Do I misunderstand ESR measurement? Frequency dependence of ESR

[joeqsmith]

I don't think anyone would argue that Gramps old TE wasn't useful and maybe even considered elegant in its day.  Like old analog scopes and meters, the old timers will defend them till their death (sounds like G0HZU may be getting close to that stage).   I gave up analog meters in the 80's and gave away my analog scope in the 90s.  I'm sure not going to recommend someone try and use them today.

Don't mess with custom spread sheets and software to run all these calculations.  Get your DER EE and get on with the actual project at hand.   

[joeqsmith]

Read the thread again. I offered this as an elegant way to do a one off sanity check of a suspect reading from a low cost hobby meter. I never suggested this could replace the meter for everyday use. Just check a reading of ESR in the ballpark of 0.1R for a 1uF WIMA cap at 10 kHz.

So if someone has a dodgy reading from a cheapo ESR meter and they don't have another ESR meter (but they do have a decent DVM and function generator) they can so a sanity check of the reading.

If someone is getting bad reads from their cheapo meter,   I sure wouldn't suggest compounding that by setting up various equipment (that they may not have) and putting together a spreadsheet or custom application like you show to run the calculations.   Certainly not for a one off.  Just buy a better low cost RLC meter.  Most places have them in stock.   

I wonder at what point did you decide you needed to automate your setup rather than just buying a meter?  I mean, how hard up are you?  The DE-5000 is under $140 and would last you a life time (in your current state, what's that?  another month?) 

[G0HZU]

joeqsmith: I wonder at what point did you decide you needed to automate your setup rather than just buying a meter?  I mean, how hard up are you?

My background is in RF engineering. I'm a senior RF Design Consultant at a major defence company having designed ESM and ECM equipment for about 35 years. Therefore, I do OK for test equipment in this area. Much better than most so no, I'm not hard up.

joeqsmith: The DE-5000 is under $140 and would last you a life time (in your current state, what's that?  another month?)

That's pretty nasty from you. None of us know for sure if we will live another day let alone a month.

[joeqsmith]

joeqsmith: The DE-5000 is under $140 and would last you a life time (in your current state, what's that?  another month?)

That's pretty nasty from you. None of us know for sure if we will live another day let alone a month.

You want to talk smack to people but get upset when others treat you the same way?  My advice if you don't like it, then treat people better. 

[G0HZU]

joeqsmith: The DE-5000 is under $140 and would last you a life time (in your current state, what's that?  another month?)

That's pretty nasty from you. None of us know for sure if we will live another day let alone a month.

You want to talk smack to people but get upset when others treat you the same way?  My advice if you don't like it, then treat people better.

Mocking the life expectancy of someone who may be close to death is a bit different to pointing out someone's obvious incompetence on a technical forum.

[joeqsmith]

joeqsmith: I wonder at what point did you decide you needed to automate your setup rather than just buying a meter?  I mean, how hard up are you?

My background is in RF engineering. I'm a senior RF Design Consultant at a major defence company having designed ESM and ECM equipment for about 35 years. Therefore, I do OK for test equipment in this area. Much better than most so no, I'm not hard up.

You mentioned that before with your uninformed torque wrench blabs.  Odd for someone with 35 years in RF,  that would even be a discussion.  When I started pushing you for details about your use of them at work, you took your ball and walked away.   On the plus side, I doubt you have to deal with us lost causes much longer.   

[mawyatt]

At low frequencies (<300kHz) and with reasonably large capacitors, (>100nF?) it is possible to measure ESR quite well using a clean signal source and a known series sense resistor in series with the DUT capacitor with the far end of the DUT cap grounded. Then measure the AC voltages around the sense resistor using a decent (true RMS reading) DVM.

We must give credit where credit is due. Our DER-5000 only goes to 100KHz, yet GOHZU claims the "elegant" method works up to 300KHz.

How about providing some "elegant" measurements around 300KHz GOHZU?? Then we'll see if anyone can reproduce these!!

Maybe your self-proclaimed "elegant" method works better at 300KHz than 10KHz, as seems no one has been able to reproduce your "elegant" results at 10KHz yet!!

Maybe the GOHZU self-proclaimed "elegant" method works everywhere, but we haven't been able to get reliable consistent results at 10KHz, and we've tried with many different capacitors and even different DMMs. Maybe our source isn't good enough (SDG2042X)??

Best

[joeqsmith]

joeqsmith: The DE-5000 is under $140 and would last you a life time (in your current state, what's that?  another month?)

That's pretty nasty from you. None of us know for sure if we will live another day let alone a month.

You want to talk smack to people but get upset when others treat you the same way?  My advice if you don't like it, then treat people better.

Mocking the life expectancy of someone who may be close to death is a bit different to pointing out someone's obvious incompetence on a technical forum.

To me its all the same.   If you want to be treated like shit, act like an ass.  Interesting enough, I watched a video the other day from a hospice  worker (end of life care giver).  They were talking about how as people get closer to death, they act more like themselves (no filter).   They were saying how bad people with narcissistic  behaviors would get.  They were addressing kids who may be trying to take care of their parents who are reaching end of life.  You are certainly in rare form lately.  Maybe we are seeing the result of what the hospice nurse was describing.   

***

I know I'm completely wasting my time but ....

Just to add,  you have posted several times about us being a lost cause and wasting your time.  Do you really want to waste your moments being a keyboard warrior?  You can, but don't expect better treatment from me as long as you continue to lash out at members.

[joeqsmith]

...

To be fair, I did post some good results when using their elegant method with my custom replicated software and hand selected resistors.  I just needed to run lower quality parts.  Still, even if I was trying to measure ESR of my alum elec, this seemed like a really dumb and obscure way to get the job done vs just buying a modern low cost meter.   I can't believe anyone would waste time automating it.

[G0HZU]

If it helps, I just tried swapping between a 10R sense resistor and a 15R resistor to measure the same 1uF WIMA cap at 10kHz. This gives three different voltages because the sense resistor is changed.

But it still measured 0.15R ESR in both cases to within about +/- 0.005R if I tried measuring again and again and finding the most extreme results. I could try adding more averaging to improve this but I never made any claims about accuracy. I just suggested it as a means to do a sanity check of a measurement made with a cheapo ESR meter.

It also measured the capacitance with excellent agreement something like 1.041uF in both cases. I couldn't be bothered to see how (really) close the two results agreed in this respect but the first 4 digits were 1.041 for the capacitance in both cases.

[mawyatt]

...

To be fair, I did post some good results when using their elegant method with my custom replicated software and hand selected resistors.  I just needed to run lower quality parts.  Still, even if I was trying to measure ESR of my alum elec, this seemed like a really dumb and obscure way to get the job done vs just buying a modern low cost meter.   I can't believe anyone would waste time automating it.

Electrolytics were about the only capacitors we measured that were even in the same ballpark as what these caps actually measured on a quality LCR meter (IM3536). Any reasonably high "Q" film caps were hopeless.

It just boils down to the measurement physics, and the uncertainty/sensitivity gets extremely large for high Q capacitors, which is why we were interested in the 1st place!! We were actually hoping this would prove out, as we hadn't heard about this technique before and admitted such, but now we know why!!

Why anyone would claim this technique as "elegant" is beyond reason, and then defiantly defend and attack those that questioned this "elegant" solution is ridiculous, but seems to fit the mold!!

Best

[RoGeorge]

I think I have an idea! 

To make the 3 scalar voltage measurements more precise, instead of measuring one float voltage (and 2 other voltages relative to ground), we can make all 3 measurements relative to GND, if instead of moving the AC voltmeter, we flip the series RC ensemble.  I mean, connect the generator, the known R, and the C to test, all in series, let's say, with R at the generator's GND at first, then:
- measure the Vgenerator (AC voltmeter between GND and Vout generator)
- measure the Vresistor (AC voltmeter between GND and the middle point between R and C)
- now flip the series RC ensemble with its capacitor side to the GND of the generator, and measure Vcapacitor (AC voltmeter remains connected the same, between GND and and the middle point between R and C)

Notice how, in all 3 measurements, the voltmeter measures relative to GND, which (I assume, but I didn't try) should give more precise measurements than when measure a floating AC voltage, particularly at high frequency (no need for a floating voltmeter any more, can be used a shielded cable instead of normal probes).

[G0HZU]

...

To be fair, I did post some good results when using their elegant method with my custom replicated software and hand selected resistors.  I just needed to run lower quality parts.  Still, even if I was trying to measure ESR of my alum elec, this seemed like a really dumb and obscure way to get the job done vs just buying a modern low cost meter.   I can't believe anyone would waste time automating it.

Electrolytics were about the only capacitors we measured that were even in the same ballpark as what these caps actually measured on a quality LCR meter (IM3536). Any reasonably high "Q" film caps were hopeless.

It just boils down to the measurement physics, and the uncertainty/sensitivity gets extremely large for high Q capacitors, which is why we were interested in the 1st place!! We were actually hoping this would prove out, as we hadn't heard about this technique before and admitted such, but now we know why!!

Why anyone would claim this technique as "elegant" is beyond reason, and then defiantly defend and attack those that questioned this "elegant" solution is ridiculous, but seems to fit the mold!!

Best

How many times do I have to tell you that the aim was to do a sanity check on a basic 1uF WIMA cap with an ESR in the ballpark of 0.1R. I never made any claims about accuracy or if you could use it for calibration/checking of an ESR meter for really low ESR caps.

Why anyone would claim this technique as "elegant" is beyond reason,

It is a simple method that doesn't require test gear that can measure phase accurately. Just a clean function generator and a decent AC DVM and a sense resistor. I thought it might be educational too in that it's possible to make vector based measurements using three scalar measurements. Something you claimed wasn't possible. At least you did eventually admit you were wrong in this respect

[joeqsmith]

If it helps, I just tried swapping between a 10R sense resistor and a 15R resistor to measure the same 1uF WIMA cap at 10kHz. This gives three different voltages because the sense resistor is changed.

But it still measured 0.15R ESR in both cases to within about +/- 0.005R if I tried measuring again and again and finding the most extreme results. I could try adding more averaging to improve this but I never made any claims about accuracy. I just suggested it as a means to do a sanity check of a measurement made with a cheapo ESR meter.

It also measured the capacitance with excellent agreement something like 1.041uF in both cases. I couldn't be bothered to see how (really) close the two results agreed in this respect but the first 4 digits were 1.041 for the capacitance in both cases.

Should that be a surprise?   As I relaxed the parts I was trying to test when evaluating your claims, not only were the results reasonable,  the effort taking the measurements diminished.   I wasn't waiting for things to stabilize,  choice of other parts were not critical, even my Marconi RF generator was fine.  But, that was never the point.  For the effort involved, I was hoping to see better results than my cheapo RLC meter as things like Q didn't matter and I knew my low cost BK couldn't measure those parts. 

My background is in RF engineering. I'm a senior RF Design Consultant at a major defence company having designed ESM and ECM equipment for about 35 years.

Sorry but I don't understand your motive of posting this.   I assume you feel your career places you above everyone else's life long accomplishments, and that everyone should blindly follow you without question.  I am used to EEs questioning everything.   

[G0HZU]

Sorry but I don't understand your motive of posting this.

You questioned my ability to afford $140 for a DE-5000. (you asked how hard up I was along with mocking my life expectancy) So I told you what I did for a living. I can definitely afford a $140 ESR meter based on my job/income although I only work part time now. I have plenty of items of high end test gear here already. nearly all of it is biased towards RF.

If you still can't grasp this find a nearby grown up to read and explain what I wrote to you in post #252 about my job and income in response you your question about me being poor or hard up. However, Joe used his flawed reasoning to misinterpret what I said and to use it as a basis for another attack as below:

I assume you feel your career places you above everyone else's life long accomplishments, and that everyone should blindly follow you without question.  I am used to EEs questioning everything.

You definitely do have a problem with your reasoning here.

[G0HZU]

Should that be a surprise?   As I relaxed the parts I was trying to test when evaluating your claims, not only were the results reasonable,  the effort taking the measurements diminished. 

Fine. We agree that this method can give good results testing a 1uF WIMA cap with an ESR in the ballpark of 0.1R at 10 kHz.

Please explain this to Mawyatt and then we can all go home.

[joeqsmith]

Sorry but I don't understand your motive of posting this.

You questioned my ability to afford $140 for a DE-5000. (you asked how hard up I was along with mocking my life expectancy) So I told you what I did for a living. I can definitely afford a $140 ESR meter based on my job/income although I only work part time now. I have plenty of items of high end test gear here already. nearly all of it is biased towards RF.

I've known people making very good money in their careers living paycheck to paycheck and having a difficult time making ends meet.  While others with blue collar jobs have done very well.  It seem less tied to how they earn their income, and more about how savvy they are with their personal finances.   

You could have just stated you could afford $140. 

Edit: Hopefully others will give this 3 Scalar Technique described by GOHZU a try, we've tried and tried to make it work, but the results continue to be disappointing. It seems the technique is highly sensitive to slight uncertainty in measurement, which is aggravated by higher "Q" capacitors, and unfortunate as high "Q" components are where it could be beneficial. 

Because you asked, and the fact G0HZU would actually waste time creating a custom spreadsheet, writing custom software, make a custom jig, and to top it off, stand on their soap box blasting everyone who dared question why, was my motive to see if there were any gains to be had.  I've seen G0HZU act this way in similar threads and I don't mind running a few experiments to see if their expert opinion actually carries any weight.  I guess I wasn't surprised.     

[G0HZU]

Sorry but I don't understand your motive of posting this.

You questioned my ability to afford $140 for a DE-5000. (you asked how hard up I was along with mocking my life expectancy) So I told you what I did for a living. I can definitely afford a $140 ESR meter based on my job/income although I only work part time now. I have plenty of items of high end test gear here already. nearly all of it is biased towards RF.

I've known people making very good money in their careers living paycheck to paycheck and having a difficult time making ends meet.  While others with blue collar jobs have done very well.  It seem less tied to how they earn their income, and more about how savvy they are with their personal finances.   

You could have just stated you could afford $140. 

Edit: Hopefully others will give this 3 Scalar Technique described by GOHZU a try, we've tried and tried to make it work, but the results continue to be disappointing. It seems the technique is highly sensitive to slight uncertainty in measurement, which is aggravated by higher "Q" capacitors, and unfortunate as high "Q" components are where it could be beneficial. 

Because you asked, and the fact G0HZU would actually waste time creating a custom spreadsheet, writing custom software, make a custom jig, and to top it off, stand on their soap box blasting everyone who dared question why, was my motive to see if there were any gains to be had.  I've seen G0HZU act this way in similar threads and I don't mind running a few experiments to see if their expert opinion actually carries any weight.  I guess I wasn't surprised.   

The 'custom jig' you refer to was made using a piece of scrap FR4 PCB material and I very quickly and crudely cut out the artwork shapes with a scalpel and a soldering iron. There are only two square islands on it. It looks a mess but it works OK. There is no MUX despite what you claimed. It looks a lot worse now because I added an extra socket to it to allow series and parallel measurements.

Writing code to read a multimeter three times doesn't take long in VB6 and I did all this about 10 years ago.

Using a Marconi 2024 to try and measure a 0.004R ESR 1uF cap at 10kHz is just so funny. What were you thinking? 

[joeqsmith]

Should that be a surprise?   As I relaxed the parts I was trying to test when evaluating your claims, not only were the results reasonable,  the effort taking the measurements diminished. 

Fine. We agree that this method can give good results testing a 1uF WIMA cap with an ESR in the ballpark of 0.1R at 10 kHz.

Please explain this to Mawyatt and then we can all go home.

Again, I don't think that was ever a question once other's posted all the details you were not forthcoming with. 

If someone wanted to repeat the this experiment for the novelty of it, that's one thing  but I'm not going to recommend analog scopes and meters just because they can still be used for certain cases.   Like Mawyatt, I was hoping to be able to use it for lower grade parts.   While you can whine about the outcome and suck your thumb raw, it won't change my stance.   

With how you were pounding your chest,  I was really hoping you would step up, procure those same parts and prove everyone wrong.  Instead here we are.   So sure, you made a very expensive, complex, and low performance RLC meter that with some custom software may compete with a modern cheap RLC meter.  We all bow to your superior skills.   Is this how a major defense company operates?

[G0HZU]

Is this how a major defense company operates?

There's definitely something wrong with your reasoning skills based on this question alone...

[joeqsmith]

The 'custom jig' you refer to was made using a piece of scrap FR4 PCB material and I very quickly and crudely cut out the artwork shapes with a scalpel and a soldering iron. There are only two square islands on it. It looks a mess but it works OK. There is no MUX despite what you claimed. It looks a lot worse now because I added an extra socket to it to allow series and parallel measurements.

Writing code to read a multimeter three times doesn't take long in VB6 and I did all this about 10 years ago.

Using a Marconi 2024 to try and measure a 0.004R ESR 1uF cap at 10kHz is just so funny. What were you thinking? 

Oh, you were still manually manipulating the leads.  I was giving you far more credit after looking at your custom software.  I didn't understand why someone with your obvious skills wouldn't just read the meter until the standard deviation was tight enough then automatically select the next channel.

The fact any time was spent on it boggles me.  Out side of some academic reason (which was my motive), I don't see the point. 

The Marconi (and others) was an easy sanity test to run on a few parts.  As I said, for the low quality parts, it and pretty much everything worked fine.  No combination I could come up with using your method worked for the higher quality parts being asked about.   

[joeqsmith]

Is this how a major defense company operates?

There's definitely something wrong with your reasoning skills based on this question alone...

Who am I to question your employers practices.  If they retain you to hear Gramps old stories of how he washed clothes with a rock down by the river, so be it.

[G0HZU]

Is this how a major defense company operates?

There's definitely something wrong with your reasoning skills based on this question alone...

Who am I to question your employers practices.  If they retain you to hear Gramps old stories of how he washed clothes with a rock down by the river, so be it.

You have definitely got a problem.

[joeqsmith]

Is this how a major defense company operates?

There's definitely something wrong with your reasoning skills based on this question alone...

Who am I to question your employers practices.  If they retain you to hear Gramps old stories of how he washed clothes with a rock down by the river, so be it.

You have definitely got a problem.

I'm not the one trolling during my final days.  That's a real problem. 

[EEVblog]

joeqsmith: The DE-5000 is under $140 and would last you a life time (in your current state, what's that?  another month?)

That's pretty nasty from you. None of us know for sure if we will live another day let alone a month.

You want to talk smack to people but get upset when others treat you the same way?  My advice if you don't like it, then treat people better.

WTF Joe, cut it out. Walk away from this thread now.