Dave Platt wrote:

I used off-the-shelf (surplus-store) ferrite tubes - very probably a
43 mix or something close to it, based on the simple inductance
measurements I did with an MFJ analyzer. They're roughly 1 inch long,
and large enough to allow three through-the-center passes of RG-8X coax.

I glued somewhere around six of them, end-to-end, to create a long
tube, and then ran the RG-8X through... creating a long three-turn
choke. Added N connectors to the ends of the coax and stuffed the
whole thing into a chunk of PVC tube with end-caps.


You might do better breaking the thing up into multiple cores and
multiple windings. That is, rather than 3 turns through 6 cores, do 3
turns through 1 core, then 3 turns through another core, then 3 turns, etc.

The logic here is that the impedance scales as the number of cores, but
doing them all in one shot means that the input coax is very close to
the output coax, so you have a lot more capacitance coupling the input
to the output.

I think Jim K9YC actually did some comparisons using #31 2.4" cores
comparing stacking to stringing em out.