Using a spectrum analyzer plugged into an 19" equipment rack, I touched my probe to that equipment rack ground bus and see lots of noise between 10khz and 100khz. This noise is as high as -70dBm. Moving the power cord to a wall outlet that is powered from the same panel I get the same result. However, if my spectrum analyzer is plugged into a battery (or UPS) isolated from the power source for the room and then I touch the probe to the ground bus, I measure almost no noise in that range (less than -100dBm noise floor on scope). My question is, what is the proper method for measuring the noise that may be on a ground bonding network in a lab? Are my initial measurements that show the large spikes accurate measurements? The battery perhaps does not provide the proper reference to the origination of the signal/power/noise?