
Do you isolate the input 230V?
Similar circuits often use a pair of standard 230 to 12V transformers back to back - to act as a cheap isolation transformer.
I did not have a separating transformer 220/220v. I put the 220 / 160v transformer in a well insulated box.
Another simple scheme of capacitor leakage tester.