Maximum ripple voltage on film capacitors is strangely low?

[grizedale]

Howcome the rms ripple voltage on these capacitors is rated so low..........(please see page 2 of datasheet, below)


...eg, top right graph, pg 2,...the 820n, 400V cap is only allowed a maximum of 20 Vrms ripple


isnt that incredibly low?


Kemet PHE426 series pulse film capacitors
http://www.kemet.com/kemet/web/homepage/kechome.nsf/file/PHE426%20Series/$file/F3294_PHE426.pdf

 

[FvM]

The said 20 V Vrms value is corresponding to 10 A Irms, which is not incredible low in my opinion. It's already limited by the pin diameter and current capability of the internal connections. The current may be too low for some applications. The Kemet specification is quite representative for similar types from other vendors, however. You won't easily find a much higher current rating for these small capacitance values.

 

[mtwieg]

Pulse capacitors generally are rated for high peak current and not RMS current ratings. But seeing them rated for RMS ripple voltage and not RMS current is sort of strange.

 

[WimRFP]

Hello Grizedale,

I use such metallized foil capacitors frequently. These values are good (better then what I have in my stock). You may check Epcos or Evox, and you will not find much better. They use PP dielectric, so you are limited in temperature rise. Values are for 10 degrees temperature rise starting from 85 degr ambient temperature.

 

[grizedale]

Thanks All

FvM:

I presume that you arrived at 10 Amps rms by assuming that the 20Vrms rating referred to a sine wave.?

v = Vsin(a)

did you then go...... i = cdv/dt ?

So the Vsin(a) went to Vcos(a)

..did you then do.... i = C * Vcos(a) ?


WimRFP, thanks for the tip, ....by PP i presume you mean Polypropylene....and this is the best for high rms current?

 

[WimRFP]

Correct, PP = Polypropylene. PP is a low loss dielectric (like PE and PTFE), so useful at 100 kHz and even above. Polyester is bad when overall loss is of importance.

10A at 20V is 200VA of blind power in a small case. The Q factor should be reasonable as you can't dissipate much (or you get significant temperature rise).

 

[grizedale]

Thanks,

Do you know how i can find the ESR of the 820n, 400V PHE426 series cap at 84KHz?

I always use ESR = dissipation factor * Xc.

However, the datasheet does not give the dissipation factor for anywhere near 84KHz,

I presume FvM managed to calculate the ripple current allowable from the datasheet. .....though i am not clear how this was done.

i thought ESR was the king of all parameters in a cap...but this datasheet does not appear to allude to it.



....Ahhhh...i see , the Xc at 100KHz is 2 Ohms...then 20/2 = 10A rms.

...but surely this cannot mean that the allowable ripple current in that cap is that massively high?
..i mean 10A rms of ripple is a massive amount of ripple current.
-it surely cannot be right?

 

[FvM]

I simply calculated Irms = Vrms/Xc. ESR or DF factor aren't specified (except for some DF limiting values). Snubber capacitors with wide low inductance terminals are partly specified with ESR values, e.g. Epcos series B32686S.

 

[WimRFP]

Hello,

You are right, they don't specify ESR. At 100 kHz the reactance for 820nF is 1.94 Ohms. The datasheet says max. 20Vrms for 10 degrees temp. rise.

So the current will be 20/1.94 = 10.3Arms.

The blind power is 206VA. I quess that you need well below 2W of internal dissipation to get 10 degr. temperature rise for the size of the capacitor. Let us assume that the internal dissipation is 1W. In that case the Q factor (1/(loss factor) = 206/1 = 206. So ESR = 1.94/206 = 9.4 mOhms (at 100 kHz).

Your 84 kHz is close to 100 kHz, so ESR will only slightly decrease.

Off topic, I am now trying to get 850VA blind power at 100 kHz into a E42 core with some copper on it. Isn't easy and maybe I go a step up (in size) to reduce the loss.

 

[grizedale]

ref your offtopic, the smps from coldamp.com purport to do what you try , and they use smaller than E42.