[TowerTalk] decoupling RF from control lines

[Pete Smith]

RFI in an LV system is not actually the problem here, Tom -- what I'm dealing with is RF cross-talk between two TopTen 6-way relay boxes in an SO2R setup.  When I bring one 6-way physically close to the other (both apparently well shielded) with 100 watts going through one into a dummy load, the receiver going through the other to another dummy load shows well over S9.  I have yet to fully work through bonding, grounding, etc., but figured I should think of coupling to the control lines as well, because the same proximity effect is visible when I bring one 6-way near the control lines of the other.

I have some small ferrite beads from an HF choke balun kit, and can get 10-15 turns of wire through each - would this be an appropriate sort of choke to try on the control leads?

73, Pete N4ZR



At 11:13 AM 8/16/2005, Tom Rauch wrote:
>used for 13 volts out to the relay box.  The .1 caps were
>suggested by TopTen for best effectiveness at 160, but what
>I was really unsure of was the inductance of the RF chokes.
>I'll certainly have a look at your app notes.>>
>
>As for component values Pete, they all depend on the
>application. The voltage or current sensitivity of the
>device to RF, the waveshape change on control lines the
>device can tolerate before switching or component life
>becomes a problem, and so on. There is no real universal
>value of filter that works.
>
>You are significantly better off with an absorbing material
>like a ferrite with a very low Q at HF rather than
>inductors. The bypass caps mostly establish a low shunt
>impedance for the series impedance of the chokes to work
>against. In virtually all cases .01uF or .05uF caps would be
>more than adequate for 160 meters. A .01 is under 9 ohms on
>160 meters. Anything less than several ohms reactance is
>really just poor engineering. If it takes less reactance
>than five or ten ohms to prevent RFI in a LV circuit the
>system clearly needs some series impedance!
>
>It is very unlikely any value disk cap when the wire leads,
>ESR, and connection path in the equipment is included would
>have less than five ohms impedance from 1.8 to 30 MHz. Why
>pick a cap that might have .9 ohms reactance on 1.8 MHz only
>when perfect (no ESR) with zero lead length?
>
>73 Tom