开云体育

Hi Lajos
I see you open phase is -0.22 degree at 250MHz and that equals 2.44ps delay. That look like you are suing the male open adaptor which in my calibration document gives has a offset one way delay of 1.62ps.
The 0.98ps is the one wat Offset delay when the SMA female adaptor is left unterminated during the calibration which after my calculation gives 1.111ps and that is so to speak the same as thee 0.1degree as the accuracy of the angle readout is limited by rounding to 1 or 2 decimal digits. There will always be fluctuations form sweep to sweep du to noise.
0.98ps is measured with the VNWA after calibration with my HP83033C 3.5 mm calibration kit.
I do not understand you comment "Now, next calculations you use 0,98 and 250, not 245,025. " where and when did I use these figures ?
Regarding the VNWA arbitrary values for calibration kits they can be converted to value to enter in the NanoVNA-saver calibration kit settings and that is a bit complicated for explaining in a mail response. I will make a document when I get time and publish of interest for those having VNWA software and want to use/purchase the SDR-Kits SMA and BNC and soon also N calibration standards with the NanoVNA .
The other way round the NanoVNA calibration kit is very easy to convert to VNWA as you just copy a calibration kit file to another name and use the simple settings and enter the two way delay data for short and open and the load data with two way delay data as indicated in my calibration document
Regarding the fringe C for load there is no easy way to find it an there might also be a delay and inductance to determine and requires a known calibration kit for determining these data.
One of my 3 NanoVNA loads are also broken
Kind regards
Kurt

-----Oprindelig meddelelse-----
Fra: [email protected] <[email protected]> P? vegne af pgc1682@...
Sendt: 13. december 2019 10:36
Til: [email protected]
Emne: Re: [nanovna-users] New version of NanoVNA-Saver: 0.2.1

Hi Kurt,

I tried your latest How_to_display_delay_shift_for_a_cable_in_NanoVNA_saver.pdf method.

Now I have many problems. I not understand what I do wrong, or what corrections/calibrations will be required.

See the attacment images. I think my nanovna calibrations not ok. Supplied coaxial cable not symmetrically spiraling inwards to centre smith diagram. Your calibration much better. I try this with different cable, also not good. Teflon dielectic cable, i think better than supplied. I currently not have semi rigid cable to better test. I will get it, but requires lots of time.

I use your method, and your calibration contans.

Now I use latest QRP 4.4 firmware, clearconfig executed, and instrument calibrated with your method.

Formerly, I have good calibrations without trick. It works well, I use many times, working good. But I lost it, I can't reproduce it. I not know why not.

Where from your 0.98 ps? I see on your document marker 1 cursor on 245,025Mhz S11 phase -0,1° so i get 1,134 ps, not 0,98ps. Now, next calculations you use 0,98 and 250, not 245,025.

Fringe C capatince not know how to compute what it causes (more than about clockvise phase shift), and i not know how to measure or compensate manually. Maybe my Hugen-H like calibrators not same? I measured load, only with multimeter. I get 49,9 ohm final value, (subtracted the internal 0,4 ohm error). Not a precise instrumnet, and not 4 leg method, but i think load resistance value good.

I see DG8SAQ WVNA arbitrary calibration formulas to oslt. I not understand this, which value where from. C0 - C3 L0 - L3 Resistance Delays and if exist loss. Many high quality oslt has loss values. I think for amateur instrument more interesting the calibrator loss. I see many cheap connector have much higher loss than precision air dielectric oslt.

So, i cannot convert nanovna values to DG8SAQ WVNA and reverse. Another problem, on wvna original male oslt damaged by other user. Only female oslt in good conditions.

I alsa have some problem on Hugen like load. I see many time contact error. I heard from another nanovna user, his load also problematic. Short, and opens always good. I have there another sma load from used mobile phone system. No data, but i think all of them is good. Contact error never exist. I tried to use these loads for calibrations but i see only minor difference to Hugen like load.

I not know the best method to convert C1-C4 to simple C value.

73,
Lajos