[RFI] Yaesu G-800DXA Rotator - RFI mitigation

[Brian Moran]

Has anyone else used a G-800DXA rotator/control box in a high-rf environment? We had two stations (both K3) going this weekend with 5B4AGN filters and about 1KW each,  and noticed that the G-800DXA (fed with unshielded cable) seems to be bring in interference (hashy noise for CW) when transmitting on 20m, and the other station rx is on 15m. With the rotator connected to the controller in the shack, transmitting on 20m causes noise on 15m to increase by approximately 3 S-units (K3); disconnecting the cable at the rotator but not at the control box drops 2 S-units; disconnecting the rotator cable at the control box reduces the interference somewhat; disconnecting at the box, and at the rotator, reduces the interference to under an S-unit. Turning the rotator control box ON when all is connected makes the problem worse. We suspect that the rotator cable is picking up RF, and re-radiating it into the shack, potentially after doing some rectification in the rotator unit and/or control box. We made a stub (~23' open) to put on the 20m station with no discernable effect before we figured out that the rotator was involved. We have endeavored to follow good practice of grounding everything to a single point; we perceive that grounding the rotator control box makes the problem worse.
We note that according to the schematic, there are diodes in the rotator itself; there are certainly non-linear devices inside the rotator box itself. Our next steps are to bypass all of the rotator cable leads with 0.01 caps at the rotator control box and as close to the tower unit as possible. We have a different issue when transmitting on 20m while attempting to receive on 40m, but thought we'd peel this onion a layer at a time.
 
Has anyone else had to mitigate issues like this with this model? What did you do? 
Thanks!
Brian N9ADG