capacitor ESL measurement

Hello,

Is there anyone know how to measure the ESL of film capacitor?

Will there be any reference standard?

Thanks, GS

Reply to
GSiu
Loading thread data ...

by checking the series resonance frequency.

Oliver

--
Oliver Betz, Munich
despammed.com is broken, use Reply-To:
Reply to
Oliver Betz

Thanks Oliver.

Our capacitor seems to be cannot provide the series resonance frequency. Is there any equipment in the market that can measure the resonance frequency?

Gsiu

Reply to
GSiu

Sorry, I don't understand what you mean.

AC generator and AC meter, e.g. spectrum analyzer with tracking generator.

Pay attention for low inductance connections and impedance matching. A PCB is useful, but not mandatory.

If you don't completely understand the above, find someone doing it correctly.

Oliver

--
Oliver Betz, Munich
despammed.com is broken, use Reply-To:
Reply to
Oliver Betz

Oliver Betz wrote in news: snipped-for-privacy@z1.oliverbetz.de:

Would a square wave with very fast edges do it? Looking at the ringing on a fast oscilloscope might show it. I mean I'm sure it would, but I'm not sure if it's easy to avoid seeing other sources of ringing. It's easy to get more inductance in the wiring than in the capacitor! Which might be why few people worry about it. Unless you're doing GHz signals, or making a fast flashlamp laser pump, you usually don't need to deal with it.

Any chance of sidestepping the problem, by looking at capacitor choices in existing equipment where this matters, and going with the same choices? That way you might not have to measure, only to know that performance is ok.

Reply to
Lostgallifreyan

Oliver Betz wrote in news: snipped-for-privacy@z1.oliverbetz.de:

Nor do I, but it occurs to me that any resonance might be heavily damped. With some dielectrics it might well be. There are several types of 'film' capacitor, mylar, polyester, etc.. they likely differ a lot in the way small transients are handled. I guess you just have to hit it hard with a very fast transient at a regular rate and see if a scope catches the response. With capacitance and inductance it ought to ring long enough even if damped, but I have no idea if it rings loudly enough to detect it by eye onscreen.

Reply to
Lostgallifreyan

maybe, but it's much more complicated.

If you want for example power supply decoupling with 1 Ohm impedance at 100MHz, what inductance is allowed?

Oliver

--
Oliver Betz, Muenchen (oliverbetz.de)
Reply to
Oliver Betz

Oliver Betz wrote in news: snipped-for-privacy@z1.oliverbetz.de:

Is it? I was thinking it sound simpler than an "AC generator and AC meter, e.g. spectrum analyzer with tracking generator." I thought most people contemplating such measures might rustle up a scope and a fast square wave transition somehow. The scope itself often provides one, though at 1V peak it might want amplifying.

As little as possible. So I wouldn't try to calculate it at all, parasitics alone might complicate the picture. This is like asking House for a precise diagnosis when it's easier to treat the patient. The results give you more chance of the diagnosis later. That's why I'm asking if the measure is really needed. I really don't know if the first poster is asking for a preventive measure, or a precise value. The latter is hard to get whatever you do, and it's not something I see that often in cap specs, I just checked four datasheets, only one mentioned ESL at all. If the makers thought it easy or required a lot, I think they'd all specify it as they do for ESR.

In the case you mention, I wouldn't try to measure because I haven't got a few grand's worth of lab gear. But I'd use an SMT ceramic cap (possibly two, a larger value with X7R dielectric, and a smaller one parallel, with COG) as close to the IC or other parts as possible. :) After all, if 'allowed' means permissible, context is everything. A filter rolls off, this sort hasn't got a brick wall response curve so it's not very precise. It's not as bad as asking 'how long is a piece of string', but it is almost certainly easier to try the best shot and see how far it goes. Sometimes a simple question cannot have a meaningful answer if it's reduced to the same simplicity. So I'd do what is easy, safe and obvious and see if it was enough, then take it from there if it wasn't.

If anything more precise is wanted about a specific capacitor, I'd identify the cap exactly and get onto the makers for specific data beyond that in the data sheets if it's not there. And if they could not or would not provide it I'd consider it 'undefined' and maybe choose a part from someone who did, because even if I did measure it, how could I guarantee the next cap I used was the same?

I once had to handle retroreflective damage on laser diodes at a time when the makers, distibuters, and damn near everyone else was in denial of the problem. It cost my a few hundred pounds that no-one repaid me for, but we do now have a mucg wider awareness that the problem exists. But that little hazard pales beside the scale of testing that capcitors involve because there are so many types, and time and gear cost so much, and I think that if this matters, it is the makers' job to do it, not ours.

Reply to
Lostgallifreyan

"GSiu"

Our capacitor seems to be cannot provide the series resonance frequency. Is there any equipment in the market that can measure the resonance frequency?

** Depends on the type of capacitor - SMD ones are very tricky to measure.

But if it is a normal, leaded part you can assume an estimate of between

10nH and 20nH.

The actual value is very dependant on how YOU mount and use the cap in your device - as the leads and the body size are what is causing ALL the inductance.

..... Phil

Reply to
Phil Allison

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.