I'm with Tom on this one. This is not a noise suppression problem The placement up high on the tower means that the common mode path is nearer a voltage node than a current node and we are talking about the kind of voltage up top there that can destroy a 4000 ohm ferrite device. As the the resistance is increased, the voltage across it increases as well. I am personally well acquainted with destroying ferrite devices in this fashion.
As Tom suggests, isolate the switch from the tower. On the shack side of the switch where the phasing is not an issue, you can use 23 turns of RG400 tightly wound around a T300A-2 powdered iron toroid. Place a 500-600 pF transmitting door knob from the coax shield on one side of the winding to the other. This will create a self resonant choke with max up about 1.95 to 2 MHz. Ground the shield on the shack side of the choke to the tower, and at the base. You can monkey with tuning against a given cap by adding or subtracting turns through the toroid. Pick the lowest point of resonance that is above your operating range. Wind it with an old piece of RG59 to get the number of turns worked out and then use the RG400. The RG400 is good for 7 kW, and with a 19 strand center conductor, is the only high power coax rated for tight bends in aircraft wiring harnesses. For the control cable, which doesn't have extension beyond the box, about half-way down, and at the bottom of the tower, loop as many turns as possible through one of the huge #31 clamp-ons. 73, Guy. On Thu, Dec 6, 2012 at 6:05 PM, Jim Brown <[email protected]> wrote: > On 12/6/2012 12:39 PM, Steve London wrote: > >> My 160 meter shunt fed tower project is essentially done. However, I have >> an issue with the 80 meter antennas hung off that tower. In a nutshell, the >> current baluns (ferrite beads) feeding these antennas don't have enough >> common mode impedance on 160 meters. They heat up, and the SWR of the shunt >> fed tower changes as they heat up. >> > > Right. 5K Ohms is a good rule of thumb for choking Z to prevent noise > coupling from feedline to antenna, and if the antenna is reasonably close > to balance, is also enough from the point of view of dissipation. A very > good choke for 80 and 160 would be 16 bifilar turns of #12 on a #31 2.4-in > o.d. core (that is, 32 turns total), connected as a parallel wire > transmission line. Tightly spaced enameled wire will yield Zo of about 50 > ohms, THHN will be closer to 100 ohms. Either presents a relatively small > discontinuity (both because it's a small mismatch and because it's pretty > short as a fraction of a wavelength). > > There's measured data for chokes like these, and for a good range of > others wound with RG8X and RG8 around multiple cores on my website, along > with considerable tutorial material about how they work and dissipation > considerations. In short, chokes overheat because their choking Z is too > low to keep common mode current low. An important element of this is the > common mode voltage, which is directly related to antenna balance, and also > to the common mode length of the feedline (that is, a resonant length that > establishes a high common mode voltage). > > http://audiosystemsgroup.com/**RFI-Ham.pdf<http://audiosystemsgroup.com/RFI-Ham.pdf> > > I strongly disagree with Tom about #31 -- it is an excellent suppression > material for the HF bands, especially on 160 and 80, when used properly. > > 73, Jim K9YC > > ______________________________**_________________ > Topband reflector - [email protected] > _______________________________________________ Topband reflector - [email protected]
