EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Shock on May 25, 2017, 01:20:52 am
-
I was looking at a few builds of homemade capacitor leakage testers (for measurement and reforming). Mostly they comprise of a variable or switchable DC power supply with an ammeter in series.
Not know much about multimeter design, what about a voltmeter in series instead, what would be the difference in practical application as well as using a digital vs analog multimeter?
-
With a voltmeter in series with the capacitor, it's going to charge kind of slowly through the 10-1000 megaohm impedance.
There isn't much practical difference between analog or digital meters for this job, since the signal of interest isn't rapidly varying. The current will start off high (limited by your series resistor, or the PSU's current limit), and then taper off to the current leaking thru the capacitor. As it reforms over the next several hours to days, the leakage normally decreases.
-
Checking the leakage of the capacitors with a multimeters(analog or digital),is not completely correct.They have low voltage.In practice,it is better to use a tester.
-
Checking the leakage of the capacitors with a multimeters(analog or digital),is not completely correct.They have low voltage.In practice,it is better to use a tester.
Please explain what has low voltage and why is it better to use a tester?
-
Shock
Decades ago when I turned spanners for a living all we had to check ignition caps was a analogue moving coil meter set in Ohms mode. Good or bad, connection would cause a rapid deflection to FS that would subside nearly to infinity IF the cap was OK. Leaky ones were easy to pick.......but we did have new replacements to compare against. :phew:
As a quick and easy check in the absence of better gear it worked just fine.
That meter was a Japanese Kaise SK-160 that I still have and use occasionally.
-
Please explain what has low voltage and why is it better to use a tester?
(http://imageshack.com/a/img924/3959/QmKwIq.jpg)
The leakage is very strong-30kOhm.
2200uF/10v.
(http://imageshack.com/a/img924/9189/i2Cn55.jpg)
(http://imageshack.com/a/img923/683/Hisaq3.jpg)
Capacitor 3300uF/10v.
The leakage is not strong-300kOhm.
(http://imageshack.com/a/img922/3259/E5oQ8f.jpg)
(http://imageshack.com/a/img924/4921/omXTUk.jpg)
Leakage at different voltage.
(http://imageshack.com/a/img922/1962/C3FU6V.jpg)
(http://imageshack.com/a/img922/3606/v3ysLs.jpg)
A good tester.
(http://imageshack.com/a/img924/2639/YQO7xo.jpg)
-
This might help -
http://www.tek.com/sites/tek.com/files/media/document/resources/LowLevelHandbook_7Ed.pdf (http://www.tek.com/sites/tek.com/files/media/document/resources/LowLevelHandbook_7Ed.pdf)
Regards, Dana.
-
Checking the leakage of the capacitors with a multimeters(analog or digital),is not completely correct.They have low voltage.In practice,it is better to use a tester.
Please explain what has low voltage and why is it better to use a tester?
It does not apply to what you are suggesting. Using a separate power supply with the voltmeter in series will work fine since you can make the test voltage whatever you want but only for very low leakages unless an external shunt is used.
-
OK.
The leakage of capacitors can be different.Therefore,it is necessary to use different divices.
Capacitor removed from the motherboard.
3300uF/6,3v
Leakage.
(http://imageshack.com/a/img924/5051/mn78JA.jpg)
(http://imageshack.com/a/img924/563/0iRFj7.jpg)
-
I have done using 287 DMM that has 10 Mega Ohm resistance at voltage mode, basically it works.
Example : The leakage chart and you can see the forming result for about 1.5 hours from a NOS polymer cap OSCON at it's rated voltage, at the chart reading, Y axis -> 1 volt equivalent to -> Volt / (10M Ohm DMM internal resistance) -> 0.1uA
(https://www.eevblog.com/forum/beginners/is-there-any-practical-or-quick-n-dirty-method-to-measure-capacitor-leakage/?action=dlattach;attach=163731;image)
More detail here -> Measure Capacitor Leakage with DMM (https://www.eevblog.com/forum/beginners/is-there-any-practical-or-quick-n-dirty-method-to-measure-capacitor-leakage/msg722332/#msg722332)
-
Here is a photograph of me burning in some NOS solid tantalum capacitors of unknown heritage. One multimeter is set to measure 2 milliamps full scale yielding 0.1 microamp resolution which is plenty for this. The power supply is configured for a very low current limit so a short will not damage the current shunt.
I would not mind having a dedicated high voltage tester for capacitors and semiconductors which measures leakage down into the picoamp range. At some point I may design something.
-
Here is a photograph of me burning in some NOS solid tantalum capacitors of unknown heritage. One multimeter is set to measure 2 milliamps full scale yielding 0.1 microamp resolution which is plenty for this. The power supply is configured for a very low current limit so a short will not damage the current shunt.
David, whats your experience with those "NOS" thru hole dipped tantalum when you did that ? As I have lots of them too.
Did they also take time to form ? Or any other interesting facts that you can share , especially for those no/unknown brand parts ? >:D
-
I have a Sencore LC-101 I got off ebay broken. I don't think much of Sencore stuff, I was just bored. The leakage test is the most used thing about it. A number of voltage steps up to 1KV. A handy test for semiconductors. It also has a reforming function, a spring on a rod that presses the leakage button to the handle.
-
Shock, if you have a multimeter with current measuring resolution in microamps and a power supply of sufficient voltage with the ability to set a current limit that is low enough to protect your meter, then that should work fine for measuring leakage. Being able to log the current values over time would be a bonus.
-
Shock, if you have a multimeter with current measuring resolution in microamps and a power supply of sufficient voltage with the ability to set a current limit that is low enough to protect your meter, then that should work fine for measuring leakage. Being able to log the current values over time would be a bonus.
This.
Any good benchtop or better class handheld multimeter is fine as long as you can measure µA's, also use a low ripple/noise powersupply, now you reliably can measure leakage current.
If the caps are old stock and i dont need a whole lot of them i usually let them sit at just a few volts for a minute or two before i ramp up the voltage to just a few volts less than the full rated voltage.
-
A good japonese capacitor. ;)
35years.1000uF/50v.
(http://imageshack.com/a/img923/874/NFEKpA.jpg)
(http://imageshack.com/a/img923/7633/2ZN46M.jpg)
-
I have done using 287 DMM that has 10 Mega Ohm resistance at voltage mode, basically it works.
More detail here -> Measure Capacitor Leakage with DMM (https://www.eevblog.com/forum/beginners/is-there-any-practical-or-quick-n-dirty-method-to-measure-capacitor-leakage/msg722332/#msg722332)
Thanks Bravo didn't see that post. Looks good enough for me.
Thanks everyone else for your suggestions and for reading the OP.
-
Here is a photograph of me burning in some NOS solid tantalum capacitors of unknown heritage. One multimeter is set to measure 2 milliamps full scale yielding 0.1 microamp resolution which is plenty for this. The power supply is configured for a very low current limit so a short will not damage the current shunt.
David, whats your experience with those "NOS" thru hole dipped tantalum when you did that ? As I have lots of them too.
Did they also take time to form ? Or any other interesting facts that you can share , especially for those no/unknown brand parts ? >:D
When I bought them from the surplus electronics store, I had my multimeter with me and checked them individually for value and resistance; none failed. When I put them on my hacked together burn in tester, their leakage at their rated voltage immediately dropped significantly below the typical specification for a solid tantalum and I left them there for a couple of days.
They replaced 16 volt solid tantalum that shorted out on a 15 volt regulated supply which is no surprise. One of them shorted out after being powered for more than an hour while I was watching. With a voltage derating of 1/2 to 2/3rds, reliability is 100% or close to it.