You'll have to wait till next week when neils returns. You don't use AI1 and AI2 when you are using rear panel IF inputs, just to be clear. And you probably need the latest PNA code for the registry entry to be recongized...
here some notes he sent to me on IF inputs:
1) Create your basic wideband pulsed RF S parameter measurements in Channel 1.
2) Copy channel 1 to channel 2.
3) In channel 2 change the IF frequency to 0.0 (Remote command: Sens:IF:Freq:Auto off; Sens:IF:Freq 0; use the GP-IB console to send them directly from the PNA), and change the IF mux to measure the external inputs. Change the trace parameter, and add new parameters to measure the receivers that you are going to hook up to. You will need to change the format to linear magnitude Actually, since the input will be a receiver (like D,1) you have to use an equation like V1=55.1*sqrt(pow(re(D_1),2)+pow(im(D_1),2)); here 55.1 is the offset needed to get 5 volts when I put in 5 volts. For a receiver such as D (or any other) the lin-mag format ill display power in watts or mwatts, so it is not really lin-mag voltage. The sqrt function computes lin-mag properly, and you can use the V1=…. To set the trace display title to V1
4) At this point you are measuring the DC values of the rear panel input with the exact same timing relative to the pulses as channel 1, but not at the exact same time.
5) You need to hook up the rear panel inputs to your DC test points.
a. The maximum voltage in the rear is +/- 0.1 VDC, so you will have to create some sort of resistive scaling network and a calibration for it. The input schematic is shown below. You can just use an RF pad or series resistor to scale down the voltage. The 55.1 is the scaling needed above in my case; I used a series 10k resistor and a shunt 100 ohm, but I don’t think the shunt 100 ohm is needed in fact.
b. Hook up a ground (ground the input) and measure the offset; hook up a known voltage and get a scaling factor.
c. Use the equation editor to apply the cal factors. In my case, the offset was very small so I ignored it.
6) The use equation editor again to create the PAE measurement.
And some notes on using Ai1 and Ai2:
With NO registry tweak, we take 4 measurements per ADC, and average them. The minimum pulse width is 17 uS.
If you change the averaging to 1 via the registry, the minimum pulse width becomes 10 uS.
The above times are with a 5MHz IFBW.
If you need a narrower bandwidth, that increases the required pulse width.
