Change in filter response starting in version r951

[John Golomb]

Hello Leif,

I was using r958 on cw over the weekend and noticed that something was not quite right with the sound of some CW notes.  I traced it down to response of the baseband filter.  I had it set to approximately 300-400 Hz wide.

The change in filter response starts in r951.  I have verified this still exists in latest r971.  The issue appears to have some dependence on FFT3 size.

Windows 10, Airspy HF+ discovery @768 kHz sampling.

Please see attached .pdf for some plots of filter response comparing r950 to r951.  I fed audio output to a separate instance of Linrad using VAC.

Let me know if you require any other parameter files.

Thank you for this fine software.

73,

John N2NC

[Leif Asbrink]

Hello John,

I do not think what you see is a problem in the filter response,
I think it is resampling spurs in rate conversion somewhere.

After applying the baseband filter Linrad resampled to a sometimes
much lower sampling rate to allow a lot of processing in the baseband.
(sophisticated CW decoding and other things.) I never completed
those efforts because digital modes made it uninteresting and
accomodating normal variations (and errors) in hand keyed CW turned
out to be very difficult. Then narrowbanded baseband, in your case
sampled at perhaps 1 kHz is up-converted to the sampling rate
of your soundcard.

I do not understand why a change in these things happened between
r950 and r951 but please send me your parameter files and a
screen dump (press "G" in linrad) and I will try to replicate here.

At the moment my work-bench is busy with a very slow project on
artifacts in phase noise measurements, but I will have a look
at the problem you have detected when finished with that.

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 on the web visit https://groups.google.com/d/msgid/linrad/48f7be0c-2d5a-4a08-ae3f-3ab7a4f1fb45o%40googlegroups.com.

[Conrad PA5Y]

Hello all. 

There is something that I have never really understood about the polarisation control. I think that it must ALWAYS assume that the 2nd RF channel is orthogonal to the 1st RF channel. So my question is this. Assuming you have reasonable phase accuracy on the RX path what should the ch 2 phase be set to if you have a 90 deg offset? Channel 2 will lead CH1 by 90 deg in my case.  I don't mean calibrating it on a known signal but as a starting point. I have never seen this documented or explained anywhere. I went to great lengths preserving the phase for both paths by individually testing each part of the system, firstly the cables and finally the LNA/BPF combintaion. In fact I was able to get the phase accurate by minor tweaks of one of the bandpass filters. This is almost certainly unnecessary I know but I just wanted to see if it could be done.

So I suppose I am talking to Leif.

Can you explain the exact procedure here? I just got an IQ+ and UADC4 combination on the system and therefore I have to do this again.

Regards

Conrad PA5Y

[Conrad PA5Y]

I mean a 90 deg physical offset due to the separation of the elements on the boom. Just for clarification.

Conrad

---

From: lin...@googlegroups.com <lin...@googlegroups.com> on behalf of Conrad PA5Y <g0...@g0ruz.com>
Sent: 17 August 2020 09:41
To: lin...@googlegroups.com <lin...@googlegroups.com>
Subject: [Linrad] Polarisation control for X mounted yagis

 

--

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 on the web visit https://groups.google.com/d/msgid/linrad/AM7P189MB1010B4455DB8624B0B92ABF7B55F0%40AM7P189MB1010.EURP189.PROD.OUTLOOK.COM.

[Leif Asbrink]

Hello Conrad,

the two antennas are supposed to be orthogonal, but that is not
really critical on the receive side. For transmit it is however.
Non-orthogonal antennas couple to each other and may give
elliptic polarization which is always a loss of power when the
other station is linearly polarized. (assuming you know what
tx polarization to set so you are not trying a 90 degree error.)

I always advocate X-configuration. Then the phase shift issue is trivial.
Listen to a DX signal in H-pol or V-pol. Point antennas towards each other
and tune hardware or Linrad parameters so you see the correct
polarization. That will give the correct amplitude ratio and the
correct phase.

With antennas in the + configuration it is far more difficult. You
will have to listen to EME signals at high elevation to get the
phase right at times when the signal is similar in both antennas.
Or, you can measure phases carefully and adjust for the offset
along the boom tube for the signal to reach a point in the forward
direction with the same phase.

In both cases which the sign of 45 degrees can be worked out from theory,
but I suggest you wait for a station that the tx polarization
calculator says you should send to on one of your antennas while
the received signal is equally strong in both. Then you will easily
see if you have the sign right on the report you get. With the
wrong sign you are likely to not be heard at all when transmitting
on the wrong antenna.

73

Leif

> To unsubscribe from this group and stop receiving emails from it, send an email to linrad+un...@googlegroups.com<mailto:linrad+un...@googlegroups.com>.
> To view this discussion on the web visit https://groups.google.com/d/msgid/linrad/AM7P189MB1010B4455DB8624B0B92ABF7B55F0%40AM7P189MB1010.EURP189.PROD.OUTLOOK.COM<https://groups.google.com/d/msgid/linrad/AM7P189MB1010B4455DB8624B0B92ABF7B55F0%40AM7P189MB1010.EURP189.PROD.OUTLOOK.COM?utm_medium=email&utm_source=footer>.

>
> --
> 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 on the web visit https://groups.google.com/d/msgid/linrad/AM7P189MB10102CE16E04229FB56A3BEEB55F0%40AM7P189MB1010.EURP189.PROD.OUTLOOK.COM.

[Conrad PA5Y]

Hello Leif thanks for getting back to me.

I am very happy with the TX side. My antennas are mounted + and - 45 degrees, I measured the phase accuracy of every single cable in the TX path and the errors are less than 1 deg. This was surprisingly easy to achieve. I measured orthogonality with an HP8752A and it is better than 30dB which is excellent so full marks to the manufacturer there. Previously I did some tests with several Xpol stations and circular gives 45 or 135 deg on their linear systems (at high ellevations) and when I switch between H and V the received pol shifts by 90 deg so that is fine.

I did try your suggested RX method with PA2CHR who has an Xpol array in the + configuration. He is 64km from me. It was difficult due to airplane scatter but I did end up with good results eventually. I have measured the RX LNAs and the cable losses and they are the same within a few tenths of a dB, however in order to get the RX polarisation calibrated I needed to have an amplitude offset of 1.4dB. Where does this come from? It bothers me and I would like to understand it. I do not expect the LNA gain to vary much from the bench measurements.

Due to the 4 relay switching arrangement for H/V and RHCP with a splitter I cannot easily TX on -45 or +45 individually.

I will have to recalibrate the RX because I am changing to an IQ+ UADC4 from an ME2XP dual transverter and Afedri set up.

Hence the subject is revisited.

Regards

Conrad PA5Y

To view this discussion on the web visit https://groups.google.com/d/msgid/linrad/20200830031446.3f643de3f5d4ea90b14bfd62%40sm5bsz.com.