[time-nuts] What phase variations to expect in a DMTD due to temperature fluctuations?

[Erik Kaashoek]

During the testing of a DMTD there appears to be a "large" sensitivity 
to temperature variations.
Opening a door in the room with the DMTD causes changes in the phase 
difference in the order of 1 ps
Blowing cold air over the encased DMTD causes phase variations up to 10 ps.
The analog part of the DMTD exists out of some coax, SMA connectors, 
active mixers and ADC's. The rest is digital and not expected to be 
temperature sensitive.
Its not clear yet which parts are causing the temperature sensitivity
As the design is symmetric (both inputs have the same components almost 
the same layout) I did not expect this.
Are these levels to be expected?
Erik.