ESR Meter confusion

If it's a quick repair, it's almost always a power supply capacitor so you can just tack a similar part in parallel and see if it starts working again.

Crude but effective, and young players don't need to worry about blowing their scopes up by mixing mains and earth connections through the scope probe.

--
Spehro Pefhany

I compared the results with three different waveforms: sine, square and triangle. Here is the readme file:

Nov 07, 2018

Two waveforms to measure capacitor ESR have been proposed. John Larkin uses a triangle wave:

w2aew uses a square wave:

I decided to investigate the behavior of these waveforms with different component values, and included a sine wave for comparison.

Triangle Wave

The triangle wave changes shape and amplitude drastically with different component values. This makes it difficult to measure anything.

Square Wave

The square wave produces pulses with variable width depending on the rise and fall times. w2aew attributes these to reflections on the test leads, but he is not using fast enough signals to produce reflections. The pulses can make it difficult to measure ESR, but you can distinguish between a good cap and a bad one.

Sine Wave

The sine wave always produces a sine wave result. The amplitude and phase shift vary with different component values, but you can easily distinguish between a good cap and a bad one.

I applied a square wave from a 50 ohm function generator; that should be obvious from the picture and the note. The capacitance of the DUT integrated the square wave current into a triangle; the slopes can be used to calculate capacitance. The ESR of the DUT produced the visible step and quantified its value. The small spike is from ESL.

My measured value of ESR is about 75% of the specified max for that part. Some caps have typical ESRs that are a small fraction of the specified max, which is good to know sometimes. Many voltage regulators are stable only over some range of ESR.

Often ESR is good.

He could have easily made a proper 4-wire connection to that cap. His test leads add ESR and ESL errors. If he had adjusted the frequency, he could have better separated the capacitive and resistive components.

I suppose if you are repairing old audio gear this sort of finding a maybe bad cap is helpful, but you can also parallel a suspected cap without de-soldering it.

As an electronic designer, not a repair tech, I want hard numbers about, usually, new parts. My square wave thing separately quantifies capacitance, ESR, and ESL.

A "bad cap" isn't a cap with high ESR. It's a cap that doesn't meet its datasheet ESR specs. A polymer aluminum with 25 mohms of ESR could be bad, and a wet aluminum with 500 could be fine.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Assuming no probe explosion, one can also scope the ripple.

Either way, the suspected cap doesn't have to be removed from the board.

As I tell my kids, measure and think; don't solder.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

These do not indicate that you used a square wave.

What was the value of the cap and ESR? What was the frequency and risetime of the generator?

I think the spike is from the rise and fall time os the generator, not the ESL of the cap and lead inductance.

You can put the same values into the LTspice file and compare the results.

Get a proper LCR meter. The TH2821B is under $200 on eBay. The TH26027 Kelvin test cable is $39.00. It will give you numbers that you can write down instead of interpreting from a scope. You do not have to make the measurements. Your secretary could do it.

Unless of course the original cap has failed leaky or short.

-- This message may be freely reproduced without limit or charge only via the Usenet protocol. Reproduction in whole or part through other protocols, whether for profit or not, is conditional upon a charge of GBP10.00 per reproduction. Publication in this manner via non-Usenet protocols constitutes acceptance of this condition.

I can't see any other way to interpret that image.

the note says it's a step.

if it was an isolated step the left side would be flat, as it's sloping it's obviously showing the end of some sort of pulse

Knowing John, probably something "insane" not that speed is particularly important with 50 ohms in series with the capacitor, The bit that shows the ESR is the little step. The presence of the step (and the overshoot) shows that the generator speed was sufficient.

In the step the impedance of the FG and the ESR,ESL of the capacitor are in series forming a voltage divider

--
  When I tried casting out nines I made a hash of it.

I thought it was obvious. I guess I could have explained it more. But ESR is dv/di at the step, and the post-it says that there was a 100 mA step.

The post-it shows the capacitance and ESR. I don't recall the frequency or rise time, but to measure ESR they don't matter much. The amplitude of the step is all you need to calculate ESR.

I used my B+K 4003A analog function generator. I like that fungen; it has knobs you can turn.

As you note, the rise time doesn't need to be super fast... just enough to make the obvious step at mid-screen.

If the frequency is high enough to make a nice triangle, the voltage drop across the cap will be small, so you can also treat the square wave generator as a current source, if you want.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

PS - the scope shows the frequency.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

On Nov 6, 2018, snipped-for-privacy@nospam.com wrote (in article):

Some handheld LCR meters measure ESR correctly, and can also handle low-Q inductors. Here is one such:

.

I have one of these, and they work quite well. Beware that they really don?t like it if the 9-volt battery is put in backwards.

Joe Gwinn

Get one of those "universal" testers some available n kit form; designations like "Transistor Tester Diode ESR MOS PNP NPN L/C/R". What the heck is wrong with Amazon, they are an order more reliable than E-bay where either a buyer or a seller can get screwed royally and ZERO recourse.

They're evil, but they're really good at it. I rarely deal with Amazon any more.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

I average about 3 Amazon deliveries per week. Test equipment, parts, whiteboard markers, light bulbs, shoes, jeans, chocolates, cookies, mini-fridges, cold plates, pumps, hardware. That saves a huge amount of driving and parking and standing in checkout lines.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Hmmm...the same way capacitance varies with voltage. V and I are swapped not only in the equations, but even in the other behaviors.

We just ordered a small optical breadboard plate. Took about 30 seconds to find it and one-click. No POs or any of that nonsense, guaranteed 2 day delivery.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Any particular maker and model number of multimeter? If you disclose that information, it might be possible to look at the specs and determine what it's really doing inside.

Ok, you're an impoverished hobbyist that needs test the ESR of caps for some reason. The cheap M328 Mega328 or LCR-T4 testers are probably good enough. There's project page somewhere on the web that I can't seem to find. Most such cheap components testers do not have an ESR zeroing feature[1], which makes the accuracy at the low end of the resistance scale a problem. Try to get one that measures at

100KHz, which is the defacto frequency for ESR specs. I make the mistake of buying the absolute cheapest M328 kits. Later versions have more features, nicer display, fewer bugs, and possibly better accuracy. I've compared the reading with 3 other ESR meters that I've accumulated. The readings are fairly close above 5 ohms, and not very good below 5 ohms.

One of the fun problem is that the ESR of electrolytics changes with temperature. I ran a few tests to see how much and ran into a problem with the ESR meter: A "good" electrolytic looks like this: A "bulging" electrolytic looks like this: Note that it would be considered bad due to high ESR at 25C, but might pass as good above 50C.

Cap tester = Measures capacitance ESR tester = Measures ESR usually at 100Khz Cap analyzer = Measures everything at various frequencies.

I assume that you're looking at the M328 or LCR-T4 collections on eBay. Yes, they're a disorganized mess. The hard part is finding one that has the latest firmware and does what you want. I could probably find a likely prospect, but I don't know anything of how you plan to use it.

Well, the LCR-T4 is a later version of the M328. However, the vendors seem to be using the model numbers interchangeably, often on the same item, so you have to look at the photo to determine which one is which. Some have color displays:

Gotta run. More later...

[1] Short the leads, push the button, and the display is set to zero ohms ESR.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

On Thu, 08 Nov 2018 13:13:48 -0800, Jeff Liebermann wrote: (...)

More on the M328 components tester:

TransistorTester with AVR microcontroller and a little more. Version 1.13k Mar 8, 2018 132 pages Includes photos and a tour of most of the Chinese clones.

Links to drawing, schematic, source code, version, etc.

Latest firmware:

Waveforms and photos:

Want to Build an Arduino-Based Component Tester?

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

** Yes, hot electrolyte is more conductive - often several time more than when tested at room temp.

Also, readings taken at 100kHz so are higher than at 100/120Hz, again by up to several times.

ESR meter readings are comparative tests, intended to identify examples that are below spec or have deteriorated.

.... Phil

Thanks for that down to earth input to this discussion! :)