Moving wires closer to ebox results in larger peaks in PSDs. Wires appear to be main source of electrical coupling.
Twisting wires does not help.
Reducing wire length running from function generator to speaker does not help.
Mounting or holding speaker to under fridge frame does not lead to reduction in peak intensity.
Moving DCRC board away from ebox with D-50 ribbon cable (pilfered from breakout box), and positioning it behind rack does not help. Covering the cable with aluminium foil has no effect. Grounding aluminium foil has no effect.
Moving speaker away from ebox results in lower peaks in PSD. However, there is no difference in peak intensity if speaker is mounted to or held above fridge frame, implying that mechanical vibrations are dampened significantly.
Holding piezo speaker to 1 K pot needle valve results in a lot of electrical coupling.
No peaks are visible when speaker is held against dewar, again implying that none of the peaks are due to mechanical vibrations.