Measuring Equivalent Series Resistance at higher frequencies

[dannyf]

Any method employed to measure anything can be incorrect and/or very error prone under some circumstances.

Yet, we use them all the time.

[The Electrician]

When you said "it can go negative, depending on how it is measured.", it was unclear what you meant by "it".  Apparently the red "it" refers to the result of a measurement, which could certainly be in error, even a negative value; this is obvious.  The blue "it" would appear to refer to the actual value of the ESR, which is never negative.  You're using one word, "it", to refer to two different things.  A person reading your post could think you were saying that the actual ESR can be negative.  I thought so, and apparently KJDS did also.

My point was that the video at about 2:18 shows the ESR of the capacitor at low frequency being about 1 ohm and at the high frequency end it reaches a value of about -8 ohms. This is not just a small error; he's doing something wrong.

Part of what he's doing wrong is using an ordinary resistor with long leads as his 50 ohm reference at frequencies nearing 100 MHz.

Knowing that ESR can never be negative should be a clue that something's wrong when a result of -8 ohms is obtained for ESR.

[dannyf]

ESR can never be negative

In general, resistance can be negative. It is questionable if the ESR of that particular capacitor is negative.

Hope it helps.

[KJDS]

ESR can never be negative

In general, resistance can be negative. It is questionable if the ESR of that particular capacitor is negative.

Hope it helps.

ESR of a passive can never, ever, be negative. It isn't questionable if that cap has a negative ESR.

Having said that, there are some semis that do exhibit bands of negative resistance, but that's not what we're discussing here.

[G0HZU]

There's something wrong with Bart's software.  In the first video at about 2:18, he says that the ESR goes negative.  This is impossible; ESR never goes negative.  If it did, we would have perpetual motion.  A negative ESR would be a source of energy, and that can't happen with a passive component like a capacitor.

I just watched it and I agree that there are lots of technical errors in that video. The person presenting it clearly doesn't have any experience of component theory or practical testing/layout or even how a typical capacitor behaves as it approaches resonance.

I learned some time ago that most hobby based tutorial videos involving RF theory/measurement are pretty dire. What's really bad about this one is that the presenter sounds confident about the results he is getting even though the results defy basic physics.

[Conrad Hoffman]

Hey, defying the laws of physics is a tough job, but somebody has to do it.

Long ago I got a nice HP 4815 vector impedance meter, good to 110 MHz. The first thing I learned was how much the test setup contributes to the measurement. In fact, it's almost impossible to design a test fixture that doesn't introduce errors you have to correct for. The second thing I learned was how little I really knew about the HF behavior of components.

[The Electrician]

I just watched it and I agree that there are lots of technical errors in that video. The person presenting it clearly doesn't have any experience of component theory or practical testing/layout or even how a typical capacitor behaves as it approaches resonance.

I just love the way he has the DUT just lying on the bench, with clip leads and scope probes connecting his box, a full lead length ordinary resistor as his 47 ohm reference and then does a sweep up to nearly 100 MHz!  Sheesh!  His setup with the BNCs for surface mount components is much better, but it's still not a proper jig.

I learned some time ago that most hobby based tutorial videos involving RF theory/measurement are pretty dire. What's really bad about this one is that the presenter sounds confident about the results he is getting even though the results defy basic physics.

After re-watching it, I notice that at that same spot (about 2:18) he outright says:

"If you look at the effective series resistance, it's gone negative....That tells us that it's gone inductive".  He doesn't see the fact that ESR measures negative as a sign of a problem; he thinks it's ok, and furthermore allows us to infer that we're above the capacitor's SRF, where, apparently, we should expect the ESR to be negative.