AnsweredAssumed Answered

N5241A & SMC+phase- group delay measurement at low input power

Question asked by d_finz on Nov 8, 2014
Latest reply on Nov 9, 2014 by Dr_joel
Hello.  I am preparing a test plan for a C-band to L-band receiver that has a input power range of -70 to -95dBm with approximately 65dB of gain.  I have a N5241A that I want to use for making the gain, group delay, return loss (SMC+phase method), C/IM3 and noise figure measurements.  I put together a mixer, amplifier and filter system using off-the-shelf hardware to emulate the receiver while preparing the test plan.
I've been working on the bench with the PNA-X and the emulator and so far I've been able to make successful group delay, return loss and gain measurements for input levels from -70 to -80dBm.  However, once I begin working at lower power levels, my group delay and input return loss measurements are indiscernible from all of the noise present on their respective traces.  I was able to reduce the noise by decreasing the IF bandwidth.  This improved the input return loss measurement but I still am having problems with my group delay.  The shape of the group delay trace is not consistent at different power levels below -80dBm no matter how much I lower the IF bandwidth; however the noise on the trace did improve.

I'm concerned that my setup of the PNA-X may be the cause of my group delay problem.  I'm using port 1 as the C-band input port, port 3 as the LO and Port 2 as the L-band output port.  The source attenuator on port 1 is set to 65dB (I did my phase reference cal and S-parameter cal at this source attenuation level).  The attenuators on all of the receivers are set to 0dB.

I'm concerned that the power being coupled to the Port 1 receivers is dropping too far below the receiver's sensitivity using this implementation (for example, at -80dBm and 15dB coupling factor, the power getting to the R1 receiver would be -95dBm).  I figured that if in fact my implementation is a problem, then I could add a 30dB pad in between Port 1 and my test cable and then drop the source attenuator to 35dB thereby increasing the power coupled to the receivers by 30dB and still maintain the input power range that I need.

So, my questions are:
1. Am I in fact affecting the measurement capability of the PNA-X by setting the source attenuator to 65dB?
2. Is my alternative method of implementation going to solve the problem?

Thanks for the help!!!