Ethernet, serial or Wifi interference with radio mics?

Has anyone come across this before?

I'm having serious issues with a pair of radio mics and receivers in a lecture theatre (Sennheiser ew300s, btw). They operate in the 830-866 Mhz range, and *something* is interfering with them. From the light patterns on other equipment when the events occur, it's something to do with an AMX Netlinx controller we have that coordinates the other equipment (power up, channel switching, volume control etc), either on the 100mbit Ethernet line it connects to a local PC, wireless AP (for comms with a remote control panel) and site network with, the wifi signal from the AP, or serial control to one of the connected devices.

Driving me batty, I keep applying fixes that seem to cure it for a day or two, but then the problem comes back. I've swapped the receiver output leads for fresh ones... moved the ethernet cables away from the receiver antennae... then the whole network switch... changed receiver channels... and changed 'em again... the last one seems to have cured it yet again, but I don't know how long it'll last.

It seems like it might be just one or two particular frequencies, but I'm not sure. The nature of it seems to change depending on how they're tuned, and it's affected the one receiver (usually on a lower frequency) more than the other. The interference usually manifests as quick blips of static, or clicks, though it has also been continuous static and strange electronic chirps - all at very high volume maxing out the affected receiver - and the main mixer & amplifier - and causing everyone in the room to jump!

Doesn't help that if they're even vaguely close to each other, the transmitter signals also clash despite being digital, heavily squelched, and quite narrow band. Previously the tuning was something like 831 or 839mhz for one, and 856/863 or so for the other (with some 3-figure decimal fine tuning that I forget... the manufacturer's pre-set channel "banks", basically). After playing with them a while I've left them set at 845 (850 suffered about as bad as the 83x original) and 866.000 (the highest settable frequency).

What in the world could be causing interference on these frequencies, unless it's a VERY high intensity, wideband burst? I know 850 (and the surrounding bands) is a US-region GSM frequency, but it's not used in the UK, which is probably why our WL mic systems use it.

We are having building work going on, but I can't see what strange equipment may be in use that would affect the WL mics alone and not make any of the other electronics or wireless devices act up.

BTW I have gone off looking for answers independently, including just googling those frequencies, and turned up nothing. I don't think it's the netlinx processor internals as it's a fairly industrial bit of kit that just passes signals around and provides a very simple telnet interface, I doubt its CPU runs above 100mhz (but AMX are unlikely to let that detail slip anyway). Plus the signals are too intermittent (often fairly regular though), apart from the one instance of continual static.

Thanks,

MP