search

jankae / LibreVNA

100kHz to 6GHz 2 port USB based VNA
GNU General Public License v3.0
1.09k stars 207 forks source link

Spikes at 31MHz and 62MHz #215

Closed Dmitrii1978 closed 11 months ago

Dmitrii1978 commented 1 year ago

LibreVNA Version

LibreVNA Version (64 bit): 1.5.0-alpha.2- OS: Windows 11 Version 22H2 CPU Arch: x86_64

Steps to reproduce

Spikes at 31MHz and 62MHz and higher multiples frequencies. spikes

Expected behavior

Is it possible to decrease level of these spikes?

Extra information & Setup and Calibration files

The frequency range is set to 1MHz..101MHz , 4501 points, IF BW 1.00kHz

jankae commented 1 year ago

These spikes are caused by LO feedthrough (and mixing products of the two LOs). I don't think it is possible to remove them without drastic changes to the hardware, sorry. You have the option to shift them to slightly different frequencies by changing the LO frequencies: Device->Configuration grafik

This could be useful if you need to look at the actual signal right at these frequencies. But I would recommend changing the LO only by 1-2 MHz, otherwise the hardware filters may attenuate the signal too much (increased noise).

YuryW commented 1 year ago

LibreVNA Version

LibreVNA Version (64 bit): 1.5.0-alpha.2- OS: Windows 11 Version 22H2 CPU Arch: x86_64

Steps to reproduce

Spikes at 31MHz and 62MHz and higher multiples frequencies. spikes

Expected behavior

Is it possible to decrease level of these spikes?

Extra information & Setup and Calibration files

The frequency range is set to 1MHz..101MHz , 4501 points, IF BW 1.00kHz

Hi, Dmitrii1978. I have a slightly different picture in this frequency range. W7_64_LibreVNA_1MHz_101MHz_-10dBm_4501_1kHz_

It is possible that I have a different LibreVNA device (six-layer board and added LO harmonic filters). https://github.com/jankae/LibreVNA/issues/161#issuecomment-1372602939

Perhaps such filters already produce better results than yours. Often bursts can be significantly reduced by changing one frequency limit, for example, the lower one at 250 kHz. This is my result. W7_64_LibreVNA_1 25MHz_101MHz_-10dBm_4501_1kHz_

Even changing the number of scanning points by one helps to shift the parasitic bursts in frequency. W7_64_LibreVNA_1 25MHz_101MHz_-10dBm_4501_20Hz W7_64_LibreVNA_1 25MHz_101MHz_-10dBm_4500_20Hz

But how many parasitic bursts there are even with harmonic filters of the first local oscillator, if you use scanning boundaries without a shift of 250 kHz, that is, as in the case you presented. W7_64_LibreVNA_1MHz_101MHz_-10dBm_4501_1kHz

And of course, reducing scanning points can completely eliminate parasitic bursts. W7_64_LibreVNA_1 25MHz_101MHz_-10dBm_450_20Hz