Finding sufficiently crappy inductors

Hi Phil, (I've been away at APS March meeting.. ~many k of physicists in Boston.)

So your filter has some gain peak at 150 kHz?

You want an inductor with more loss there?

Some crafted piece of pcb? A sheet of copper gives you loss above the skin depth... could you craft some piece that had selective loss?.. or maybe with discrete's? I think this might be the 'same' as with3rd's transformer idea... a shorted turn, but only some places in frequency space.

George H.

Reply to
George Herold
Loading thread data ...

Hi Phil

Sorry if this idea is somewhat speculative (I haven't actually tried it out in real hardware), but have you thought about applying cancellation to the unwanted resonance "signal" by adding the opposite polarities?

Your circuit is a pi filter as I understand - a C to ground, followed by an L in series, and than by another C to ground.

This type of circuit will invert the phase at resonance, so a signal that matches the self-resonant frequency will appear in opposite polarities at the C before the L compared to at the C after the L.

The voltages may be non-equal (voltage ratio depends on capacitance ratio), but the polarity should still remain opposite at resonance. This should allow both sides of the signal to be summed to zero (with appropriate summing ratio, as necessary) and to cancel the signal.

If your switching regulator gets its feedback node from only the input (or from only the output) of the pi filter, the (sensitive) feedback node will get a lot of interference at resonance. If however the feedback is taken from the sum of signals before and after the pi filter (and the summing ratio is correct) the resonance signal should be strongly suppressed at the feedback node. The disturbance signal is added to the inverse of itself, resulting in near zero total.

Now, to the summing ratio. You wrote: 4.7 uF each, on either side of the L, so this makes the filter symmetrical. So the voltage ratio at resonance would be 1:1 and the summing ratio needed for cancellation should also be 1:1. In practice however, the different source and load impedances will influence the filter, so it won't be truly symmetrical. Also the phase shifts on the source and load side will prevent a full cancellation. Some measurements and experimentation may be needed to obtain the best summing ratio to cancel as much of the resonance as practical.

Regards Dimitrij

Reply to
Dimitrij Klingbeil
[...]

Hi, Dimitrij. Nice to see you back again! Don't waste all your talents on the s.e.r. discussion group alone.

--
This message may be freely reproduced without limit or charge only via  
the Usenet protocol. Reproduction in whole or part through other  
 Click to see the full signature
Reply to
Cursitor Doom

Isolation resistance isn't important, so a center-tapped inductor would do as well as a transformer. Not sure how easy it is to shoehorn one of those in, or where to uuy it Coilcraft does have, though, loosely coupled items that might suit both inductor and transformer roles.

Reply to
whit3rd

You'll get much the same effect with a large inductor paralleled with a resistor and a small inductor (representing the leakage inductance) in series with the pair (or in series with the resistor)

It could perhaps be a ferrite bead in series with the parallel resistor.

--
  When I tried casting out nines I made a hash of it.
Reply to
Jasen Betts

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.