What an analog ESR meter will and won't do

[Th3_uN1Qu3]

I'm sure you've all seen those analog ESR meters which use a microamp meter (or really, any meter with a needle, mine for example uses a cheap dB meter), and some of you may wonder whether they will be enough for finding bad caps that aren't bloated.

Short answer: Yes.

I have built this one: http://ludens.cl/Electron/esr/esr.html and with a few small modifications it can easily spot differences of 0.1 ohms, the other end of the scale being at roughly 15 ohms. This makes a good all-rounder, and it's suitable for checking high voltage capacitors as well, which generally have a few ohms ESR.

What this type of meter won't do is spot a cap that is low-ESR and not yet bad, just out of its tolerance. It does not have enough resolution for that (unless you've got your hands on a really big meter movement). If you work only on computer stuff with ultra low ESR caps, you're better off buying a digital ESR meter, but i bet 95% of capacitor problems can be diagnosed with one of those analog meters built from scrap parts...

Most of the non-commercial digital ESR meter designs i've seen so far are pretty dodgy, so if you really want a digital one, i'd say buy it. Or wait for me to design one.

[brethin]

Re: What an analog ESR meter will and won't do

I would prefer a meter with a range of .001 to say 5 ohms for it to be useful to me.

[Longbow]

Re: What an analog ESR meter will and won't do

I'd say the above information is misleading at best. The Blue meter is the only ESR meter that comes to mind that gives a 2 digit readout below 1 ohm. There may be more accurate digital readout meters, but the cost would be prohibitive and unnecessary. Is the Blue meter precise to the 100th of an ohm? I don't know and it doesn't matter anyway. People incorrectly assume that a capacitor's ESR must closely equal the manufacturer's claimed value, or else it is faulty. In fact, the ESR value can often be 5 to 10 times the advertised value before it affects circuit operation. As with most test equipment, an experienced technician is required.

For the most part, the equipment is a troubleshooting aid and not a laboratory calibration standard. Analog meters have one advantage - you can get an instant relative measure and do not have to concern yourself with actual numbers. Therefore, an analog meter such as the Capacitor Wizard has that advantage.

In fact, the one thing an ESR meter cannot do is detect a shorted capacitor, since it will read a low value on the meter, and look like a perfectly good ESR reading.

[Th3_uN1Qu3]

Re: What an analog ESR meter will and won't do

A design i have seen even included shorted cap detection. I've never gotten that particular one to work though. You DO have a continuity function on your multimeter though, don't you.

In fact, the ESR value can often be 5 to 10 times the advertised value before it affects circuit operation.

Exactly. And this is something that a simple analog meter WILL detect reliably.

[brethin]

Re: What an analog ESR meter will and won't do

I did say prefer, and if caps with ESR at that ultra low range didnt matter beyond the 100th then why even bother having specs for them to the 1000th. My guess is because sometimes it does matter.

[ipman]

Re: What an analog ESR meter will and won't do

There is an ESR-meter made by somebody here in Romania which has 2 decimals.
It's spot-on compared with a Fluke 189, which also can measure with 2 decimals.
Not bad for a device that costs around 30 euros, but still not as stable and calibrated against standards like the Fluke.
Here's a picture measuring a (still) good low-ESR 470uF/25V cap, Hermei LL series.

To add some on-topic: with a digital ESR meter you can choose the best piece in a batch of caps.

[japlytic]

Re: What an analog ESR meter will and won't do

As with voltmeters, I prefer the digital ones, since they are easier to read (no parallax error).

[Th3_uN1Qu3]

Re: What an analog ESR meter will and won't do

As with voltmeters, I prefer the digital ones, since they are easier to read (no parallax error).

Me too. I don't even own an analog multimeter... But you do need all those fancy goodies like TrueRMS to replace a good analog meter. An analog will not lie regardless of the waveform of the voltage it is measuring. For a digital one to do the same you need some extras.

@ ipman: Isn't that the one from x-roast.home.ro? I wanted one myself but i wasn't sure how well it did.

[Longbow]

Re: What an analog ESR meter will and won't do

I did say prefer, and if caps with ESR at that ultra low range didnt matter beyond the 100th then why even bother having specs for them to the 1000th. My guess is because sometimes it does matter.

It was the original question that caught my eye, namely what an ESR will and won't do. Apparently a lot of people expect them to do everything. But it is a fact that a piece of test equipment no matter how wonderful is only as good as the person operating it.

Objectively speaking, manufacturer's data sheets are almost always massaged to indicate the best possible spec under conditions that will never exist in the real world. Any specification may be claimed on these data sheets, and often can be grossly incorrect. Designers do not take these numbers as fact, but rely on their own testing to establish the working values of capacitors, semiconductors, etc. Having said as much, the data sheet is all you have as a starting point for selecting components. Also, good engineering practice always adds a comfortable de-rating factor even if the claimed specification is accurate.

As far as an ESR meter that could be accurate to the 1/1000 of an ohm - would that level of accuracy assist anyone in repairing faulty equipment, or in the evaluation of an individual part? Answer: no. I've never seen such a unit, but you can be certain it is not a troubleshooting aid. Similar to high end digital multimeters, resistance reading can be made at extremely low ranges, and usually demand 4 leads, careful attention to lead lengths, shielding and connectors. So if it were possible, we could cull out the questionable capacitors with an ESR of .008 ohms, from the good ones showing an ESR of .002 ohms?

[ipman]

Re: What an analog ESR meter will and won't do

@ ipman: Isn't that the one from x-roast.home.ro? I wanted one myself but i wasn't sure how well it did.

Yes, this is the one. Works just fine.

I've just aquired today a 1964 analog volt-ampermeter for two bucks. Works like new, hard to believe.


@Longbow: such fine resolution and accuracy will be required only in very special and particular applications and caps will be expensive and big. Never seen one of these, but as far as I've encountered caps (until now) had no such low values. In usual device repairs, like monitors, TV's, motherboards, etc.

[jsog]

Re: What an analog ESR meter will and won't do

There is an ESR-meter made by somebody here in Romania which has 2 decimals.
It's spot-on compared with a Fluke 189, which also can measure with 2 decimals.

Wait--there's a Fluke portable meter that does ESR?

I'm happy with my analog ESR meter, but a digital version would be handy at times. A open-source digital ESR meter that uses a standard two-line LCD display, has good input protection, and is based on an AVR would be great. Haven't found one of those yet, though.

[Th3_uN1Qu3]

Re: What an analog ESR meter will and won't do

A open-source digital ESR meter that uses a standard two-line LCD display, has good input protection, and is based on an AVR would be great.

If i make one it's going to be PIC based, so i can't help you on this one.

[jsog]

Re: What an analog ESR meter will and won't do

If i make one it's going to be PIC based, so i can't help you on this one.

If the input circuitry is well-separated from the microcontroller and the code well-commented, porting it to an AVR would be an interesting project for me. Not that I need another project, but still.

[dumpystig]

Re: What an analog ESR meter will and won't do

If i make one it's going to be PIC based, so i can't help you on this one.

I would definitely like to see this...

[Th3_uN1Qu3]

Re: What an analog ESR meter will and won't do

I have another device that deals with caps to make first (shh... surprise), i just have to convince myself to start building it.

[dumpystig]

Re: What an analog ESR meter will and won't do

I have another device that deals with caps to make first (shh... surprise), i just have to convince myself to start building it.

c'mon then, get a move on - I'm wanting to build my own, hopefully after somebody else (you?) has built and tested it first...

[tom66]

Re: What an analog ESR meter will and won't do

Wait--there's a Fluke portable meter that does ESR?

I'm happy with my analog ESR meter, but a digital version would be handy at times. A open-source digital ESR meter that uses a standard two-line LCD display, has good input protection, and is based on an AVR would be great. Haven't found one of those yet, though.

It's just an AC ohmmeter so will test ordinary resistors too, can be used as a reference against a DC ohmmeter fluke.

[lti]

Re: What an analog ESR meter will and won't do

I haven't seen an AC ohmmeter. Does the Fluke 289 have the same function?

[dumpystig]

Re: What an analog ESR meter will and won't do

Me neither. How do they work?

[Th3_uN1Qu3]

Re: What an analog ESR meter will and won't do

I haven't seen an AC ohmmeter.

ESR meters are AC ohmmeters. They work by passing high frequency AC instead of DC thru the tested component. The idea is that a resistor will still be a resistor in AC, but for a capacitor (which cannot pass DC from one side to another, so it just charges up with it), you will instead see how much resistance there is when passing the AC signal - ie, the ESR.