In Reply to: Self-resonance of X capacitor. posted by Al Sekela on August 29, 2005 at 11:56:19:
That low? I should have thought a self-resonance in the 40-50MHz.
That said, I checked attenuation curvers of standard filters from Schaffner
and Timonta and, if attenuation curves show a plateau from 15-20Mhz
depending on the model, attenuation stays above 40dB up to 50MHz, which is
what counts ;-)
As for chokes, RFI attenuation chokes use a mix of core materials that show a
high dissipation factor above the self resonance frequency (which is rather
low as you expected it, say in the low MHz). These high core loss lower the
Q of the self resonance. A typical attenuation (when fed by a 50½ source
towards a 50½ load) is 50dB at 1MHz going down to 20dB at 50MHz.Furthermore, there is a very neat trick to get rid of most the ESL (serial
inductance) of an attenuation cap. By using a "transformer" made of a few
spirals on both sides of the pcb, (without core!), and by connecting the
"primary" in serial with the incoming signal, and the "secondary" in serial
with the X cap, you can get cancellation of the ESL since this xformer,
thŽvenin-transformed, behaves in its cap branch as a ... negative
inductance. I read a very interesting paper on this in IEEE Power
Electronics Society Proceedings. Amazing how layout can be in,strumental in
building a good filter. If you want, i can check if I can get it in
electronics form. Obviously, I can't put it on a public forum without
infringing IEEE rights, which I won't... Send me an email.
I cannot go further right now, lack of time. Cheers
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Re: Self-resonance of X capacitor. - Jacques 01:11:21 08/30/05 (5)
- Re: Self-resonance of X capacitor. - john curl 15:00:03 08/31/05 (3)
- Re: Self-resonance of X capacitor. - Jacques 07:29:01 09/01/05 (2)
- Re: Self-resonance of X capacitor. - john curl 07:55:28 09/01/05 (1)
- Re: Self-resonance of X capacitor. - Jacques 08:14:42 09/01/05 (0)
- Thanks for the comments. - Al Sekela 14:15:55 08/30/05 (0)