I had a customer call me the other day to tell me what piece of s@!%*^ my company's amplifier is because it had such terrible IP3 performance. I ran through the usual list of questions about how she was testing it - power supply, coaxial cable & connector integrity, spec an setup, etc. All seemed well, so I dared to ask about following ESD precautions and caught an earful for impugning her skills as a lifelong technician.
Just as I was about to give up and send a replacement (and have to admit as she implied that, yes, I was yet another male that cannot admit being wrong), I remembered her reading off the power supplies being used. It was two 6V supplies in series that had a 6V center tap to a separate controller board. After a little more questioning, I figured out she had the negative line of the PA plugged into the center tap of the two supplies, but was measuring the voltage with the DMM connected between the ground reference and the second series 6V supply. So the DMM read 12V while the PA saw only 6V.
The coax cables didn't pull the PA reference to ground because the DC connections are isolated from the case. Kind of a strange situation, but it's one I'll be sure to ask from now on.