Measuring the energy used to recharge a 1mF electrolytic capacitor

[treez]

Thanks, another point is that if you have a current pulse with a peak of 150A and duration 200us, then what ESR does the electrolytic capacitor present to that pulse?...its anyone's guess......as you know, we would need to know the ESR so as to do the calculation i(rms)^2.ESR.

 

[c_mitra]

...then what ESR does the electrolytic capacitor present to that pulse?...

This is a question that really does not have an answer; because

1. ESR, is part of the impedance model and is frequency dependent (unlike a real resistor)

2. If you apply a pulse it is equivalent to applying a broad range of frequencies and the ESR is going to be a just another parameter

3. It is simply going to fit a model.

 

[CataM]

Thanks, another point is that if you have a current pulse with a peak of 150A and duration 200us, then what ESR does the electrolytic capacitor present to that pulse?

Why don't you go to manufacturer's datasheet of the capacitor you have and take a look at the table they present with the ESR the capacitor has, as well as the graphs they show for the ESR vs temperature. (e.g. vishay's capacitors)

However, as c_mitra says, a pulse has lot of frequencies meaning that the graph of impedance vs frequency is useless (in case it was tested aginst sinusoidal input... which I do not know).

Even though ESR varies with frequencies, you can take the Impedance vs frequency (in case it was tested against pulse input) graph and model your capacitor as such: ESR---ESL---C

• 
  ESR : freq independant once is calculated at the working frequency, temperature dependant

• ESL : independant on everything

• C : temperature dependant and then only use the C/Co vs temperature graph.

 

[Warpspeed]

Electrolytics always have an rms ripple current rating, and that may be a good starting point for estimating acceptable temperature rise.
It may be 150 amps peak, but what is the rms value of charge + discharge over a complete one second cycle ?

A fairly ordinary current transformer may be a big help here. Current transformers are rated by how much the winding can continuously carry. A 20 amp rated CT will easily handle a 150A peak, as long as the rms is below 20 amps, and the burden resistor is made low enough to not exceed the rated maximum output voltage (in volt microseconds).

I understand how sales departments and management really like to talk about high power density, but with something like this that may come at the cost of long term reliability.
An injection of reality at the next planning meeting may be required once you can nail down some facts.

If you do point out potential problems, and are overruled, and later the whole thing turns into a disaster, your arse is then covered.

 

[c_mitra]

...Electrolytics always have an rms ripple current rating, and that may be a good starting point for estimating acceptable temperature rise.
It may be 150 amps peak, but what is the rms value of charge + discharge over a complete one second cycle ?...

This is the practical way to look at the problem. Originally, the electrolytic capacitors were designed for filtering ripples.

If you rapidly charge an electrolytic (Al electrolyte) to the rated voltage and rapidly discharge it and then leave it open circuit. You will notice that over a few ms some charge appear and the capacitor shows some voltage that may be a significant part of the original charging voltage (often more than 10-20%). The ions cannot move fast enough and the capacitor discharged over a period of 10us is not fully discharged if seen on a scale of 10 ms.

They are a strange beast in the lot.