Capacitor Rabbit Hole. Just How Do You Measure?

[Woofers and Tweeters]

37 minutes ago, CECAA850 said:

Chrysler, Dodge, Jeep, Ram.

By money, I was lured between a couple of Chrysler dealerships in the late 80s and early 90s. I was factory trained in automatic transmissions, and the A604 was job security. 

[CECAA850]

3 minutes ago, Woofers and Tweeters said:

By money, I lured between a couple of Chrysler dealerships in the late 80s and early 90s. I was factory trained in automatic transmissions, and the A604 was job security. 

That was before my time with the brand although the brand today has nothing in common with the brand then other than the name.  I started here in '97.

[Woofers and Tweeters]

13 minutes ago, CECAA850 said:

That was before my time with the brand although the brand today has nothing in common with the brand then other than the name.  I started here in '97.

It's a good thing too.

Every US auto manufacturer in the 70s and 80s seemed to try to rival each other as to which one could make the worse car, Chrysler seemed to try to take that challenge to a new level.  

[CECAA850]

2 minutes ago, Woofers and Tweeters said:

Every US auto manufacturer in the 70s and 80s seemed to try to rival each other as to which one could make the worse car, Chrysler seemed to try to take that challenge to a new level.

Fiat is cutting edge in that department.  State of the art.

[jimjimbo]

Aren't there car threads elsewhere in this forum?  Would be nice to keep this to the cap issue, or, non issue, whatever.

[Woofers and Tweeters]

1 minute ago, jimjimbo said:

Aren't there car threads elsewhere in this forum?  Would be nice to keep this to the cap issue, or, non issue, whatever.

Aren't there grouch threads elsewhere in this forum?
😋

You must be new here. 

[Dave A]

52 minutes ago, jimjimbo said:

 

Dave, do you plan on purchasing a cap kit from them and doing some measurements, or is that already in the works?  I will be happy to contribute to that effort.....

Yes I am interested in trying a set. Right now though I don't have anything requiring recapping. It's a shame that these just became available as I am pretty serious about not fixing up old speakers anymore. Too much trouble to find, too expensive when you do and always a trip on top of it all.

[Bubo]

I have to believe that the Cap mfgs must run frequency and DC leakage tests for their products to characterize them.

And that the equipment mfgs must bring in the top 10 meet or exceed spec caps to test and listen to  in the design.

 

Thanks to Deang for his measurements

 

This is the spec sheet from Nichicon's premium "audio" capacitor Muse

Measurement made at 120HZ

Not sure I understand Tangent of loss spec, loss across cap uniformity at tested frequencies ?

http://www.nichiconcapacitors.com/pdf/e-ukz.pdf

 

https://www.nichiconcapacitors.com/product/nichicon-kz-muse-1000uf-25v/

[Dave A]

3 hours ago, RandyH000 said:

the AA crossover is not a car with computers , sensors and complex electronics , the AA  is  not even a complex  network like the latest AL5  or AK6  , and  if JEM can   fix these new networks -then they are experts in crossovers -

 

the AA   XO was operative on the mids and woofer but fainter on the tweeter and , the tweeter was not the issue , what is the only  part  left  ?  , the cap for the tweeter ----you want  to send in a crossover , pay shipping back and forth , plus labor , for 1 bad cap -----my  8 year old niece could have called this one  - 

What you said earlier today did not mention diodes as a problem but did say the only thing left possible was the capacitor. I remember even if you don't. Bad information is far worse then no information.

[wstrickland1]

23 minutes ago, jimjimbo said:

Aren't there car threads elsewhere in this forum?  Would be nice to keep this to the cap issue, or, non issue, whatever.

When the weather warms up I'm going to need you in the Tiki Hut Forum of b.org. They're all in hibernation right now so go ahead and finish this up.

[Bubo]

Capacitor Self-resonant Frequency and Signal Integrity

https://resources.pcb.cadence.com/blog/2019-capacitor-self-resonant-frequency-and-signal-integrity

 

Frequency dependent capacitor for AC sweep analysis in LTspice

https://electronics.stackexchange.com/questions/410388/frequency-dependent-capacitor-for-ac-sweep-analysis-in-ltspice

 

Capacitor AC Response

The frequency dependent impedance of a capacitor is called capacitive reactance.

http://hyperphysics.phy-astr.gsu.edu/hbase/electric/accap.html

[Bubo]

 

At different frequencies electrolytic capacitors have different capacitance. Suppose it is 50 uF 16V capacitor (type K50-16), then the measurement results are (using russian E7-14 RCL tester):

1) At 100 Hz - 57,31 uF
2) At 1 KHz - 38,20 uF
3) At 10 KHz - 3,56 uF

So my question is - which frequency the manufacturer had in mind?

measurement capacitance

share improve this questionfollow

edited Mar 4 '17 at 11:59

Autistic

10.5k22 gold badges2222 silver badges4747 bronze badges

asked Mar 4 '17 at 11:18

Ernestas Gruodis

17711 silver badge1010 bronze badges

•  
  At a frequency where its ESR is negligible compared to its capacitive reactance. Look up its ESR in the datasheet and see if it explains the anomalous results at higher frequencies. – Brian Drummond Mar 4 '17 at 11:23
• 1
  Special electrollytics should be used when high frequencies are applicable, such as with switch-mode circuits. – skvery Mar 4 '17 at 11:40

•  
•  
  Hard to find info about K50-16 russian capacitors, but I have noticed, that for similar types like (K50-15) here it is clearly mentioned test condition: "Capacitance tolerance (25 °C, 50 Hz), %". So I suppose I should use 50 Hz test frequency (in my case only 100 Hz available, but it is closer to truth). – Ernestas Gruodis Mar 4 '17 at 12:37
•  
•  
  Why measure an electrolytic? If it's new and bought from a reliable source just use it. If it's old then don't use it. How much do you value your time? – Andy aka Mar 4 '17 at 14:29
•  
  If its good, why to replace? But how do you know if its good? So this raises the question how to test it properly. – Ernestas Gruodis Mar 4 '17 at 14:42
•  
  You have your question backwards. Instead of asking what conditions you should reference, you should instead start from the conditions which you wish to deal with. If your cap is a filter on the output of a bridge capacitor with a 50 Hz line frequency, 100 Hz is the nominal condition of interest. If you are using the cap as a blocking cap in an audio amplifier, then your audio frequency range is what's important, If you don't know how your cap will be used, it is impossible to tell what test condition is appropriate. – WhatRoughBeast Mar 4 '17 at 15:28
•  
  Yes, I understand. Just want to test as it was tested at factory conditions - at specific test frequency, as they did. Its like testing radio tubes - at some written conditions in datasheet. – Ernestas Gruodis Mar 4 '17 at 17:00

 

The answer to a slightly different question, which is "what is the standard measurement parameters for electrolytic capacitor capacitance?" is as follows:

Modern electrolytic capacitors use a standardized measurement method to determine their capacitance. This method is 20°C and 0.5VAC at either 100 or 120Hz (which is derived from the traditional common usage of electrolytic caps, which was to smooth the output of a bridge rectifier. That ripple is going to be 2*50Hz or 2*60Hz, depending on the country). The exception is photoflash capacitors, which typically have their DC capacitance measured.

Virtually all electrolytic capacitors will cease being capacitive at all before 100kHz, and beyond that, will be inductive. This is universal to all electrolytic capacitors. Generally, an electrolytic capacitor will gain about 10% capacitance if the DC capacitance is measured. It will lose about 10% by 1kHz (or more in the case of your capacitors - but they say they are designed for DC and ripple applications, so they are probably not optimized for anything above 120Hz).

But to actually answer your question, "at what test frequency should electrolytic capacitors be measured?" They should be measured at the frequency of interest for your application. That's the frequency they should be measured at. Why would you ever measure their capacitance at anything except the frequency of most interest to your planned use?

Which makes me wonder - why are you even doing this? It sounds like you are trying to verify if a capacitor is within spec. There is no reason to do that however, and you are just wasting your time if you are.

The failure mode of electrolytics does not involve a reduction in capacitance, so measuring it and comparing that value to what it was designed to have is not useful.

Beyond that, virtually every electrolytic capacitor, including those russian ones, have pretty awful tolerance. They can vary from -20% to +80% the rated value. So even if loss of capacitance WAS a valid way to determine if an electrolytic had gone bad or not (which it isn't), then a capacitor that reads the right value might have began life at +80% that value, and has lost nearly half of its capacitance, but you'd never know. Not that this ever happens though.

The failure mode of electolytic caps is that their electrolyte dries out. This does not have a meaningful impact on their measured capacitance. What it does do, however, is cause ever increasing ESR, until eventually the ESR is so high that they are no longer useful as capacitors. Even then, they will have the 'right' capacitance. So I can think of nothing useful being accomplished by verifying the actual capacitance value.

Now, if you don't know the value of a capacitor, because the label has fallen off, the lettering rubbed off, or whatever, then that's when you can do something useful by measuring the capacitance. Or if you need to verify the capacitance at a specific frequency of interest.

https://electronics.stackexchange.com/questions/290252/at-what-test-frequency-should-electrolytic-capacitors-be-measured

[Bubo]

Capacitor Comparison using Bass guitar and test measurements

Found no real differences in test or audible

Instrument repair tech

https://www.talkbass.com/threads/capacitors-and-tone.1442449/

 

The musician boards seem keen on which caps to use with instruments

[Dave A]

8 minutes ago, Bubo said:

Capacitor Comparison using Bass guitar and test measurements

Found no real differences in test or audible

Instrument repair tech

https://www.talkbass.com/threads/capacitors-and-tone.1442449/

 

The musician boards seem keen on which caps to use with instruments

Do you have any similar information on poly caps?

[Bubo]

2 hours ago, Dave A said:

Do you have any similar information on poly caps?

After I finish my slice of carrot cake and PG tips tea

I may dive back in........

 

On my long list of things is setting up a test rig for caps just for fun

Resistors is another one that easier to demonstrate noise and design.

 

My thought is to send a tone through the cap and measure using an oscilloscope on the in and out and comparing

Within frequency range, you may be able to characterize a cap the same way you would characterize an amplifier using test tones,

[Dave A]

Also thinking about those of you who asked if I am going to test these caps from JEM. Please understand I am not an expert on these and all I know to do is measure capacitance and ESR. I have no way of measuring crossovers and all I can give you is my subjective opinion on what I hear with my ears when done recapping things. I have no personal knowledge to know what the design intent parameters for Klipsch crossovers are nor how to match that. I either like what I hear or don't and in general make it a rule to replace capacitors with high ESR which has always made things sound better to me. All I will be able to do is measure the capacitance and ESR and tell you if I like them. I like Dayton caps too and others whose opinion I respect do not so you might not agree with what I find.

[Bubo]

1 hour ago, Dave A said:

Also thinking about those of you who asked if I am going to test these caps from JEM. Please understand I am not an expert on these and all I know to do is measure capacitance and ESR. I have no way of measuring crossovers and all I can give you is my subjective opinion on what I hear with my ears when done recapping things. I have no personal knowledge to know what the design intent parameters for Klipsch crossovers are nor how to match that. I either like what I hear or don't and in general make it a rule to replace capacitors with high ESR which has always made things sound better to me. All I will be able to do is measure the capacitance and ESR and tell you if I like them. I like Dayton caps too and others whose opinion I respect do not so you might not agree with what I find.

My thought is to send a tone through the cap and measure using an oscilloscope on the in and out and comparing

Within frequency range, you may be able to characterize a cap the same way you would characterize an amplifier using test tones,

 

I'm no xover guy....but the caps roll off at different freqs, combined with inductors becoming more resistant with frequency

The faster they roll off the steeper the crossover point......

 

It should be possible to characterize a cap like an amplifier within it's range.

[jimjimbo]

4 minutes ago, Dave A said:

Also thinking about those of you who asked if I am going to test these caps from JEM. Please understand I am not an expert on these and all I know to do is measure capacitance and ESR. I have no way of measuring crossovers and all I can give you is my subjective opinion on what I hear with my ears when done recapping things. I have no personal knowledge to know what the design intent parameters for Klipsch crossovers are nor how to match that. I either like what I hear or don't and in general make it a rule to replace capacitors with high ESR which has always made things sound better to me. All I will be able to do is measure the capacitance and ESR and tell you if I like them. I like Dayton caps too and others whose opinion I respect do not so you might not agree with what I find.

I had thought that you were also considering using a calibrated mic and REW or some such thing.  I understand you were looking for input on that.  I don't think capacitance and ESR values are the question here, but please correct me if I'm wrong.

[jimjimbo]

1 hour ago, wstrickland1 said:

When the weather warms up I'm going to need you in the Tiki Hut Forum of b.org. They're all in hibernation right now so go ahead and finish this up.

Yeah man, I can't wait.

[Dave A]

1 minute ago, jimjimbo said:

I had thought that you were also considering using a calibrated mic and REW or some such thing.  I understand you were looking for input on that.  I don't think capacitance and ESR values are the question here, but please correct me if I'm wrong.

That is what started this whole thread was the question of can an audio analysis of the effect each different type of cap you use be done with a UMike and REW or TrueRTA. I wanted to know if that could be done and I have no idea how to set up and do just that. I was hoping for an answer there but so far no answer.