When measuring signal noise, it's important to be selective and data mine the most important terms.
Jeff and his team found the first hint of what might
be going on in these lines from the TDR (time domain reflectivity)
response of each line in the two pairs. Figure 2 shows the TDR response,
converted to a first order impedance profile.
Figure 2. TDR response of the microstrip line (red) and the stripline (in blue) converted to impedance.
This shows the microstrip (red) as a uniform line, with an impedance
slightly under 50 Ohms, of about 47 Ohms. The stripline trace (in blue)
has a very large, high impedance discontinuity at the beginning and end
of the line. What structure could be creating this impedance and could
it be related to the far end cross talk, Jeff wondered?
The second clue came from a close inspection of the artwork for the launches to the microstrip and stripline.
Figure 3. Close up of the launches to the microstrip and stripline traces.