Ripple current (how to measure)

[joble]

Hi,

I'm working at an SMPS and since the capacitors are the most critical
components for the life time, i'd like to calculate the lifetime of
those components.

If i look at the datasheets the ambient temperature and the ripple
current are the parameters to calculate the lifetime. It's no problem
to measure the temperature, but the ripple current is not as easy to
measure.

How can i do this?

With the currentprobe and a scope i measure a 100Hz ripple (i'm in
europe) with 130kHz added to it (= Switching frequency of the smps).
So what is the max ripple current i should take, the one at 100 Hz and
100kHz.


Can somebody explain me some about ripple currents and lifetime of
cap's.

Thanks

 

[Rene Tschaggelar]

I'm working at an SMPS and since the capacitors are the most critical
components for the life time, i'd like to calculate the lifetime of
those components.

If i look at the datasheets the ambient temperature and the ripple
current are the parameters to calculate the lifetime. It's no problem
to measure the temperature, but the ripple current is not as easy to
measure.

How can i do this?

With the currentprobe and a scope i measure a 100Hz ripple (i'm in
europe) with 130kHz added to it (= Switching frequency of the smps).
So what is the max ripple current i should take, the one at 100 Hz and
100kHz.


Can somebody explain me some about ripple currents and lifetime of
cap's.

The manufacturer have plenty of Appnotes about that subject.
I'd think both values are to be compared.

Rene

 

[R.Legg]

to measure the temperature, but the ripple current is not as easy to

measure.

How can i do this?

With the currentprobe and a scope i measure a 100Hz ripple (i'm in
europe) with 130kHz added to it (= Switching frequency of the smps).
So what is the max ripple current i should take, the one at 100 Hz and
100kHz.

The current probe will give an accurate measurement of ripple current.
This can be read on an RMS-reading scope or by plugging the current
probe into an RMS meter.

As the two component frequencies of the ripple current are widely
spaced, they could be separated by fairly simple filters, before
measurement. Depending on your topology and the capacitor location,
one is likely to dominate over the other.

Some topology capacitor stresses are fairly well documented in the
literature, allowing an indirect calculation, but you'd have to
indicate where and how this part is used before a specific reference
can be cited.

Can somebody explain me some about ripple currents and lifetime of
cap's.

This is better done by the capacitor manufacturer:

http://www.cde.com/tech/appguide.pdf

RL