"stealing" power of a 10Ghz pair

[david_ber]

Hi,

I need an advice.
I need to measure rms/ppk of a 10G signal. The signal is an output of a TIA of a 10G PIN diode. The TIA output is a neg/pos pair. The output is connected through caps to a clock data recovery chip (a point-to-point connection).

My board material is rogers 4000 series. I was wondering if I could create a "coupler" by splitting the traces on the board, and be sure that the 10G signal is still intact.

I can allways take one TIA output and use that as an input to the detector, but then the CDR would get only the other pair.

 

[vfone]

You can make a coupler, but the coupling factor cannot be less than 3dB. So, in your measurement you have to take into account this loss.
You can use a resistive splitter, but the insertion loss is 6dB.
**broken link removed**

 

[david_ber]

Thank you vfone. you gave me some ideas.