Rfi caused by AC units in dome

oct05

In oct05 it was found that the ac units in the gregorian receiver room produce rfi. It was seen at 327 Mhz and 430Mhz. The rfi consists of:

There is a 1 Mhz comb. The individual spikes are not resolved in a 95 hz channel width.
The comb can drift by up to 30 Khz/hour.
At 327 Mhz it got up to 8 times the noise power in a 95 hz channel.
The birdie does not go away if you turn off the ac with the remote unit. This will cause it to jump in frequency.
The birdie will go away if you turn off the breaker to the ac units. This shows that the birdie is coming from the digital control units of the AC.
There are two combs that drift at different rates. Each ac unit (there are two in stalled) produces its own comb.
Attempts at fixing the problem:

There is a switch to turn off the power to the old AC unit. This will make the birdie go away. It also causes the room to heat up.
We tried putting capacitors in the line that went down to the compressor on the service platform. This made little difference.
We will eventually replace the digital portion of the AC unit with an analog one. The units will also be moved up on top of the dome.

28mar06: look at the new ac units. See if comb is still there.
06dec05: The 1 Mhz comb from ac Unit is also seen in the 430 Mhz receiver. The comb is seen in the 430 Mhz dome receiver. The strongest members are about 3% of Tsys.
16oct05: 1 Mhz comb in 327 receiver (coming from AC units in dome). The birdie width, drift rate, and how it changes while turning the ac Remote on/off was first measured at 327 Mhz. This was done with 95 Hz channel widths.
20oct05:327 Mhz Dynamic specra over a 25 Mhz band. This shows the extent of the comb (although the wide channels are not very sensitive to the narrow birdie).

home_~phil