开云体育

Hi pgc1683 (do you have a name or are you an alient ?)
I have repeated the measurement for sending you some s1p files. I have also made a report where the measurement are seen and exposed to ZPlot. Instead of implementing anything in the NanoVNA saver it is much simpler to save to s1p and import in ZPlot followed by a delay which is standard for ZPlot with selection of the cable types in question
For the VNWA software there is a simple interface to ZPlots where the measurement are automatic polled to the ZPlots
I respond to you questions below

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af pgc1682@...
Sendt: 6. december 2019 14:37
Til: [email protected]
Emne: Re: [nanovna-users] New version of NanoVNA-Saver: 0.2.1

Hi Kurt and others,

Sorry for wrong copied new post. I deleted it. Now the the correct reply:

Thanks for the detailed reply. You can send to me the measurements in file s1p format? You can measure this with professional grade VNA and export s1p format?

I frequently use zplots and sim smith chart ( ). Both program useful function extensions for vna-s. Many amateur grade instrument not have useful functions (Markers, cursors, graphs, etc.). I like nanovna-saver, many function available without external program. Some amateur grade commercials vna says: Laboratory grade. This is not true, i test the R&S entry level (!) handheld vna, much better.

Of course, many bad connector, adapters, cables exist. For example, my unkown (i think, Hugen, but not 100%) clone nanovna sma thru adapter very bad. Very lossy and not 50 ohm. I have another sma thru, much better. Near ideal with nanovna.
<<<Yes indeed the FF adaptor is very poor but from end November Alababa.com supply Hugen-H with low loss FF adaptor anf I have one it is very good
See my document for calibration data:
and female calibration NanoVNA-H.pdf


Professional adapters and calibration kit have low loss. I see many cheap adapter lossy. I think, this erros maybe compensatable. But easiest way change to good adapter.

I use lossy port extension trick on low frequncy (if osl calibration impossible), 1,8Mhz to 52 mhz. I not understand square wave signal how many bad. I think not good. Harmonics, for example.
<<< In principal terms the squarewave signal does not harm as mixed to an audio bandwith IF %KHz as I remember

Another questions, the calibrations. Nanova-saver see raw input values? Or selectable? I think not, use only 101 point nanovna calibration. This is bad for expert user, double calibrations. And 101*10 = 1010 point data, interpolate between 101 original calibration points not a best way. I think some ripple from this way, but i will test it. Professional instrument not recommended to use interpolated calibration point. If change frequency range, it works, but displayed the "uncalibrated" warning.
<<<< as all SOL standard has smooth running impedances traces 101 point are enough point and the interpolation is very close to exact calibration. NanoVNA does not use raw data but does small changes to the NanoVNA calibration. That is why you always must calibrate the NanoVNA with same setup of test cables and do a full calibration followed by a save to C0
Then start using NanoVNA saver and do calibration and measurements:

On nanovna, i can not find correct calibration. For example open short load delays not user definable. I have not zero lenght N osl calibration kit. Not all sma short has zero delay. I think requres firmware mod.

<<< the NanoVNA requires ideal calibration standard meaning open (precompensated with 50fF or 2.5ps one way) 0ps short and pure 50 ohm load. It does not matter whether this is SMA as the supplied kit or N, BNC or PL259 etc. But the trouble is such devices does not exist apart from N female short. But there is a simple stunt to do as when using N, BNC, PL259 flange adaptor the grind down the center conductor flush with the rear side for open. Solder a 2 mm Cu or Brass disk to the center conductor and the rear for short and for open use a single 49.9 ohm 1% SMD resistor as load (in the NanoVNA-saver you need anyway to create a calibration kit file and the define load as 49.9ohm) else use two 100 ohm 1% SMD resistor and in both cases grind down the center conductor to a 0.5mm height. Now you have to figure out the fringe C for the open and the shunt C for the load. More on that later but in the nanoVNA calibration kit file it is "forbidden" to use shunt C for load so the shunt C must be converted to a negative inductance after the formula L in pH= - C-shunt in fF*2.5. Then the delay from the rear of the bulkhead adaptor to the calibration plane has to be calculated and for that purpose a caliper is used.
measure in mm e.g. from the front of the female N down to the calibration plane subtracted from the entire length form front to rear and multiply by 4.79 and you have the one way delay = the offset used in NanoVNA. This way you have a super N kit. The fringe C for load is a bit dependent the amount of solder used but about 150fF and the open fringe C is 80fF
Calibrating the VNWA is just using the fabricated N standard and in the NanoVNA enable the negative electrical delay with the calculated adaptor delay. Then the calibration to the rear of the adaptors are pulled back to N adaptors calibration plane. If only the calibration of the NanoVNA prior to using the NanoVNA saver then forget about the Electrical delay as that only facilitate fairly correct measurement using the NanoVNA as standalone without compensating for the fringe C of open and load.
If you want to do it the complicated way see my document:
to fabricate a N Male and Female calibration kit rev2.pdf
Kind regards
Kurt