hugen79 / NanoVNA-H

NanoVNA-H based on edy555 design, provides effective measurements up to 1.5GHz.
http://nanovna.com/
515 stars 124 forks source link

Open calibration corrupts S21 measurements using 401 points at 315MHz center with 30MHz span #61

Open jeraymond opened 2 years ago

jeraymond commented 2 years ago

When I do a full calibration at 315MHz center with a 30MHz span using 401 points the S21 measurements is corrupted and looks to be showing noise.

Steps to reproduce:

  1. Configure Trace 0 as S11 logmag and Trace 1 as S21 logmag.
  2. Set Sweep Points to 401
  3. Set Stimulus center to 315M and span to 30M
  4. Reset the Calibration
  5. Calibrate Open, Short, Load, Isoln, and Thru then tap Done and Save 0.

Expected:

S12 and S21 measurements (e.g logmag) display correct values.

Actual:

S21 measurements look corrupted and show what looks like noise. S12 measurements look normal.

Other Info:

Uncalibrated, notice the blue S21 trace look fairly low noise.

1  uncalibrated

Open calibrated, now the blue S21 trace looks much noisier (corrupt). 2  open calibrated

Finished the full calibration and attached bandpass filter as a DUT. S12 logmag look correct, but S21 is noise. 3  full calibrated with DUT - S21 corrupt

Expected S21 output (but uncalibrated). 4  expected - uncalibrated

metan-ucw commented 2 years ago

I can see similar effect but at slightly lower frequency it's centered around 300Mhz. Looks like the 1.0.71 release works fine, but the 1.1 is broken this way.