ESR Meter Questions

[gowf67]

I'm curious to pick one up for some cap and battery testing but as always there's folks who think a particular product is snake oil.

In this thread there's a gentleman who seems to know what he's talking about but it's all just left me confused. Any ideas and apologies if this has been beaten to death already.

His points to summarize:

<sub>I'm sure the so-called "ESR" meters have some use, since a few people on the internets swear by them. I've never figured out what that real use is, and the explanations offered by the companies making kits are pretty bogus.

All "pros" I know use proper capacitance meters, which also measure ESR, and give consistent and accurate results. These used to be expensive, so there was at least some motive to buy a cheap "ESR" meter just for troubleshooting. However, now that you can get a decent cap meter for $30, or so, I would ignore the ESR things. They are primitive, inconsistent and don't measure what they claim to.

-k

Sorry if I was brusque... the Internet has a way of doing that to me. I do understand how capacitors are constructed, specified, measured and used. And I stand by what I said about "ESR" meters. Refering to the type of "ESR" meters sold often in kit form for troubleshooting and repair jobs, I've tested a
few of them and examined the circuits of others. I don't think they are either accurate or cost effective, and most don't really measure what users
think they do. Sure, they can help to locate open or shorted caps, but that's about all. Some specifics,

1- The very first and cheapest of these attempted to derive ESR from charge/discharge time constants. This approach has been largely abandoned as inacurrate and inconsistent.

2- The ESR meters we are talking about probe the impedance at some fixed high frequency, usually about 100KHz, and report that as if it were ESR.
But what if it is Inductance? What if it is the self-resonance in a small lytic. What if Z(C) > Z(R), at the test frequency? In all cases, tough luck, you get the wrong answer.

3- Typical decent capacitance meters use a balanced bridge method and a seperate pulsed probe signal for ESR. I just checked, and plenty are available in the $30 range on Ebay. Just search "LCR Meter," then sort by price. This kind of meter will be vastly more useful in the bench.

4- There's a cap meter function on my handheld Fluke, but it is rather poor, and also doesn't report series impedance.

5- Lab-grade instruments combine more complex bridge topologies with complex algorithms to determine the exact series impedance and equivalent circuit model. They are always 4-wire devices.

6- By convention, and to avoid the complexities of properly defining ESR, modern cap meters mostly report DF, (Damping Factor). A simple calculation
will give you an ESR value.

7- Just FYI, I use two Genrad (now IET) Digi-Bridges as references, along with a custom network analyzer built around an AP System One DSP. (Plus a 1 MHz Digi-Bridge for non-audio cap testing.) I also have hand-held B&K LCR meter that is very nice for the price.

Having just revisited the ESR measurement issue in another context earlier this year, I am happy to discuss/debate the topic as long as it would be
useful.

-k</sub>


http://www.audiokarma.org/forums/showthread.php?t=235016

[Richard Crowley]

What is the context here?  WHY are you wanting to measure ESR?
Perhaps if we had some background here it would help all of us understand what may (or not) be important.

[gowf67]

Ala Dave did in his one of his monitor repair videos, filter caps and battery testing to see if one of the aforementioned should remain in service or replaced; and for learning.

[commie]

What is the context here?  WHY are you wanting to measure ESR?
Perhaps if we had some background here it would help all of us understand what may (or not) be important.

Well, effective series resistance(ESR), is simply the ac resistance which exists between the two plates of every practical capacitor at a particular frequency, for the perfect capacitor it's ESR would be zero. So, if you have an ESR or LCR meter handy, you can measure the ac resistance (ESR) of more or less any type of capacitor and what you are looking for, is the lowest ESR. Perfect capacitors and inductors do not consume energy they simply store it.

Cheers
Commie

[PedroDaGr8]

What is the context here?  WHY are you wanting to measure ESR?
Perhaps if we had some background here it would help all of us understand what may (or not) be important.

Well, effective series resistance(ESR), is simply the ac resistance which exists between the two plates of every practical capacitor at a particular frequency, for the perfect capacitor it's ESR would be zero. So, if you have an ESR or LCR meter handy, you can measure the ac resistance (ESR) of more or less any type of capacitor and what you are looking for, is the lowest ESR. Perfect capacitors and inductors do not consume energy they simply store it.

Cheers
Commie

I don't think Richard was asking for a definition of ESR. I think what he was asking is WHY does OP want to know ESR, how will he use the meter, etc. This way any advice given can prioritize things he needs and ignore things he doesn't.

[measurementpoint]

I'm sorry, so much has been written that the author himself confused. This topic has all the characteristics of capacitors?
This topic ESR?
Known methods for me:
1) Industrial LCR-meter;
2) Analog probe;
3) LED probe;
4) LCR-meter on the ATmega.

1 Important! Your business and your money. Buy it more expensive or cheaper. Dear unit will be slightly better display and a robust housing.

2 Important! All theories about the accuracy of the measurement can be left alone. Because the quantity will depend on 1) capacitance and 2) its voltage. You have already received an entire table ... Do not have a headache. Headache here - add here 3) various parameters will be from different manufacturers, and 4) of different parties capacitors.
No headache?
Add the item 5 - capacitors discharge! The discharge must!
I hope not confused in the calculations? Smile.
Successful measurement!

This suggests that the best method would be a comparison, rather than the absolute accuracy of the method.

[PedroDaGr8]

Well, effective series resistance(ESR), is simply the ac resistance which exists between the two plates of every practical capacitor at a particular frequency, for the perfect capacitor it's ESR would be zero. So, if you have an ESR or LCR meter handy, you can measure the ac resistance (ESR) of more or less any type of capacitor and what you are looking for, is the lowest ESR. Perfect capacitors and inductors do not consume energy they simply store it.

Cheers
Commie
[/quote]

I don't think Richard was asking for a definition of ESR. I think what he was asking is WHY does OP want to know ESR, how will he use the meter, etc. This way any advice given can prioritize things he needs and ignore things he doesn't.

As for OP, there are two directions that you can go with this. You can go the bargain basement route which involves an open-sourced device designed by a couple of people that post on this forum. This is the device you often see running around $20-30. It will test ESR, plus a few other things, as well as do component ID. These are relatively capable devices with active development and support. The intent is to be a very simple hobbyist level device and they succeed at that.

The next step up takes you to the ~$100 range and is a huge step-up in quality. This device is the Der EE DE-5000. It is a true LCR in every sense of the word, with user selectable frequencies, parallel/series modes, ESR, tan D/D/DF/etc. can measure inductance, low resistances, etc. Many of us on here have one of these and these are real deal LCR meters. They are the same as the IET Labs DE-5000 and DE-6000, Der EE is the OEM for IET Labs.

[gowf67]

I'm sorry I realize now there is too much in the original post.

This is the main point I'm confused about:

2- The ESR meters we are talking about probe the impedance at some fixed high frequency, usually about 100KHz, and report that as if it were ESR.
But what if it is Inductance? What if it is the self-resonance in a small lytic. What if Z(C) > Z(R), at the test frequency? In all cases, tough luck, you get the wrong answer.

To me this seems like he is saying ESR meters, like Bob Parker's design, are of no use. That is what I'm trying to clarify in the context of checking for bad/failing capacitors and/or batteries.   

[measurementpoint]

I've already written - do not expect absolute precision and do not pursue it.
Therefore, paragraph 6 - run on different frequencies.

[commie]

I don't think Richard was asking for a definition of ESR. I think what he was asking is WHY does OP want to know ESR, how will he use the meter, etc. This way any advice given can prioritize things he needs and ignore things he doesn't.

Thanks for pointing that out, actually I had figured this might be the case, sorry Richard just ignore me.

 

[unitedatoms]

I'm sorry I realize now there is too much in the original post.

This is the main point I'm confused about:

2- The ESR meters we are talking about probe the impedance at some fixed high frequency, usually about 100KHz, and report that as if it were ESR.
But what if it is Inductance? What if it is the self-resonance in a small lytic. What if Z(C) > Z(R), at the test frequency? In all cases, tough luck, you get the wrong answer.

To me this seems like he is saying ESR meters, like Bob Parker's design, are of no use. That is what I'm trying to clarify in the context of checking for bad/failing capacitors and/or batteries.   

The point of the sentence 2- is that single value (number reading in scalar or complex form) does not characterize device under test completely enough. There is no way to tell how complete equivalent circuit looks like if you just have a single point measurement at single frequency.

In other words Network Analyzer > LCR Meter.

1. Simple LCR Meter will assume network of 1 component L, C or R <- these are too basic or highly specialized. No use for repairs
2. Typical LCR Meter will assume network of 2 components: reactive and resistive <- this is optimal cost/complexity wise for repairs
3. Advanced LCR Meter will assume network of 3 components for some measurements: 2 reactive and 1 resistive, like for example leakage inductance of transformer together with ESR and distributed capacitance <- may be too much overkill for simple repairs
4. Full network analyzer: Complex impedance spectrometry of linear circuits of any complexity (transformers, speakers charting, filters, crystals etc.), transfer characteristics measurement <- overkill
5. Universal testers with advanced software analysis to do something else: to run experiments on varactors for example or bio samples measurement or whatever

TLDR: For batteries, and bad caps 2 component equivalency is enough. You can tell cap from LC tank by eyeballing in say in Audio.

[measurementpoint]

Bob Parker - you're talking about numbers or a linear mapping diodes?
If there is no objective research, only to reject the capacitor - it works!
It does not work in the one who needs to sell for 1000.
Catch up passions about security.
Nobody violates safety! Driving previously discharged!

[Richard Crowley]

As with most things in the Real World, there are different ways of measuring ESR, and some of them may not be relevant, accurate, or even possible for different cases.  Measuring the ESR of a giant backup battery cell the size of a small car will be different than measuring the ESR of the dielectric of a microscopic virtual component in an integrated circuit.

I remember back when we were getting strange results from measurement of the gate oxide capacitance of the transistors in the memory cells of the latest DRAM product. We had always measured this capacitance with the classic Boonton 72 meter. But it turned out that the structures we were now making evolved into the resonance range of that fixed-frequency meter.  So we had to change our strategy and start using more sophisticated instruments to accurately measure gate capacitance.  The point being that you must select the measurement METHOD to suit not only WHAT you are trying to measure, but WHY you want to measure it and HOW you are going to use the information.

[gowf67]

For me it is simply a matter of go/no go for a component. It seems the general consensus is that these type of ESR meters ~$80-100 (Bob Parker designs) are fine for that. Just don't expect precision results because that is not what they are designed for.

[measurementpoint]

Yes / no question - of course, the probe can be solved.
If you have not used before - every meter needs to adapt - but important - to defuse the scheme!

[commie]

I've already written - do not expect absolute precision and do not pursue it.

Hey, I like this statement absolute accuracy is unobtainable, Gauss knew about this before he was 18 years old.

[gowf67]

Thanks everyone for taking the time to respond, I guess I'm back to one more point in question:

- But what if it is Inductance? What if it is the self-resonance in a small lytic. What if Z(C) > Z(R), at the test frequency? In all cases, tough luck, you get the wrong answer.

Is he saying the meter will return an inductance or self-resonance value of the cap under test and mischaracterize it as an ESR reading? 

[unitedatoms]

I've already written - do not expect absolute precision and do not pursue it.

Hey, I like this statement absolute accuracy is unobtainable, Gauss knew about this before he was 18 years old.

Lord Kelvin:
"There is nothing new to be discovered in physics now, All that remains is more and more precise measurement."

[measurementpoint]

Hey, I like this statement absolute accuracy is unobtainable, Gauss knew about this before he was 18 years old.

[measurementpoint]

Author, any inductance in the capacitor?
The principle is that if the resistor capacitor appeared - and it breaks all the fluctuations. And if it's not there ... You're in luck! And cheaper!

Forcing purchase. And what cost - is also written above - depends on the tasks.
 

[PedroDaGr8]

For me it is simply a matter of go/no go for a component. It seems the general consensus is that these type of ESR meters ~$80-100 (Bob Parker designs) are fine for that. Just don't expect precision results because that is not what they are designed for.

If you are paying $80-100 for the Bob Parker designed meters you are getting ripped off. The $20 LCR meters that were mentioned are significantly better quality. The Bob Parker meter was revolutionary when it was designed and made because NOTHING out there was in the sub-$100 price range for measuring ESR. These days, technology has advanced and you are MUCH MUCH MUCH better off with either the $20 LCR meter or if you don't mind stepping up to $100-ish, the Der EE DE-5000. The DE-5000 just demolishes the Bob Parker designed meters in accuracy, features, reliability and specs. It isn't even close.

[gowf67]

This guys tries everything we've talked about.

https://youtu.be/ESijVVY_GjA




Thanks guys I think I'll go for the 5000.

[markce]

I got the DE5000 as my main LC meter, and build an analog  in-circuit ESR tester for
first checks of electrolitic capacitors. It's cheaper to repair and better protected. A good combo,
don't need more.

[gowf67]

goodness gracious there are so many options on ebay for this model leads/no-leads meter only package  any recommedations        etc.. I'm just as confused when I started   

[measurementpoint]

Let's not forget that there were statements on a particular topic. Sooner or later, will be understanding - accuracy / visibility. Difference!
Even if the accuracy is in thousandths. These are just numbers. What do they tell you? It's nothing.
When ranking the same batch of capacitors - visibility play a positive role.
When working with the full range of manufacturers, and work schemes are unknown to you ... I was always amazed absurdity of striving to measure exactly - the object is not accurate. You can continue. Or you can think about.
As well as in different countries - is very different prices for certain devices. The choice is the master of your money.