LCR meter and ESR for small capacitors

[Elerion]

Hi.
I've bought my first LCR meter, capable of testing at 100 Hz to 100 kHz.
It can measure ESR in Cs mode (capacitance, series measurement).
For electrolytics, it seems to work fine.

But, if test a good 100 nF ceramic capacitor, it measures hundreds of ohms resistance, which it obviously must be wrong.
For a good 1 uF ceramic, it tells me 16 ohms (again, wrong, it should be an order of magnitude lower).
The D factor is so low for all of them. Just the ESR is abnormally high.

Also, for the some working caps, I sometimes see a lower D factor at some frequency, but a higher ESR.
This also happends for different electrolytics. Some have a better D factor, but worst ESR. Does it make sense?
Could it be the meter?

 

[Nigel Goodwin]

ESR is for electrolytics, don't worry about it for other types of caps.

High ESR is by FAR the most common failure mode for electrolytics, and the most common fault in electronics generally.

 

[Elerion]

ESR is for electrolytics, don't worry about it for other types of caps.

Yes, I know. But I wanted to test if the meter is working fine, and in the way, learn a bit more about how they work.

The discrepancy between D factor and ESR does concern me.
Isn't a bad ESR capacitor suppose to have a high D factor?
Why could a higher ESR capcitor read better D factor on a meter?
There's a relation between them, as far as I know from theory.
I heard some people replacing electrolytics based only on the D factor, not the ESR.

 

[Nigel Goodwin]

I'd never even heard of the term until you mentioned it, and I've never seen an ESR meter that displays it.

ESR meters are extremely common in the service trade, and I've used them thousands of times locating faulty electrolytics - the other option is to use a scope, assuming the unit powers up for scope readings.

 

[kubeek]

Try measuring say a 1ohm smd resistor, it should have very little capacitance or inductance, but should show exactly 1ohm.

 

[The Electrician]

But, if test a good 100 nF ceramic capacitor, it measures hundreds of ohms resistance, which it obviously must be wrong.
For a good 1 uF ceramic, it tells me 16 ohms (again, wrong, it should be an order of magnitude lower).
The D factor is so low for all of them. Just the ESR is abnormally high.

What make and model is your new meter?

You didn't say at what frequency you were making the measurements. What style is your 100 nF caramic? Is it an older disc type, or a more modern surface mount? Try measuring your 100 nF ceramic cap at 100 Hz, 1 kHz, 10 kHz and 100 kHz. What do you get?

 

[Elerion]

It is a handheld Tenma, capable of 100kHz and ESR measurements.

I tried all frequencies, from 100 Hz up to 100 kHz, but focussing on 1 kHz.
100 nF older disc type, yes (leaded). Not multilayer.

 

[Nigel Goodwin]

It is a handheld Tenma, capable of 100kHz and ESR measurements.

I tried all frequencies, from 100 Hz up to 100 kHz, but focussing on 1 kHz.
100 nF older disc type, yes (leaded). Not multilayer.

As I said above, you're wasting your time - ESR testing is for electrolytics, not small capacitors.

 

[Elerion]

Sure. I just answered the questions.
The main point of all this was testing if the meter was operating correctly.

Remember it is a LCR meter, not just an ESR meter.

 

[Nigel Goodwin]

Remember it is a LCR meter, not just an ESR meter.

I do, but it doesn't mean that ESR is relevant for non-electrolytic capacitors.

 

[The Electrician]

It is a handheld Tenma, capable of 100kHz and ESR measurements.

I tried all frequencies, from 100 Hz up to 100 kHz, but focussing on 1 kHz.
100 nF older disc type, yes (leaded). Not multilayer.

What values did you get at the various frequencies?

 

[The Electrician]

It is a handheld Tenma, capable of 100kHz and ESR measurements.

I tried all frequencies, from 100 Hz up to 100 kHz, but focussing on 1 kHz.
100 nF older disc type, yes (leaded). Not multilayer.

You should expect the ESR for an older disc type ceramic cap to be high at low frequencies. Here is an image showing the ESR and impedance magnitude at frequencies from 100 Hz to 5 MHz of a 100 nF ceramic disc cap. The impedance is shown in green and ESR in yellow. Note that the scales are logarithmic.

The ESR at 100 Hz is 385Ω, at 1 kHz it's 38Ω, at 10 kHz it's 5 ohms and at 100 kHz it's 1.4Ω

Be aware that the ESR of capacitors of various types typically vary with frequency, sometimes by orders of magnitude. There's nothing wrong with your meter.

 

[Elerion]

Here is an image showing the ESR and impedance magnitude at frequencies from 100 Hz to 5 MHz of a 100 nF ceramic disc cap.
...
There's nothing wrong with your meter.
View attachment 110460

Thank you! That was exactly what I meant.

Several hundreds ohm resistance is very high. So, does this mean that ceramic capacitors are unsuitable for low frequency filtering?

 

[Nigel Goodwin]

So, does this mean that ceramic capacitors are unsuitable for low frequency filtering?

Is this why you've been trying to check ESR of small capacitors?.

For decoupling you basically stick a high value capacitor across the supply for low frequencies (which won't be any good for high ones), and a low value capacitor across the supply for high frequencies (which won't be any good for low frequencies), the two together cover pretty well all you need.

No need for tests or measurements, it's simply standard design practice.

 

[kubeek]

Most esr meters don't discern between esr and capacitive reactance. That is why the esr number looks horrible at low frequency, even though that is the reactance and esr could be much less than an ohm. Will try with two lrc bridges on monday to be sure

 

[The Electrician]

Most esr meters don't discern between esr and capacitive reactance. That is why the esr number looks horrible at low frequency, even though that is the reactance and esr could be much less than an ohm. Will try with two lrc bridges on monday to be sure

It's not the reactance that Elerion's meter is measuring at low frequencies; it is in fact the ESR.

Elerion's meter isn't specifically just an ESR meter; it's an LCR meter. He didn't give us the model number, but I'm pretty sure it's this one: http://www.newark.com/tenma/72-10465/lcr-meter/dp/26W7021

The fact that it can measure at frequencies of 100 Hz, 1 kHz, 100 kHz and so forth differentiates it from so-called ESR meters which almost always measure only at 100 kHz. It's an LCR meter, not just an ESR meter.

The reason Elerion got a high value for ESR at low frequencies is because those old leaded disc capacitors actually have a high ESR at low frequencies. I showed this with the impedance analyzer sweep in post #12. You should get the same result with a bridge. His meter can discern between ESR and capacitive reactance.

This Tenma meter is one of a breed of meter that various manufacturers have begun selling since the nifty Cyrustek chip became available: http://www.cyrustek.com.tw/product-2-6.htm

This makes it easy and inexpensive to manufacture a handheld LCR meter covering the standard frequencies of 100 Hz, 120 Hz, 1 kHz, 10 kHz and 100 kHz. Besides the Tenma meter, the DE-5000 meter is a prime example of such a meter that can be bought for less than $100: http://www.ebay.com/itm/DER-EE-DE-5...179223&hash=item25ed243ceb:g:JpcAAOSw~QRaLNG0

These meters based on the Cyrustek chip can distinguish between reactance and ESR at low frequency, medium frequency and high (100 kHz) frequency. They also can measure ESR at 100 kHz, and they can measure capacitance and inductance. They do a surprisingly good job given their low cost.

 

[Elerion]

Elerion's meter isn't specifically just an ESR meter; it's an LCR meter. He didn't give us the model number, but I'm pretty sure it's this one: https://www.newark.com/tenma/72-10465/lcr-meter/dp/26W7021

It is.
The meter seems a good alternative to very expensive units.
But measuring high inductances (several Henry and up) is tricky. The reading value does not stabilize. It very slowly drifts for so much time, I end up taking an estimation. Sometimes the "Lp" value slowly goes down, while the "Ls" value slowly goes up.
Is this normal?

Is this why you've been trying to check ESR of small capacitors?.

No. Just testing the LCR meter.
I bought the meter to test power and audio output transformers and inductors, and, as I have no esr meter, to also detect faulty electrolytics.
The DCR measurement for small resistance is better than my true-rms DMM (I just miss that it could make four point test measurements).

 

[The Electrician]

But measuring high inductances (several Henry and up) is tricky. The reading value does not stabilize. It very slowly drifts for so much time, I end up taking an estimation. Sometimes the "Lp" value slowly goes down, while the "Ls" value slowly goes up.
Is this normal?

This can happen when you try to measure the inductance of a winding which has an iron core, such as a power transformer winding, or a large crossover inductance with an iron (or ferrite) core.

The inductance of iron cored windings varies with the excitation level. Since your meter uses a fixed excitation level you can't do the experiment where you can see the different measured inductance as you change the excitation, but it does exist. When your meter suddenly applies a sine wave to the inductance, there is a transient component that decays with time and causes the slow drift in the measurement.

Try this: hook up your meter to an inductance with iron core and wait for the reading to stop drifting. Then disconnect the meter and reconnect with the leads reversed. You should see a different reading which drifts a lot for a while, then should settle to approximately the same reading you got before reversing the leads.

There's not really anything you can do to avoid this phenomenon. You just have to realize that measurements of iron cored inductances are very approximate, and will change with time and with the magnetic state of the core.

With regard to that last bit, try this: measure the inductance of a winding on a small power transformer. Place a decent size magnet in contact with the core momentarily then re-measure the inductance. It will probably be quite different.

 

[Elerion]

This can happen when you try to measure the inductance of a winding which has an iron core, such as a power transformer winding

Yes, that's the case. Chokes, power and audio output transformers.
It also depend on the measurement mode; if measuring series or parallel, which is something I'm still learning how to use properly.

Oveall the meter is good value for the money, and it is very usefull to measure parasitic inductance and capacitance without having to setup a test circuit, signal generator and oscilloscope.

 

[Elerion]

I've been working on several medium wave receivers. I began using a homemade variable capacitor (using tubular cardboard and aluminium foil), then I got an old variable air capacitor. With it, the radios behave better; the output sound is louder.

Even though Nigel said that ESR is of concern only for electrolytics, I've been comparing these totally different capacitors, and (obviously) very different ESR values (at same frequencies).

At 100 kHz, my homemade cap reads:
D = 0.006
Q = 154
ESR = 45 ohm
angle = -89º

The old good air capacitor:
D = 0
Q = >2000
ESR = 2 ohm
angle = -90º

I still don't understand why some people say that D factor is a good indicator of a good/bad capacitor.
I even read sometime that if a capacitor's D is below 0.1, it is a "good cap".
My homemade capacitor is surely inferior to most commercial ones, and has a high ESR even though its D=0.006 (which is low). I've seen this situation with other capacitors.
But AFAIK, D and ESR are related.
Why sometimes D is low while ESR is high?