🚀 go-pugleaf

I need a little help please.

I an trying to work out what the parasitics in electrolytics
capacitors do in terms of degrading the filtering of a standard
bridge rectifier-capacitor filtering unit.

In the catalogues two or three figures of merit are quoted:

	The ESR, equivalent series resistance,
	The loss tangent
	The impedance at ususally 10KHz.

Having looked up some application notes I discovered the following:
(all taken from Phillips/Mullard notes)

Model of an electrolytic is:
                              _____
	-/\/\/\/\------| |----|   |---
                    | _____ | -----
	            --|   |--
                      -----

A series inductor,L, a series capacitor C, a series resitor Rs, and
a leakage resisior Rp.

The capacitance is measured at 100Hz along with the ESR (ie Rs).
The loss tangent is also measured at 100Hz and is defined by
    tan (delta) = 2*pi*100*C*Rs
The impedance mentioned above is given by
	Z^2 = Rs^2 + (2*pi*f*L - 1/(2*pi*f*C))^2
where f = 10,000Hz.

The problem is that the quoted values of Rs (ESR)
and impedance Z do not fit the model (we have ignored the leakage resisitor
as this is a few meg at least). In essence I cannot calculate the value
of L from the data given.

for example a standard 10,000uF 40V capacitor from a big UK supplier has an
ESR of 18mohms and Z of 13mohms. Just thinking about this and the equation
for Z there is a problem because how can Z be less than Rs ie 18milliohms?

Am I missing something, am I better off getting a few capacitors and
measuring the resonance to get L?

At the end (apart from building the final item) I want to model the
psu in SPICE so I can actually find the "best" solution to getting good
filtering as well as smoothing.

If anyone has another helpful hints like inductive filtering these would
be welcome.

Thanks in advance

Raggy

-- 
Raggy                 | God smiles upon the little children, the alcoholics, and
u9dmlb@ohm.york.ac.uk | the permanantly stoned. S.King.