Author Topic: Impracticalites of ESR in repair work  (Read 1695 times)

0 Members and 1 Guest are viewing this topic.

Offline Fred BassetTopic starter

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Impracticalites of ESR in repair work
« on: June 20, 2020, 10:58:00 am »
It seems everywhere there is so much information on ESR, what it is, how to measure it, how to visualize it, how important it is and so on.  Thanks to what I have read on here though, my eyes are opened now to the shear size of the variations in ESR.

So pulling a data sheet for a particular range of caps when you are designing makes a lot of sense - But I was trying to compile a chart of ESR values at which a cap should be replaced when doing repair work, but I quickly come to a halt.  The variation even for normal electrolytics kind of makes a nonsense of my chart.  So many unknown Chinese makes (rebrands?) makes obtaining data sheets impossible and even if I did, I am not very confident that the caps will be within spec. even when new.

So although I have a great little ESR meter, what use is it on anything other than well-known and respected brands?  In typical reapir works, it is still quicker and cheaper to replace all likely electrolytic caps with known good ones, or should I persevere?  What are your thoughts please?
 

Offline ocset

  • Super Contributor
  • ***
  • !
  • Posts: 1516
  • Country: 00
Re: Impracticalites of ESR in repair work
« Reply #1 on: June 20, 2020, 11:22:30 am »
As you know, you  have to measure ESR at the ripple current frequency, otherwise the esr value you get is wrong.
With many ripple current shapes  being complex, and not sinusoidal, it means that there are in fact a range of frequencies in the ripple...making it even harder to assess.
With electrolytic caps, the situation is harder because an el cap doesnt "form" until  its got at least 10% of its rated voltage on it...so if your  esr measurer doesnt impose at least this voltage, then again the reading is wrong
 

Offline Fred BassetTopic starter

  • Regular Contributor
  • *
  • Posts: 52
  • Country: us
Re: Impracticalites of ESR in repair work
« Reply #2 on: June 20, 2020, 11:54:16 am »
The meters do seem to be settling now on results at a test frequency of 100KHz.  But so many tables do not state the test frequency - Usually the cheaper ones, that it just seems to add another variable.  I was trying to design a table at 100KHz test frequency, but as you rightly point out, it should be at the ripple frequency - But that also varies so much

I suppose it is not so bad if you are repairing the same boards from the same manufacturer all the time, but for me it is just a hobby now, so I work on anything.  From the day I first found out how bad the tolerances are on them, I have always felt that electrolytics have never been domesticated - They are not housebroken and would be called ferral if they had ever been domesticated enough in the first place...

I do not like blanket changing of electrolytics, so I am testing them, but using my experience to decide if the reading is acceptable or not.  I would love to see what an AI system would make of the collected results over a few years.

Thank you for your thoughts, it has given me more to consider.
 

Offline kripton2035

  • Super Contributor
  • ***
  • Posts: 2713
  • Country: fr
    • kripton2035 schematics repository
Re: Impracticalites of ESR in repair work
« Reply #3 on: June 20, 2020, 12:20:08 pm »
in standard SMPS, I usually replace caps that have an esr of more than one ohm. no need to have any chart for this.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9337
  • Country: fi
Re: Impracticalites of ESR in repair work
« Reply #4 on: June 20, 2020, 12:37:51 pm »
Of course you need at least some kind of a chart. Sizes of elcaps in "typical" SMPS vary by at least two orders of magnitude, so do expected ESRs.

2 ohm ESR is just fine for a very small capacitor which can be on the control side.
0.5 ohm ESR is likely a total showstopper for a large low-voltage output capacitor, say a 4700µF 6.3V thing. Such a beast might be rated at ripple current of, say, 5A, and if, say, 4A ripple is actually applied, 0.5ohm ESR would dissipate 8W!

Fixed 1-ohm limit only works if you only consider a certain range of capacitors. This is your "hidden" knowledge.
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 8276
  • Country: de
  • A qualified hobbyist ;)
Re: Impracticalites of ESR in repair work
« Reply #5 on: June 20, 2020, 01:29:13 pm »
Use your common sense! If the output filter caps of a SMPSU have a high ESR value then replace them. Those electrolytics are low-ESR types and deal with 50kHz or more. The big high voltage filter cap at the primary side sees only 100/120Hz ripple and is still fine in most cases. When you find bad electrolytics look for the same types and check also other caps used in the same part of the circuitry. After repairing a few devices you'll learn what and where to check. The ESR tables are meant as a rough good/bad indication, not as exact must-be-lower-than limits.
 

Offline trobbins

  • Frequent Contributor
  • **
  • Posts: 826
  • Country: au
Re: Impracticalites of ESR in repair work
« Reply #6 on: June 20, 2020, 02:27:55 pm »
As you have now realised, if you don't have a datasheet for the cap you are measuring then it is guesswork as to how to interpret a measurement.  As you will have read, ESR levels vary very widely between brands/models, temperature, measurement frequency, meter (if not using an actual ESR measurement technique), and other factors.  If you don't have all those ducks in a row and accounted for then don't try to guess based on ESR.  Just plucking a value out of the air, or because someone uses a particular line-in-the-sand value is imho just perpetuating a guess.

If you got sold an ESR meter based on advertising spiel, then watch out for other snake oil advertising.

If you are fortunate enough to have made notes on brands and cap values and their ESR values, and come across the same model/value caps in other equipment, then you may be able to assess outlier parts, similar to measurements on a part that you have a datasheet for.

ESR, no matter what frequency it is measured at (within reason), can degrade with age/abuse.  So can cap leakage and cap value.  If any one of those parameters changes outside accepted tolerances then you have an indicator, and can make a judgement call on whether or not to replace the cap, and for what reason.
 

Offline TheMG

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: ca
Re: Impracticalites of ESR in repair work
« Reply #7 on: June 20, 2020, 02:59:31 pm »
ESR meter is a great tool for repair work! I use mine quite regularly and it has never failed to pinpoint bad electrolytic capacitors in-circuit. Also many instances where capacitors (even out of circuit) measured completely fine for capacitance using a DMM.

Don't worry so much about the exact ESR values for the capacitors you are measuring. As long as it's in the general ballpark for that size of capacitor, it should be good.

The user manual for the Peak ESR70 has a handy table of "typical" values for various voltage and capacitances:

https://www.peakelec.co.uk/resources/esr70_userguide_en.pdf

I haven't yet run into a situation where a capacitor is marginally out of spec causing a circuit to malfunction. Any bad capacitors I've found that were causing problems have been very obvious, 5 ohms for a cap rated at less than 0.2 ohms (2400% over spec!!!) to give an example.

If you're worried about finding capacitors that are just barely beginning to degrade in order to ensure long time reliability, you're probably better off just going through and replacing all the capacitors with new ones.
 
The following users thanked this post: Fred Basset

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: Impracticalites of ESR in repair work
« Reply #8 on: June 20, 2020, 07:03:49 pm »
If I can't find a data sheet on the part to be replaced, I look in my collection of caps, new or used, low ESR or not to get a rough measure.  In the end, a bad part will have an ESR that's easily an order of magnitude greater than others of similar value, ratings and size.

If I think the design is too close to what the capacitor(s) is capable of, I'll either uprate it or parallel another part if there's room.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf