AnsweredAssumed Answered

N9923A S21 and Math operations

Question asked by Whatsamatteru on Oct 19, 2011
Latest reply on Oct 19, 2011 by am95405
I am working on a project with the N9923A (6 GHz) and noticed that the S21 measurement does not equal the B/R1
  equivalent math function. After a preset with no calibration, using 201point sweep and 300 Hz IF Bandwidth
  here's what I get:
With both ports terminated, S21 is about -90 dB (averaging set to 64); R1 is approximately -15 dB (traces varies within
  +/- 3 dB of -15 dB); B is approximately -116 dB (averaging set to 64).  Storing R1 into memory and then when looking
  at B Data Math function was activated using Data/Mem, the resulting trace ranges from -98 to -104 dB, which isn't
  even close to the S21 -90 dB values seen.

From the FieldFox help:
  N9923A S-Parameter and Raw Receiver Measurements Table of Contents
  S21 = B / R1. The percentage of power that was measured at the DUT output
    (as measured at the B receiver) versus the power that was sent out port 1
    (as measured at the R1 reference receiver).

Is there a reason for this discrepancy?