Calibration with RSPduo

[Scott T]

Trying to calibrate my RSPduo with a pulse generator at 50 MHz (because I have terrible noise on 6m).

I am fortunate enough to have a great deal of test gear, so I am using a pulse-modulated signal generator with the RF center frequency set to 50.125 MHz.

I am able to generate pulses with 30 dB SNR, although I have to increase the PRF to 300 Hz in order to get to the "collect pulses" screen.  RSPduo settings are:  fs=2000 kHz, FFT1 BW=100 Hz, FFT1 size=32768, ZIF mode with 200 kHz IF bandwidth.

Once I do get to that screen, it appears that the LO/DC feedthrough is right in the middle of the magenta amplitude envelope...and the help warns against narrowband signals in the passband due to phase discontinuities--and I do seem to have a lot of those! (Pic attached).

Question 1:  should that LO/DC feedthru even be there?  In the RSPduo setup I do select DC offset correction and IQ imbalance correction (parameter file attached).  In any event I am never offered the calibration A task (Calibrate I/Q phase and amplitude), even when I deselect the aforementioned "self-correction" items.

Question 2:  is the solution as simple as offset tuning?  Which leads to:

Question 3:  what is the magenta passband?  IF bandwidth?  Baseband bandwidth?  FFT2 bandwidth?  Or ?

Thanks, so far a great learning experience and I hope to get over the goal line

[Leif Asbrink]

Hello Scott,

Something is very wrong.Have a look at this video:
https://youtu.be/rPQARPt6r_8?t=1419
How does your first screen look, the one at 23:39 in the video?
Maybe your calibration pulse is too long? It must be significantly shorter
than 1/fs, no longer than 100 ns in your case.

Do you have an oscilloscope that allows you to look at the pulses?

If you de-select the "self-correction" items you should get the
calibration A task (Calibrate I/Q phase and amplitude) The reason you do
not get it is that sdrplay3.c sets the flag DIGITAL_IQ in the variable
ui.rx_input_mode. That tells Linrad to not manage mirror images.
I am not the author of the RSPduo support in Linrad so I can not
easily correct that.

When you try to calibrate the pulse response it might be possible that the
IQ balancing in the RSPduo causes changes in the impuse response
during the calibration procedure. A good idea could be to de-select
the "self-correction" items to see if you can get proper calibration then.
That should be possible although you would then have image spurs. Maybe
you could then apply the automatic image correction without loss of
blanker performance(?)

When you press enter on the first screen you should get something like
the screen at 24:17 in the above video (although you should see two
graphs for two RF channels.) The image you sent does not resemble
that at all. It seems to me that you have used the next screen to make
a parameter update that was totally wrong.

Question 1: The center peak should be there, I do not know whether
the automatic DC suppression would remove it.

Question 2: I do not know if offset tuning will help.

Question 3: The green trace is phase vs frequency and the
magenta trace is amplitude vs frequency.

In this video at 16:28 you can see the typical response for a direct
conversion radio using a soundcard: https://youtu.be/WrWFZVbs79I?t=988
The amplitude is zero near the midpoint because the soundcard is
AC coupled. At the exact center there is a peak because of the
DC offset of the signal. The phase drifts away as the amplitude
goes down near the center. That is because of the phase shift
in the highpass filter that blocks DC to the soundcard.

Videos on calibration and setup are listed here:
http://sm5bsz.com/linuxdsp/videos.htm#setup

73

Leif

> --
> There is an excellent Linrad User Guide by Gaetan, ON4KHG, at:
> http://w3sz.com/Linrad%20Installation%20&%20Configuration%20User%20Guide%20-%20V1-0.pdf
> ---
> You received this message because you are subscribed to the Google Groups "Linrad" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to linrad+un...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/linrad/386a92d1-62bc-4b57-b11b-a6f70d7ca3fbn%40googlegroups.com.

[Earl Shaffer]

Hi  Scott.

Linrad is very good at removing power line noise and to some degree, precipitation static. It has been about 15 years since I have been able to detect power line noise because the broadband noise that is created by just about every modern device like LED lights or switching power supplies is much stronger. Linrad even with a properly calibrated smart noise blanker does not blank it. In fact if the dumb blanker does not help your noise the smart blanker is not likely to do so either. If I understand it correctly the smart blanker which requires the calibration you are trying to do helps the noise blanker work when strong signals are present which the dumb blanker can fail at. You can look at Linrad in the oscilloscope mode and decide if there are large amplitude noise  pulses present that might be blanked. I live in an Urban environment and I have not been able to get Linrad to be of any help for 2 meter EME. Fortunately with a very low sidelobe array, I can still work EME at high elevations where I reject terrestrial noise. I have sent files with a properly calibrated system of my noise to Leif and he was unable to do anything useful with them. My noise consists of many different sources of spread spectrum noise. I believe manufacturers frequency modulate their noise sources to sidestep FCC regulations. Spreading the noise over a wide range of frequencies can allow them to pass FCC limits but it also ensures that we can not get away from those noise sources by hunting for a clear frequency. There ARE none. In theory with advanced tech, even these LED lights might be blanked but we are not there yet.

IMHO your best solution is to find the source of your noises and remove them. The second best would be to use an antenna with lower sidelobes. If you have a single outstanding source of noise that you can not blank, you may be able to combine a noise antenna in the proper  phase and amplitude to cancel it. That would not work here because I am assaulted by noise from many directions and many sources. 25 years ago it was much different.

73, WB9UWA.

--
There is an excellent Linrad User Guide by Gaetan, ON4KHG, at:
http://w3sz.com/Linrad%20Installation%20&%20Configuration%20User%20Guide%20-%20V1-0.pdf
---
You received this message because you are subscribed to the Google Groups "Linrad" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linrad+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/linrad/386a92d1-62bc-4b57-b11b-a6f70d7ca3fbn%40googlegroups.com.

--

Earl Shaffer,  WB9UWA

 

My EME array photos

https://www.dropbox.com/sh/bbn2ktonko8e3az/AACHBBOhf6Djk5Pf2c86Br5ja

Detailed array photos

http://www.gm4jjj.co.uk/WB9UWA/

Facebook

http://www.facebook.com/earl.shaffer

Linkedin
http://www.linkedin.com/pub/earl-shaffer/12/881/735