A Resistive choke will *always* reduce CM currents; a Reactive choke may
not.
Here's an extract from an EMC handbook:
"For high frequency applications, ferrites should be viewed as frequency
dependent resistors. Since they are magnetic components that exhibit
significant (and useful) loss over a bandwidth of over 100MHz, ferrites
can be characterized as high frequency, current operated, low Q series
loss elements. Whereas a purely reactive (i.e., composed only of
inductors and capacitors) EMI filter may induce circuit resonances and
thus establish additional EMI problem frequencies, lossy ferrites cannot."
If chokes experience significant heating it does not mean they are
working; it means they have inadequate impedance, and have not reduced
the CM current sufficiently. Properly designed ferrite-cored chokes
should easily handle US maximum power levels - their core flux is not a
direct function of the differential power they are handling.
73,
Steve G3TXQ
On 19/11/2010 04:10, Rob Atkinson wrote:
> It certainly is true that the ferrite core c.m. chokes have much more pure
> resistance and are useful over a very wide range of frequencies compared to
> a large coil of coaxial cable (a.k.a. air core choke). Each design has its
> pros and cons. A lot of ferrite chokes experience significant heating
> because of the pure resistance (which means they are working I guess) and
> that may be okay for low duty cycles and power levels. As i mentioned
> earlier, a long 160 m. coaxial cable coil has the loss of its long run. I
> operate at a higher d/c and power level and have over heated some ferrite
> chokes. Because I only needed functioning on one band and an air core does
> not heat much, I used it. I maintain that if properly designed, they
> perform at an acceptable level--I will grant you that I may have not made
> mine well--but more important, the best choke is the missing one because it
> is not needed. With an unbalanced monopole and ground system, you don't
> need one if you have enough radials on the ground (roughly 30 or more).
>
> 73
>
> Rob
> K5UJ
> _______________________________________________
> TenTec mailing list
> TenTec@contesting.com
> http://lists.contesting.com/mailman/listinfo/tentec
>
>
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|