AGC in AirSpy HF+ #airspyhfplus #airspy


sm6fhz
 

Hi.
Can someone describe the difference between the AGC-Low and AGC-High in AirSpy HF+ and Discovery?
When to use one or the other?
Any specific use cases?
Different implementation?
I have not seen any description of the differences and I am curious about it.
I have not been able to notice any differences between them either. Maybe because I do not know what to look for.
73 / Ingolf, SM6FHZ


sm6fhz
 

Thanks Vince.
Mystery cleared.
/ Ingolf


Kriss Kliegle KA1GJU
 

FWIW,

This is what happens to mine with it Low to Off, and back to Low again:



YMMV

73 Kriss KA1GJU


AM BC stations galore!


Jim Smith G0OFE
 

I have had similar experiences, especially on LF.

It's wierd... sometimes if the AGC is off, I get what looks like overload and images etc on, say, 160m, so I turn it on. QSY to 40m and I get again, what looks like oveerload and images, so I turn the AGC on, and everything is fine.

================================
Jim, Bournemouth, Dorset. IO90BR.
Elad FDM-S2, Airspy HF+, SDRPlay and Funcube Dongle SDR Recievers
8-element OP-DES for Band 2, 8-element LFA for 2m, 3-element LFA for 6m, 20m Windom for HF,. Wellbrook 1530 Loop for LF


On 21/01/2020 22:45, Kriss Kliegle wrote:
FWIW,

This is what happens to mine with it Low to Off, and back to Low again:



YMMV

73 Kriss KA1GJU


AM BC stations galore!


Alan
 


It's wierd... sometimes if the AGC is off, I get what looks like overload and images etc on, say, 160m, so I turn it on. QSY to 40m and I get again, what looks like oveerload and images, so I turn the AGC on, and everything is fine.

Jim,

Not really weird, the ACG sets the gain correctly.
If you turn it off YOU have to set the gain correctly.

73 Alan G4ZFQ


Kriss Kliegle KA1GJU
 

The ATT works as stated in the GUI, AGC must be off for it to function... But, the Pre Amp states the AGC MUST be off, yet it works with the AGC on or off. 🤷‍♂️

73 Kriss KA1GJU


Larry Dighera
 

Werner (DC4KU) tested the Airspy HF Discovery.and found its AGC
operation unsatisfactory:
https://groups.io/g/airspy/wiki/Werner-%28DC4KU%29-tested-the-Airspy-HF--Discovery.and-found-its-AGC-operation-unsatisfactory

"Unfortunately, the Airspy HF + Discovery shows the same faulty
AGC hysteresis as the HF-Airspy HF+. If the SDR is operated on a
powerful antenna, quite large signals may occur at the input of the
SDR, so that the Airspy after that remains in its down-regulated state
and small can no longer be detected."

On Wed, 22 Jan 2020 05:24:13 -0800, "Kriss Kliegle"
<kliegle@myfairpoint.net> wrote:

The ATT works as stated in the GUI, AGC must be off for it to function... But, the Pre Amp states the AGC MUST be off, yet it works with the AGC on or off. ????

73 Kriss KA1GJU


Alan
 

Larry,

I note that firmware 3.02 was installed in the one HF+ Discovery that was tested. AGC revisions have been included up to the current "R3.0.6 - 2019-08-20 Optimized the high AGC threshold". Also the software has been updated.

Saying that I have had a HF+ since it was introduced and now a Discovery and have not noticed the effect mentioned.

Neither has it been mentioned on the Airspy group. So I presume if it was present it would be a very uncommon occurrence.

73 Alan G4ZFQ

Werner (DC4KU) tested the Airspy HF Discovery.and found its AGC
operation unsatisfactory:
https://groups.io/g/airspy/wiki/Werner-%28DC4KU%29-tested-the-Airspy-HF--Discovery.and-found-its-AGC-operation-unsatisfactory

"Unfortunately, the Airspy HF + Discovery shows the same faulty
AGC hysteresis as the HF-Airspy HF+. If the SDR is operated on a
powerful antenna, quite large signals may occur at the input of the
SDR,


jdow
 

On low HF with a good antenna detecting small is not a function of the noise figure of the receiver. You have 20 dB or so of margin. On 75 meters I always run my ProII with 20 dB attenuation inserted on RX, for example. The world is NOISY down there and believe it or not most of that noise is not man made.

{o.o}

On 20200122 06:58:14, Larry Dighera wrote:
Werner (DC4KU) tested the Airspy HF Discovery.and found its AGC
operation unsatisfactory:
https://groups.io/g/airspy/wiki/Werner-%28DC4KU%29-tested-the-Airspy-HF--Discovery.and-found-its-AGC-operation-unsatisfactory
"Unfortunately, the Airspy HF + Discovery shows the same faulty
AGC hysteresis as the HF-Airspy HF+. If the SDR is operated on a
powerful antenna, quite large signals may occur at the input of the
SDR, so that the Airspy after that remains in its down-regulated state
and small can no longer be detected."
On Wed, 22 Jan 2020 05:24:13 -0800, "Kriss Kliegle"
<kliegle@myfairpoint.net> wrote:

The ATT works as stated in the GUI, AGC must be off for it to function... But, the Pre Amp states the AGC MUST be off, yet it works with the AGC on or off. ????

73 Kriss KA1GJU


Larry Dighera
 

Hello Alan,

Thank you for providing your keen observation and analysis. I will
pass this information on to Werner.

I have observed that the moderator(s?) of the Airspy group regularly
censor posts that are potentially derogatory toward their products.

Best regards,
Larry


On Wed, 22 Jan 2020 15:51:22 +0000, "Alan via Groups.Io"
<alanzfq=yahoo.co.uk@groups.io> wrote:

Larry,

I note that firmware 3.02 was installed in the one HF+ Discovery that
was tested. AGC revisions have been included up to the current "R3.0.6 -
2019-08-20 Optimized the high AGC threshold". Also the software has been
updated.

Saying that I have had a HF+ since it was introduced and now a Discovery
and have not noticed the effect mentioned.

Neither has it been mentioned on the Airspy group. So I presume if it
was present it would be a very uncommon occurrence.

73 Alan G4ZFQ


Werner (DC4KU) tested the Airspy HF Discovery.and found its AGC
operation unsatisfactory:
https://groups.io/g/airspy/wiki/Werner-%28DC4KU%29-tested-the-Airspy-HF--Discovery.and-found-its-AGC-operation-unsatisfactory

"Unfortunately, the Airspy HF + Discovery shows the same faulty
AGC hysteresis as the HF-Airspy HF+. If the SDR is operated on a
powerful antenna, quite large signals may occur at the input of the
SDR,


Alan
 

I will pass this information on to Werner.
Larry,

After re-reading the article I wondered why he had not tested with the (Faulty?) AGC turned off. This option is useful for using the HF+ for measurement and possibly for being measured. After all the AGC is at least partly there to make the Airspy easy to use.

I have observed that the moderator(s?) of the Airspy group regularly
censor posts that are potentially derogatory toward their products.
Yes, a few have complained their posts are not allowed but I'm not sure of the content.
On many groups there are posts from people who make bad judgements that might make a product seem faulty when it is not.
I guess I'd be wary of letting ill informed posts from such people make my product seem bad, and that's what "prog" has said in reply.

73 Alan G4ZFQ


Larry Dighera
 

My comments in-line below:

On Thu, 23 Jan 2020 14:56:10 +0000, "Alan via Groups.Io"
<alanzfq=yahoo.co.uk@groups.io> wrote:

I will pass this information on to Werner.
Larry
Werner's response:


To: Larry Dighera <LDighera@att.net>
Subject: Re: [SDR-Radio] AGC in AirSpy HF+ #airspyhfplus #airspy
From: dc4ku <dc4ku@darc.de>
Date: Thu, 23 Jan 2020 17:50:57 +0100

Adam Farson stopped all his measurement at the Airspy HF + and
sold it.
https://www.ab4oj.com/sdr/airspy/airspy_hf_notes.pdf
Best regards,Wernerdc4ku


And Adam responded with this comment:


To: "'Larry Dighera'" <LDighera@att.net>
Subject: RE: [SDR-Radio] AGC in AirSpy HF+ #airspyhfplus #airspy
From: "Adam Farson" <afarson@telus.net>
Date: Thu, 23 Jan 2020 06:45:04 -0800

Hi Larry,

My test report confirms Werner DC4KU's findings on AGC hysteresis
and indifferent NPR test results.

I sold my Airspy HF+ a few days after completing my tests.

73, Adam VA7OJ/AB4OJ


Perhaps they will grace us with the reasons they did not also test
with the AGC off, as you have suggested, and redeem their perception
of impartiality. Here's hoping...

As a point of information in this topic, there is this (Note: Firmware
version unstated, but it was probably an early one as used by Werner):


https://groups.io/g/airspy/message/31749
Re: Discovery - First Light
prog
08/13/19 #31749

On Wed, Aug 14, 2019 at 08:00 AM, <dx@...> wrote:

I'm curious about the "HF AGC Threshold High" feature. Under
what circumstances is it best to use this feature?

Level "Low" tells the AGC to start adding attenuation 6 dB earlier
than level "High" as the signal power increases.
Basically, level "High" gets you more sensitivity while "Low" gets
more linearity.

Note that the tuner has unit uses hysteresis to prevent the ATT
frop jumping back and forth between two successive levels. This
means you will have to wait a few seconds to see the effect of
changing the threshold.


After re-reading the article I wondered why he had not tested with the
(Faulty?) AGC turned off. This option is useful for using the HF+ for
measurement and possibly for being measured. After all the AGC is at
least partly there to make the Airspy easy to use.
I have CCed Werner on this message in the hope that he may answer your
astute question.


I have observed that the moderator(s?) of the Airspy group regularly
censor posts that are potentially derogatory toward their products.
Yes, a few have complained their posts are not allowed but I'm not sure
of the content.
On many groups there are posts from people who make bad judgements that
might make a product seem faulty when it is not.
I guess I'd be wary of letting ill informed posts from such people make
my product seem bad, and that's what "prog" has said in reply.

73 Alan G4ZFQ
In my opinion, implicit in censorship of on-topic, reasonably accurate
information that can be seen as derogatory toward the product is an
attempt at deception. Such lack of impartiality is not an admirable
quality in a moderator (or judge either, to wit Moscow Mitch). Rather,
such posts might be seen as an _opportunity_ to respond with
additional product information that enlightens and informs an opposing
view, and exonerates the product, as your point about firmware version
has somewhat accomplished. It is disappointing when emotion overrides
their otherwise rational actions. Perhaps I expect too much ...

To be fair, perhaps this moderator trait may be fueled by having been
inculcated in a local culture that lacks a guarantee of the right to
free speech.

Disclaimer: I am a supporter of Airspy products who owns several. I
wish them success and look forward to their future products.


Alan
 

Larry,

I did some crude tests and it does seem the AGC does not make the best choices at high levels of a single signal. I did not make it lock up.

I am not able to make any laboratory grade tests but feel it works well in practice. I do not think I've noticed any strange AGC effects in use. However I do often use manual attenuation preferring to have some idea of the signal levels.


Perhaps they will grace us with the reasons they did not also test
with the AGC off, as you have suggested, and redeem their perception
of impartiality. Here's hoping...
I would be interested if someone could do this.

In my opinion, implicit in censorship of on-topic, reasonably accurate
information that can be seen as derogatory toward the product is an attempt at deception. Such lack of impartiality is not an admirable quality

As I said I do not know what was said in the censored messages. Certainly some ill informed derogatory remarks can put people off. But the knowledge that censorship has been used is troubling.

Disclaimer: I am a supporter of Airspy products who owns several. I
wish them success and look forward to their future products.

Agreed

73 Alan G4ZFQ


Larry Dighera
 


Hello Alan,

Thank you for your additional tests.

Werner has been kind enough to further elucidate his findings in his e-mail message below.  Thank you both for providing further data regarding these issues.

------------------------------------
To:       <LDighera@...>
Subject:  Airspy HF+ and Airspy HF+ Discovery - AGC-Error
From:     "Werner Schnorrenberg" <Werner.Schnorrenberg@...>
Date:     Mon, 27 Jan 2020 14:47:52 +0100


Hi Larry,

I try to explain it again.
I had already discovered the AGC error in the Airspy HF+ and reported it to Airspy. But Airspay did not answer. Then later I tested the new HF+ Discovery and hoped that the error had been fixed. But this was not the case; it shows the same AGC error. Here is my IFSS, IMD3 curve.


If you increase the input 2-tone signal from -70dBm to -10dBm (green) and then reduce it in the same steps back to -70dBm, the declining curve goes completely wrong (red). The IMD3-curve no longer goes back to its high sensitivity of at -141dBm, but remains at -120dBm. It looks like an "ADC blocking". Afterwards the receiver will not return to its original setting. Therefore you have to switch off the receiver completely and start it again to get back to the original settings (high sensitivity). If you change HF-Threshold from Low to High, the error can be reduced a little bit, but an AGC hysteresis remains.

If you do not measure the curve in that way, you will not understand the AGC-error. This is the problem, many users don't even notice the error, because a remaining sensitivity of -120dBm/500Hz  on the 160m, 80m or 40m band is enough, because a good antenna produces much more noise.  I suspect that even Airspay did not perform this kind of measurement and therefore does not know about it. Maybe Airspay made this setting on purpose (which I can't imagine) but then the company would have to describe and explain it.

In the next two pictures you see a 2-tone signal of the Airspay HF+, both generated with constant 2 x-35dBm generator level. In the left picture the intermodulation distance is 75dBc (not so good)  and in the right picture 100dBc (good value). But which value is correct? 75dBc intermodulation distance is created when I reduce the 2-tone level down to 2x-35dBm and 100dBc distance when I increase the level up to 2x-35dBm. I haven't seen such a control error with any receiver so far. This must not happen.

     
              
Pi=2x-35dBm, Delta IM3=75dBc                                                                   Pi=2x-35dBm, Delta IM3=100dBc        

Another resulting problem is that important specs of the receiver such as SBN (Side Band Noise) and RMDR (Reciprocal Mixing Dynamic Range) can no longer be determined due to the undetermined sensitivity.

When I ran my test suite I first downloaded and installed the then-current firmware.

Because the Airspay HF+ and the Airspy HF+ Discovery shows the same kinds of errors, the receivers are not usable for me and I sold them both.

73, Werner

DC4KU

27 January, 2020

http://www.dc4ku.darc.de/airspy-hfplus.pdf

http://www.dc4ku.darc.de/AIRSPY%20HF+%20DISCOVERY%20-%20Testpdf

http://www.dc4ku.darc.de/AIRSPY%20HF+%20DISCOVERY%20-%20Test_English.pdf


 [PDF attached]

----------------------------------------------------------------------------

On Saturday, January 25, 2020, 7:38:05 AM PST, Alan via Groups.Io <alanzfq@...> wrote:


Larry,

I did some crude tests and it does seem the AGC does not make the best
choices at high levels of a single signal. I did not make it lock up.

I am not able to make any laboratory grade tests but feel it works well
in practice. I do not think I've noticed any strange AGC effects in use.
However I do often use manual attenuation preferring to have some idea
of the signal levels.


> Perhaps they will grace us with the reasons they did not also test
> with the AGC off, as you have suggested, and redeem their perception
> of impartiality.  Here's hoping...
I would be interested if someone could do this.

>In my opinion, implicit in censorship of on-topic, reasonably accurate
information that can be seen as derogatory toward the product is an
attempt at deception. Such lack of impartiality is not an admirable quality

As I said I do not know what was said in the censored messages.
Certainly some ill informed derogatory remarks can put people off. But
the knowledge that censorship has been used is troubling.

>Disclaimer: I am a supporter of Airspy products who owns several. I
wish them success and look forward to their future products.

Agreed

73 Alan G4ZFQ