AnsweredAssumed Answered

Segmented Sweeps for Better DC Extrapolation at 10MHz for Time Domain IFFT

Question asked by VNA123 on Jul 28, 2015
Good Morning,

I learned about segmented sweeps recently. I also know that DC extrapolation is far more than taking the lowest frequency value of an S-Parameter and extrapolating that towards DC. When I gather VNA data, I typically always have artifacts such a "sloping" and DC offsets when doing reflective ifft in software like PLTS. I would like to know if there is a paper specifically on the DC extrapolation methods offered in Keysight VNAs and/or PLTS. Why? Well, I figured that if the extrapolation method uses certain frequency points (say 10MHz-50MHz), then I would like to see if there is an advantage in lowering the IF Bandwidth and also the Number of Points for low frequency region used for extroplation (whatever that may be), maybe throw in some averaging, and then continue with a another sweep setup for the rest of the frequencies of interest. For instance:

Typical measurement in my lab for say a 5m Length Lossy DUT: 10MHz-26.5GHz, 1KHz IF, 5MHz Steps, Power at 0dBm.

What if I did something like:

10MHz-100MHz: 500Hz IF, 500KHz Steps (181points)

105MHz-26.5GHz: 1KHz IF, 5MHz Steps (5280points)

all @ 0dBm power level.


Any advantage here? I'm on my way to test!