Best Capacitor Tester?

[Generalquark]

Hi,
 I was looking for advice from individuals with more experience than myself. The last couple pieces of equipment I have worked on ended up being bad electrolytic capacitors. I discovered them because of the voltage drop I saw in the circuit. I traced the drop back and swapped them, finding it fixed the problem. They both had checked out ok with a LCR meter, reading in the range that I would expect. This leads me to find another way to measure these capacitors. I should have checked ESR on these before getting rid of them, but I didn't think of it until after I had lost track of them. I was looking for a setup to test a capacitor in every way it could possibly fail. Currently I have
Handheld BK Precision 878
An AnaTek Instuments Blue ESR Tester by Bob Parker (I just picked this up as part of a lot I purchased and have not even used it yet)
I also have a QuadTech 1920 Precision LCR Meter

I don't believe any of these instruments can check Leakage, and from what I am aware of, that is one of the 3 things I should be checking.
Can someone confirm the three things to check to ensure a capacitor is good would be capacitance, ESR, and leakage?

Is there a device on the market that can verify a capacitor is good or bad 100%?
I was looking a bit at the Sencore LC75, LC102, and LC103.
I have a Sencore PoweRite with a leakage probe, but I am thinking the voltage range is limited for that purpose.

I did have another question. I doubt there is anything handheld on the market that could test Leakage, but I was curious what you guys though about the DE-5000 vs the BK 878.
Browsing through the forum it seems that is one of the most highly esteemed handheld LCR meters.
Is there anything better in a handheld package?

I have read the documentation on the Anatek ESR meter and still am unsure of where it fits in. I believe it claims the ability to troubleshoot in circuit and things like Detection of shorts on PCB's and shorted coils in transformers. Is that unique to this device or is this something that the DE-5000 can duplicate?

Lastly,
 I found an interesting device called a leakseeker 89 that claims to be able to calibrate to detect exactly on the pcb a short may be. That sounds appealing if true. Does anyone have any experience with something like that? I am looking for something of an all in one device if such a thing exists. Something to completely test all capacitors, inductors and transformers down to a 1 coil short. Perhaps all this is already possible with equipment I have and I just don't have the knowledge, or perhaps it is similar to the handheld LCR/DMM combo meter I wish existed, but no one seems to make one for some odd reason.

Any help would be greatly appreciated. I always love seeing your replies on this board. I am thankful for the time you have invested to helping others out with answers.

[bob91343]

Testing capacitors is only part of what you want to do.  Yes, leakage testing is not addressed in many units.  In fact I have two devices that test for leakage but I really only use one, the GR 1644 megohm bridge.  It allows any voltage up to 1000 V to measure leakage.  Actually it measures leakage resistance but it does it at any voltage I select.  This is important, as some bad parts don't start leaking until the applied voltage reaches some point.

When a capacitor leaks, it draws in-phase current and so its dissipation factor increases and it warms up.  If this is too much, the generated heat can increase leakage due to higher temperature, which turns it into a potential runaway situation.

ESR is highly overrated.  Mostly you don't need to measure it.  Capacitance is important of course, but so is self resonant frequency.  The latter is not likely to change as a part ages so is best used as a test for suitability for a particular application.  An ESL measurement can reveal this.  You need to measure ESR when you are deciding which capacitor to use in a given circuit.  Too high a value can defeat the purpose of the part.

The bottom line here, is that you ought not to try to buy an instrument to measure something until you fully understand what you should measure and why, and the limits on an acceptable measurement.

[coromonadalix]

The problems with some of the mentioned lcr's,  some can do esr dcr tests,  others don't

The Sencores are very good lcr's, but they kinda get outdated and are pretty complicated to calibrate or repair ... i had an LC102,   loved the 1kv capacitor test, but i got stuck with other tests the lcr can't do

The BK'S  lcr's like the 878  885 886 889 are special beasts, only the 885 and 886 schematics are now known,  got the help here to find most of the ic's in it,  helped me to complete another thread here, who identified the dds chip in it.


I work with an DE-5000 and im very satisfied for the moment, can do serial or parallel modes

You have tons of lcr's models, think about calibration and repairs if you need to   EX: forgot to discharge a capacitor ...  lots of damage in a defunct lcr

Some models can have an added dc bias option ...


For the leakseeker : 
I personally don't believe a gadget like this, could be some kind of a signal tracer

A good meter with an option of zeroing the leads can do a good job,  an thermal camera can do a good job too.


There is no ultimate appartus to test and find or report everything. Like the previous thread  you have to know your equipment(s) limit(s)  and work with them

Your 878 can be good for you  bad for the others, an DE-5000 is good for me        and  you're gonna start an endless debate over this and that  loll


You want to be sure of a passive part like a resistor, a capacitor  or a coil, remove it from the pcb and test it,  you can be fooled by many things on a pcb

I work on 27 Mhz RF stuff, they are transformer coupled in many places of the rf stages in the board,  i see tons of shorts with an ohm meter, but it's not,  it work at 27Mhz  not in DC  loll

[TMM]

You generally only need to test leakage on high voltage caps (>100V). 1mA leakage at 350V = 0.35Watts which may cause the capacitor to heat significantly which in turn causes it's leakage to worsen and heat even more - a run away condition until the capacitor goes bang or your circuit gets cooked from overcurrent. 1mA at 3.3V is only 0.0033W which is generally a non issue. For this i just have a low current, high voltage supply made of a small low voltage transformer and cockcroft-walton multiplier which gives me 50VDC taps all the way up to 450V. An ammeter measures current. You connect the cap onto the appropriate tap and after a minute or so see what the current has settled to.

If you don't want to build something then an insulation resistance meter (e.g. Fluke 1507/1587) could be used to measure leakage but you're far more limited in terms of output voltage selection.

For ESR/Capacitance measurement I use a DE5000.

[bdunham7]

I use a BK 886 which has variable test levels of 50mV, 250mV and 1V, the idea being you can often -- but not always--get an accurate reading in-circuit at the lower test levels.  It also is 4-wire and has test frequencies from 100Hz to 100kHz.  It is also a pretty old design and quite expensive, although they apparently still make and sell them.

For leakage, which can only be done out-of-circuit, I simply have an HP 6515A variable HV power supply and a handheld DMM with a microamp range.  If I'm really feeling silly, there's the picoammeter.  That way I can test any capacitor at exactly its rated voltage.  I see no need for a special-purpose tool for this basic test.

[David Hess]

I also just use a power supply and multimeter to test leakage when it is warranted.

I have found dissipation more relevant than ESR for quickly determining if a capacitor is bad but they are effectively measuring the same thing; dissipation is roughly constant with frequency while ESR is not.

I would not mind finding an instrument which measures capacitance, ESR, D, and leakage, but have not so far.

[graybeard]

Since I bought my DE-5000 I use it for measuring capacitors about 95% of the time.  I like the multiple frequency options it has.

For checking capacitors in tube gear at voltage and checking their leakage I use a ZM-11/U 1KHz impedance bridge.

A few times I have checked leakage of capacitors down to the pA level with my Keithley 617 electometer and a Faraday cage shielded box.

I have used current probe with my Siglent scope and function generator to characterize capacitor impedance.

I have also used my VNA to get the ESR and approximate resonant frequency of capacitors.

I even occasionally use a multuimeter's capacitance function.

For just measuring capacitance I have a nice comparison of several multimeters, the DE-5000, and ZM-11/U here.

Although I have all of these options when I am trouble shooting my DE-5000 has been my go to instrument since I bought it two years ago.

Chris

[kripton2035]

I can answer fot the short finder part...
http://kripton2035.free.fr/Projects/shorty-display.html

and I also have a schematic for a leakage tester (I never tested it !) that works wor low voltages only
http://kripton2035.free.fr/LCR%20meters/lcr-capaleakmete.html
http://kripton2035.free.fr/digital%20esr/esr-sprut-de.html

[joeqsmith]

Normally, I will see the alum elec physically change.   Bulging, leaking, etc.   I am not one to normally use the shotgun approach and wildly change out all the caps to try and troubleshoot a problem.  Some time back there was a large thread about a video that Carlson's Lab had made showing a capacitor tester he had made.   It was a feeling driven debate and was locked.  I found the topic interesting and started it up again but tried to keep it more data driven.  Maybe you can find something of value there:
 
https://www.eevblog.com/forum/testgear/using-low-voltage-levels-to-test-hv-capacitors-leakage-current/

Most handheld meters can measure C.  I have one that can also measure L but it's pretty sketchy at best.   I also have a couple of BK RLC meters which have their limitations.   This video shows me attempting to measure the ESR for a high Q ceramic capacitor using the BK and a couple of low cost VNAs.   Not what you are asking about but you may find them of interest.  Contrary to what I am often told, the VNA isn't just for measuring antenna SWR.   
 

[bdunham7]

Some time back there was a large thread about a video that Carlson's Lab had made showing a capacitor tester he had made.   It was a feeling driven debate and was locked.  I found the topic interesting and started it up again but tried to keep it more data driven.

I've seen Mr. Carlson's video on his leakage tester, but missed the threads here on it.  A while back I acquired a NVOS (New Very Old Stock) BK Precision Insulation and Ground test meter that has a selectable 250V/500V/1000V test voltage.  My intent was to use it as a source for testing capacitors.  First time out of the box I fired it up and watched the meter climb to about 950 volts and then a pop noise and no output.  The 1.2kV-rated output film cap had shorted.  I only need that one data point to know that low-voltage testing is incomplete.

The OP said:

I was looking for a setup to test a capacitor in every way it could possibly fail.

That's ambitious, but I certainly wouldn't want to overlook the occasional HV flashover or breakdown even if leakage were typically linear.

[HighVoltage]

My vote goes to the Sencore LC103
It is the best capacitor tester I ever had.

[joeqsmith]

My vote goes to the Sencore LC103
It is the best capacitor tester I ever had.

Found your thread along with a manual for it.  Looks like it's limited to 0.1 ohm.  My BKs can't make these measurements.  Even the cheap VNAs struggled. 

http://ftb.ko4bb.com/manuals/71.201.191.60/LC103_Meter_Operator_Manual.pdf

[HighVoltage]

One specialty of the LC103:

=======
"RINGING TEST (U.S. Patent #3,990,002)
A dynamic test of inductor quality determined by applying
an exciting pulse to the inductor and counting the
number of cycles the inductor rings before reaching a
preset damping point."
=======

The only bad thing about the LC103 is the price, it has slowly increased over the years.

[kripton2035]

did not know this ringing test was patented. there are a lot of other test devices that does it too.
bob parker did one.

[Generalquark]

Thank You for all the replies.
I will need to read over the other threads and watch the video in the next couple days here.
I see that the DE-5000 is very popular. I will probably ditch my BK 878 and get one of them. It seems to have a bit more functionality.

That would give me capacitance, dissapation, and ESR or I could use the Quadtech for those measurements.
Do I need an ESL measurement on the capacitor? If yes, is that done on a scope with a function generator?

I like the idea of using a power supply and measuring current on a DMM
I wonder if I could use my Sencore PoweRite to vary the ac output voltage and run it through a rectifier to convert to DC then to the meter and capacitor? Do you guys think this would work? It could get caps all the way up to the 130V area. I don't mind purchasing a power supply that can hit the higher voltages if necessary. Is the HP 6515A the best power supply for the job? I don't think I would ever be testing caps above 1000V that I can think of. I am not too familiar with high voltage supplies and which ones are out there.
Does this setup mimic a bridge?
If this would work, that would take care of the leakage measurement.
I would probably need to get a different DMM. I currently have a Fluke 179, but would probably need to go with a 87V instead to get into the current range that would work well with all caps. Am I reading the specs correctly that the 87V goes down to 10nA?

Are there any other readings that I should be taking?
I have done some looking at the LC103's but the prices are high as stated, and it looks like there seem to be alot of issues with calibration.

kripton2035 Thanks for the link. I may build the shorty in the future.
Also, you proved me wrong. There IS a DMM/LCR The DE-5004 has both. Unfortunately, it is a somewhat stripped down version of both instruments. Now if they could combine the full functionality of the DE-5000 with something like a Fluke 87V I would purchase it for sure and pay more.

Forgive my ignorance, but how do you test HV flashover or breakdown?

Thanks Again for all the suggestions and sharing your knowledge.

[bdunham7]

That would give me capacitance, dissapation, and ESR or I could use the Quadtech for those measurements.
Do I need an ESL measurement on the capacitor? If yes, is that done on a scope with a function generator?

You certainly don't need ESL as a routine diagnostic tool.  Your Quadtech 1920 should blow away any other LCR tool we're discussing, but you'll need to use the PC functionality (LabView or the Quadtech program) and really learn how to utilize it.  You want a plot of Z and θ to start. 

I like the idea of using a power supply and measuring current on a DMM
 Is the HP 6515A the best power supply for the job? I don't think I would ever be testing caps above 1000V that I can think of. I am not too familiar with high voltage supplies and which ones are out there.
Does this setup mimic a bridge?
If this would work, that would take care of the leakage measurement.
I would probably need to get a different DMM. I currently have a Fluke 179, but would probably need to go with a 87V instead to get into the current range that would work well with all caps. Am I reading the specs correctly that the 87V goes down to 10nA?

I'd recommend a standalone HV power supply, preferably one with limited current output, and a cheap DMM with a 200uA range.  The Harbor Freight free one is as good as anything for this.  You want separate, inexpensive instruments for the leakage part just in case something goes wrong--no point in blowing up an expensive LCR tester or your best multimeter.  The HP 6515A just happens to be what I have, but you probably only need 600 volts for most things, 1000V if you think you need it.  And insulation tester ("megger") can provide that fairly inexpensively, but usually only a few select voltages.  If you do get an HP 6515A, look for one with 3-way binding posts so they'll take regular leads--the BNC-like HV connectors that many of them have are just an annoyance unless you like making custom cables.

Forgive my ignorance, but how do you test HV flashover or breakdown?

Just do the leakage test, if it shorts out one of the above has happened.

[mr.fabe]

Although it's not lab grade equipment, the Sencore LC's would cover all the troubleshooting tests that you listed including the HV breakdown testing on diodes.  The LC101, 102 and 103 can test cap leakage up to 1kV and the 103 can do in-circuit testing.  Calibration is easy on all LC's except for the LC103.  The Sencore PowerRite is an isolation autotransformer and not the best instrument to use for the leakage testing of capacitors.

[Smoky]

The Sencore LC53 is no slouch either!

This meter requires an electrolytic capacitor replacement job, but when you're finished, you'll find this meter to be extremely accurate. No calibration was done to mine whatsoever.

I haven't touched the insides of my LC102, but as it sits, its capacitance measurements can't hold a candle to the LC53. The LC53 tests caps up to 600v (verified) and also has the ringing test.

I have two reference 1% capacitors that I sample every time I use my LC53 and it never fails to hit the mark.

This is a 1% Cornell Dubilier 100pf 500v Mica:



Treat the LC53 to some quality electrolytics and you'll be one proud and happy owner

[HighVoltage]

did not know this ringing test was patented. there are a lot of other test devices that does it too.
bob parker did one.

Yes, I read the patent a few years back, when I got to know the Sencore LC103.

The LC53 and the LC102 are also some great instruments. But these also had a healthy increase in price over the years.
 

[Generalquark]

Has anyone used the LV Carlson Capacitor tester. If you have used it, did it catch every bad capacitor you have come across?

[Dwaine]

I used a fully restored heathkit IT-28.  Also the new hantek 1833c.

[Generalquark]

If I were to go the power supply route and manually test leakage that way, does anyone know what the max normal leakage rating of any capacitor I would see would be? I would like to select a power supply that had at least that much current capability. I was also curious if anyone has used electrophoresis power supplies? These supplies looked interesting and probably decent capability for the price. Any other suggestions?
https://www.ebay.com/itm/292547051057?epid=23017593285&hash=item441d29be31:g:8fkAAOSwkbxa6esC
https://www.ebay.com/itm/224183808984?epid=20037062981&hash=item34326573d8:g:mE8AAOSwcDBd8-xb
https://www.ebay.com/itm/203463522425?hash=item2f5f5ef879:g:tKYAAOSwtC1gqUBI

[bdunham7]

If that first one can put out 200mA, that should cover even a relatively leaky 3 kilofarad double layer supercap. Your concern with those caps would not be the current required, but the need to accurately regulate the voltage at a low level.  Actually, anything with leakage over a few milliamperes would likely only be a very low voltage capacitor, so you could simply use a bench PSU in those cases. 

[paulbt]

Hi !

What about the ESR70 model from Peak Electronic Design? Does anyone own one? If so, does it worth the money?

[kripton2035]

What about the ESR70 model from Peak Electronic Design? Does anyone own one? If so, does it worth the money?

it will not give you any leakage info, only esr and capacitance