Not Enough Data on Capacitor Datasheets?

Probably 10-100mohm and SRF = 1-10MHz, depending on value, with impedance being asymptotic away from resonance. ESL is basically overall length.

Dissipation factor at least is given, at 10kHz, so you can estimate ESR in the asymptotically capacitive range.

Well, if you need to know exactly, then you'd better buy from a company that cares to specify. (I've found Epcos MKP's are usually well rated and well specified, and they're fairly cheap at Digikey of all places!)

If you don't need the specs in writing, then you can guess. I would guess they'll take around 1-2A RMS, 30ish peak, for a part similar to above (0.1uF

I've never had a polypropylene puke on me, although I'm definitely stressing these buggers:

Tim

On a related note... how are you supposed to know what ceramic caps are good for? Now, some polypropylenes come with stuff like Z vs. F plots... ceramic caps even more rarely do that. You're lucky if you have a [peak] dV/dt rating. Most of the time all I see is dissipation factor, and sometimes not even that.

Typical example... I'm always thumbing through the capacitor pages in catalogs, thinking, what would this do if I stuck it to my induction heater tank capacitor? What if I want something high frequency, but still cheap? I've looked, and no one makes MKPs that handle rated voltage over about

So I'm thinking, maybe ceramic caps will work. Now, I already know ceramic is going to suck. On principle, I avoid buying Z5U. I usually get X7R, which is +/-20%, but that's still more delta C than can possibly be healthy. Typical dissipation factor is huge, like 3.5% or something, more than ten times that of a typical polypropylene.

So the next thought comes: what ceramic actually works? C0G of course. Stable as hell, and generally considered to be lossless, I guess. But s*it, low K = huge caps = big bucks, right? Welllll, since no one actually rates their products for current, when do I know I've drawn too much? What if I line up a couple 1206 MLCC's on copper bars with heatsinks and just blast the s*it outta them? Will they survive, is it just a matter of internal heating?

Tim

The way I see it, it's bench test or request more data. Bench testing sucks cause it makes garbage. What's the success rate in receiving more capacitor data from the manu? I guess I should try someday. 'Dear Panasonic: Send me everything. Thank you.'

It could be a conspiracy. The smps companies could be in bed with certain capacitor makers. Less data is put on datasheets to retard competitors.

btw.. I was blowing up these 1812 MLCCs :(

Yup.. I'll try shopping at Epcos. Thanks..

Rough app data on film caps from Panasonic - graphs at end.

Limit specs are labelled as 'handling precautions':

from here:

Most product categories are provided with app data. You just have to look.

All categories:

RL

I found that page too. The ECWF(B) series are not on those graphs. It's a PP film cap for high freq.

The closest I got to finding a ZvsF graph for the WCWFB series on the panasonic site is:

Digikey has ECWFB

However.... ECWFS 1uf resonant at about 1kHz. ECWFL 1uf looks about resonant at about 1khz ECWFA 1uf looks about resonant at about 1khz also..

There's a pattern and it's probably a good assumption that the ECWFB series is like the ECWF (S) and (L) and (A).

"Dope from BC"

A ** 1 Farad ** electro has a self resonance at about 1kHz !!

Read the damn data properly.

The real number is around ** 1MHz ** !!!!!!!!!!!!

What use is published data to a fool who cannot read.

..... Phil

How blowing up, peak current? Using them for CW multipliers? ;-)

Tim

Oh no ..it's one of those days.. :( I wrote 1khz . That's wrong. It's supposed to be 1MHz...

Nope.. Smps app. Chip was destroyed by ripple current. I suppose it's to be expected with super thin metal in these things..

good

eramic

es not

ter

eap?

just

ramic

X7R,

thy.

s*it,

lly rates

if I

al

cx7r.pdf

Network analyzers have component testing fixtures and built in modeling software. Not cheap. I agree this data should be supplied. Of course a secondary question is how does the impedance change after being soldered, at least for surface mount.

D from BC a écrit :

Probably not thin metalization but just that X7R caps dissipation factor is somewhere between 1 and 2%.

Just size your caps appropriately.

Isn't it the dissaption factor that you need? This is the energy lost each cycle AIUI, so will set the upper frequency limit (due to heating from power losses).

Yeah, but what's appropriate? 1/4W per 1206? Some number of amps? Volts? No one says!

Tim

I can do that just fine, but how much heating:

Figure same as chip resistors, like 1/4W? 1/2W if you heatsink em? Is barium titanate less conductive than whatever resistors are made of? Can they take any internal heating at all?

Tim

WIMA probably has most of the curves you're looking for. You need to find yourself an HP4194 impedance analyzer. It also doubles as a gain-phase meter. Hard to design switchers and pulse amplifiers without a proper Z analyzer. You can find the HP4194 on Ebay.

Download AVX's SpiCap program. Gives all sorts of good information on ceramic caps.

Once again, search Ebay for an HP4194 impedance analyzer.

Aren't they thousands of dollars.. I'll probably figure out what I can do with a digital scope and DDS generator to do a Z sweep.

Will check it out..