Paul,
A good first order approximation for all cores is:
Effectiveness = (n^2)*u*length*log(OD/ID)
where n^2 = #turns squared , u = permeability (43 matl.=850, 31 matl.=
1500), length = dimension that is perpendicular to the hole, OD and ID
are the outside diameter and the hole diameter (all in same units).
I agree that for your application #31 matl. is probably best but if you
have a bunch of #43 matl. cores and there is enough room for an extra
turn or 2 the difference may be pretty small. I have a hundred or so
#31 type cores and maybe a thousand #43 cores. Over the years many
measurements I have made confirms the above formula pretty well. Bottom
line is try for at least 1000 ohms of total impedance for any frequency
you are interested in.
I am probably preaching to the choir but the impedance of the circuit is
just as important as the impedance of the choke when it comes to
reducing common mode currents. A choke with 20 ohms of impedance will
halve the common mode current in a 20 ohm circuit as will a 200 ohm
choke in a 200 ohm circuit. The circuit impedance is not easily
measured but a study done by a company in Britain about 10 years ago
said that 95% of the circuits had impedances ranging from 10 ohms to 300
ohms. Also the circuit impedance varies with frequency which should be
obvious since external wiring is involved.
Another thing that can be done if you have a cooperative environment to
do the tests is to pick a frequency and gradually ramp up the power
until you see (hear) the problem. For example if you have a problem at
the 200W level, but you want to run 1500W, the current induced in the
wires is 2.74 times more at 1500W than at 200W. Therefore the common
mode choke needs to have a high enough impedance compared to the circuit
impedance that the common mode current is reduced by more than 2.74
times. BTW 1000 ohms of choking impedance in a 300 ohm circuit will
reduce the current by slightly more than 4 times. If by some rare case
you find than you need to reduce the current by more than 10 times
(20dB) you may not be successful with common mode chokes alone and may
need some shunt capacitors from my experience.
Hope you are successful.
73, Larry, W0QE
Paul Christensen wrote:
> I am looking for small toroid cores in a #31 mix to construct my own
> bifilar-wound CM chokes for audio RFI abatement applications. From what I
> glean on the Fair-Rite website, #31 material is only available in a snap-on
> core type, elongated cable cores, and large donut cores.
>
> Anyone know a source of small toroid donut cores in #31 material? Do they
> even exist? Is #31 material exclusive to Fair-Rite? In addition to
> Fair-Rite, I tried Amidon and CWS ByteMark - and no luck there either.
>
> Paul, W9AC
>
> _______________________________________________
> RFI mailing list
> RFI@contesting.com
> http://lists.contesting.com/mailman/listinfo/rfi
>
>
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|