Change in Detector

Sonarflash, Duncan, Vancouver Island B.C. Canada.
Mounted on the wall outside my room, an Aware Electronics RM80 - LND7313-3 pancake detector was ruptured by a kitty's sharp little claws.
Over the past few months, the background count had risen to 34 CPM, approximately 0.085 microsieverts per hour with that detector.
I've replaced the RM80 it with a Russian SBM19/STS6, which is a 185 mm long by 18 mm diameter metal GM tube for beta and gamma. It is now giving an average of about 62 CPM, roughly converted to the same level of microsieverts by a division factor of 709.

Unfortunately, when operating on 20 metres, my ham band transceiver feeds harmonic RF into the detector, USB interface or cables despite layers of shielding and grounding. So, the count sometimes jumps over 100 - 200 CPM. Please just ignore my alert high readings. Hoepefully, I'll resolve this issue in the next few months.

My recommendation is to create a twisted pair cable to the Geiger tube, easily done by placing ends of wires in a vise, the other ends in a hand drill chuck, and twisting. Then wrap 5-10 turns of the the cable around a ferrite toroid, preferably no. 31 mix. Forget shielding and "grounding," which is ineffective at RF, and might make it worse.
If that doesn't fix it, you might also need to wrap the power leads around a similar toroid.
See https://www.dxengineering.com/parts/dxe-csb-525p
https://palomar-engineers.com/ferrite-products/ferrite-cores/ferrite-mix-selection
Probably also less expensive sources.