PXI-5112 problem with phase and impedance

[normanshi]

Hi , I need some help with using PXI-5112
My Friend and I are trying to measure the sinewave signal which is given by the PXI-5401.

1. We find the phases are slightly different when connecting the two channels of 5112 to the same signel souce of 5401.
We try to measure the phase between ch0 and ch1, and the phase changes about 0 to 3 degrees.
Sometimes the differences are very small, but some are large, reaches about 3 degrees.
Since we're just measuring the same signal, we want to avoid this unpredictable phase change.
We find that when the frequency is 50KHZ the phase differents about1 degree
                          ''                             70KHZ                   ''                       3 degree
How could this situation happen?

2. Are there any possible reasons that might change the input impedance of the PXI-5112?
We use 5401 to generate a sinewave signal and connect it to the circuit belows:
      ______________________
     |       |                       
    +      R                                 +
  FG      | ----------+            ch0 = V1
    -        |         ch1 = V2           -
    |       R             -
    |____|___________________
Since the R is the same V2 should be half of V1 and this is true when the frequency is not high(100KHZ).
The values of V2 becomes smaller than half of V1 when the frequency reaches about 1000KHZ .
When the frequency is about 1MHZ, the value of V2 becomes very small compare to V1.
We think that the input impedance might change as the frequency increases, and thus, chaning the value
of V2.


3. If there is any possible way that we can measure the input impedance of 5112, besides our current medthod?
It is very important for us to know the exact resistance and capacitance of the 5112 because our measurements involves high frequency.
For this reason, we're afriad that the total result of the changes of the input impedance of 5112 could affect our measurements.
We would set signal frequency(about 1000KHZ to 10MHZ or even more) as high as possible.

[gwd]

As the specs say, the input impedance can be either 50 ohms or 1 M depending on how it is set (in software). Again referring to the specs, this is in parallel with 30 pF PLUS whatever capacitance is associated with the cabling. For coax this is typically 30 pF/ ft.

There are two possible causes for the phase shift. One is that you are not actually doing simultaneous sampling. If that's not it then significant phase shifts can result when the two signal paths are not equal in some way. Certainly that is true with your resistor divider circuit, especially if the resistors are high value.