Testing electrolytic capacitors

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
I want to determine if a couple of capacitors from a TV power supply are still good. My DC tests below suggest they're fine. Do I need to do anything more?

Both caps are rated at 1,000µF and 25v.

1) Do they hold a charge? Both were touched across a laptop power brick, ie. 20v. They were both over 14v more than a week later. Normally I would take use this test alone to judge them "good". Negligible parallel resistance.

2) Do they have rated capacitance? For this I set up my LabJack to follow the voltage on the cap as it discharged through a resistor. The LabJack itself has about 1.3M impedance and this was in parallel with 3 tests with added resistors; none, 100K and 220K. (The latter were measured on my DMM at 100.5K and 223K Ω.) The starting voltage was again set with the power brick to ~20V.

After fitting the data, I was surprised that the caps came in at 1,132µF and 1,174µF or slightly above spec.

3) Do they allow DC to pass? I hooked the caps up to the DMM as an ammeter and the power brick, all in series. After charging the cap, the current eventually fell to well under 1µA, for instance 0.2µA. Near infinite series resistance.
 

#12

Joined Nov 30, 2010
18,224
There is also "equivalent series not-capacitance" (resistance and/or inductance).
You're going to have to use a higher current/low impedance test to find out what internal reluctance there is to passing current. I don't have one of those in my mind right now but I'm sure one can be found.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
Yeah, I was afraid I needed some sort of AC test. Not sure how to do it without a specialized meter.
 

#12

Joined Nov 30, 2010
18,224
http://fullnet.com/~tomg/esrscope.htm

fullnet.com/~tomg/esrscope.htm

These appear to be the same link, but they aren't.
The basic principle seems to be a conductivity meter using a low voltage square wave. It can be calibrated with simple resistors from .1 to 10 ohms. A "good" cap will show around .1 ohm and "bad seems to be 2 ohms and up. Depends on your needs.
 
Last edited:

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
These appear to be the same link, but they aren't.
Huh? I don't get that.

I've read that post before and must admit that, without a picture, I'm a bit lost trying to follow the setup. I understand the 555 running at 100kHz and 100mV p-p.

My "oscilloscope" is a computer sound card good to about 5kHz. That frequency might be appropriate for a power supply filter capacitor?
 
Last edited:

#12

Joined Nov 30, 2010
18,224
Huh? I don't get that.

My "oscilloscope" is a computer sound card good to about 5kHz. That frequency might be appropriate for a power supply filter capacitor?
I posted 2 links. One to a method to use a scope for an ESR meter, and one to a schematic of a DIY meter. Therefore, the links go to different places, ie, they aren't both the same link.

The important part of an ESR meter is the fast rising edge of the square wave. An audio range sound card is going to have a bad time trying to track that.
 

#12

Joined Nov 30, 2010
18,224
Look at post#4. This site recognized the first link and made it blue. Click and it goes.
The second link (which is on the second line of the post) is still black. You will have to copy and paste it into a new tab to get to the schematic of an ESR tester.

and yes, I believe you can't use a sound card to measure fast digital edges. Better check out the second link.
 

vk6zgo

Joined Jul 21, 2012
677
I want to determine if a couple of capacitors from a TV power supply are still good. My DC tests below suggest they're fine. Do I need to do anything more?

Both caps are rated at 1,000µF and 25v.

1) Do they hold a charge? Both were touched across a laptop power brick, ie. 20v. They were both over 14v more than a week later. Normally I would take use this test alone to judge them "good". Negligible parallel resistance.

2) Do they have rated capacitance? For this I set up my LabJack to follow the voltage on the cap as it discharged through a resistor. The LabJack itself has about 1.3M impedance and this was in parallel with 3 tests with added resistors; none, 100K and 220K. (The latter were measured on my DMM at 100.5K and 223K Ω.) The starting voltage was again set with the power brick to ~20V.

After fitting the data, I was surprised that the caps came in at 1,132µF and 1,174µF or slightly above spec.

3) Do they allow DC to pass? I hooked the caps up to the DMM as an ammeter and the power brick, all in series. After charging the cap, the current eventually fell to well under 1µA, for instance 0.2µA. Near infinite series resistance.
If you are doing this at work,there is an additional question to ask yourself:-

"How much am I costing the Boss,spending time messing round with these electrolytics,when it would be faster to just replace them & be on the safe side?"

When I was fixing a lot of Picture Monitors,I had the luxury of a "magic box" which measured C & R values & losses,but after a while,I didn't bother testing electrolytics,as it was more cost effective to replace the "usual suspects"---scattergun style.

Of course,if it is your own gear,all bets are off,as you are doing it for fun!:D
 

#12

Joined Nov 30, 2010
18,224
What? Those were 2 different pages when I put those links up!:mad:
I even did it over because it looked wrong!

Pencil got the second one right in post #10. That is what the second link was supposed to be.:)

Sorry to mess you up. I'd never do that on purpose.:(
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
Of course,if it is your own gear,all bets are off,as you are doing it for fun!:D
Well, it's long past "fun" at this point but I'm not on a clock, and this is DIY. I'd be happy to simply replace them, but unless the TV then works, I won't know if they were a contributing factor. And I've heard so much about bad capacitors that I've been wanting to learn how to confirm a bad cap.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
Pencil got the second one right in post #10. That is what the second link was supposed to be.:)

Sorry to mess you up. I'd never do that on purpose.:(
No sweat, #12. I'm just glad it wasn't me. Thanks to you and Pencil for the link. I might be able to rig up something similar without too much trouble.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
I finally got around to making myself an "ESR tester". It appears to work but not quite as I expected. I think it's more of an inductance meter. See if you agree, and I'd appreciate any ideas on how to reduce the sensitivity to inductance relative to the ESR.

Here is the schematic:

I use a 555 at 100kHz to supply the signal.

There are a handful of DIY ESR circuits you can find with a little searching. Most include circuitry for driving a meter. I didn't want to bother with that and just wanted to use my DMM for the display.

According to an interesting discussion of their value, none of the DIY circuits are really capable of judging a low ESR capacitor against its specs. On the other hand, a failed cap may have a huge - and detectable - increase in ESR.

Anyway, my circuit above with a 555 running on 9V shows a voltage of about 3.7V at the output. Placing a short across the test leads reduces the voltage to 330mV, but not zero. And the surprising thing was that the length of the jumper wire affected the voltage, longer wire equals higher reading (still less than 1V).

The caps I needed to test (1200µF, 25V, low ESR) gave readings of 540mV. I confirmed that jumpering across the leads dropped the 540mV down the 330mV. So I'm detecting something. Brand new Panasonic low ESR caps I bought as replacements actually gave higher results, about 640mV, but the leads are still attached and not cut short.

A 0.2Ω, 15W resistor gave a reading of 2.83V. :eek:

I assume I'm looking at inductance as much or more than "ESR".
 

Attachments

Last edited by a moderator:

RamaD

Joined Dec 4, 2009
328
And the surprising thing was that the length of the jumper wire affected the voltage, longer wire equals higher reading (still less than 1V).
That is likely to be due to the skin effect. At 100kHz., the longer wire resistance could be significant compared to the resistance (ESR) being measured! And probably that lead to the higher reading.
A Litz wire can be used, which reduces the ac resistance due to multiple strands offering more surface area. Or, even simpler is to use thin gauge multiple winding wires bunched, as these could be easily available.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
My saga with the ESR tester continues here. It's better than I though it was. Results are tough to interpret but it can definitely see differences between capacitors.
 

Thread Starter

wayneh

Joined Sep 9, 2010
17,496
One final follow-up. I went ahead and used a regulated power supply to give me 9.00±0.05V for power. This produced a steady 3.80V at the ESR meter output. Then I ran through my stocks of 1-100Ω, 1/4W carbon resistors to calibrate the meter. Values below 1Ω were obtained by placing 1Ω resistors in parallel.

Results are pretty good!
Screen Shot 2014-10-06 at 5.06.41 PM.png

Then I used the meter to test a bunch of caps I had pulled from a failed SMPS I'm working on. There were four identical size ones rated at 47µF and 25V. The ESR I measured varied from 5 to 38Ω, even though these same caps all showed similar capacitances. I don't know what the specs are but I'm replacing them all. No way I'd put the high ESR ones back on the PCB.

Bottom line, the circuit works and I think this is a useful tool is you have no other ESR meter.
 
Top