AnsweredAssumed Answered

E5071C IF bandwidth problem

Question asked by GW on Feb 15, 2008
Latest reply on Mar 24, 2008 by OsamuKusano
I have an E5071C with the latest firmware and option 08. In one setup it is interfaced to an E8257C that is using the 10 MHz reference from the NA. This setup utilizes the external generator as a higher purity source for a measurement that is spur and harmonic content sensitive. The NA is set for absolute D(1) and frequency offset mode (FOM) where the generator is substituted for port 1.

During an evaluation of a new design at high temp using this setup I increased the IF bandwidth of the NA to 200 kHz to accommodate the temperature drift of the UUT's imbedded LO (measuring wide band converter response) and got very unexpected results. Troubleshooting this resulted in a very simple setup in which the signal generator was simply connected to the NA. The attached screen shot shows the interface cable loss (yellow memory trace) and looks exactly the same at all IF settings from 10 Hz to 100 kHz. When the bandwidth is increased to 200 kHz (cyan trace) the response changes dramatically showing correct measurements at a few points but with the bulk of the sweep at a much lower level as if the IF filter was not centered properly on the generator's output. I suspect this because the same thing occurs at higher IF bandwidths but to a lesser degree. I also slowed the sweep way down to confirm that the generator was not being missprogrammed. It's my understanding that in frequency offset mode (FOM) the IF response is exactly centered on the measurement point whereas in normal ratio mode this may not be the case but doesn't matter since its a ratio measurement. What then is happening at 200 kHz and above? Can this be fixed quickly?