SNR fluctuation

[Rick, W4XA]

Howdy all, 

I have been noticing a seemingly strange SNR fluctuation when using my MAC book Pro running Linux. (KDE NEON, Ubuntu based, using the Linux applmage) 

When talking to a station in Spokane from Western WA, I have noticed from observing the Reporter (and reports from the receiving station) that my SNR fluctuates from what I would call a fairly "normal" 10-15 dB and then momentarily dropping to 0, -1,-2,-3 Then back up to 10-15dB.

The receiving station reports an extremely LOW ambient noise level (S0-S2) and he tells me that my received signal is steady at S-9 or better all the time.

On my end, his signal is S-9 or better and his SNR at my end never fluctuates more than a few dB (20 or so +/- 3dB) 

When  I switch computers to a rather new ACER Core-i7 laptop running Winblows 11 or use my older (2015) MacBook Pro running Monterey,   His SNR is simlar on my end but MY SNR at his end no longer fluctuates at all or than a few dB 

It seems that my MacBook Running Linux is the problem.

Has anyone else seen something like this?

73/Rick

W4XA

 

[Mooneer Salem]

Hi Rick,

Does the other station hear anything odd with the audio when the SNR drops below 0? For example, any dropouts in the received audio.

Thanks,

-Mooneer K6AQ

--
You received this message because you are subscribed to the Google Groups "digitalvoice" group.
To unsubscribe from this group and stop receiving emails from it, send an email to digitalvoice...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/digitalvoice/1fcb6d74-b9d3-4f66-a3dc-7d218a068557n%40googlegroups.com.

[Rick, W4XA]

Hi Mooneer, 

Actually yes.  The signal strength as observed on the S-Meter doesn't drop or fluctuate much but the SNR at his end does.  Momentarily dropping from 10-12dB  to around 1 or 2dB  and then briefly 0, -1, -2, etc with a little garble before steadily climbing back up to around 10dB or so..  

He said he'd "lose" a couple of words when the SNR was down around  0, -1, -2 etc.

Receiving  on my end, his signal  (IC705 S-Meter)  stays at around S-9 to S-9+10 with little or no fluctuation. And the  SNR hovers around 20-24dB .

When I switched to either the ACER Laptop(WIN11) or the Early 2015 Macbook Pro (Monterey)  his received SNR didn't fluctuate a lot nor did it drop to near zero dB at any time.

It only seems to do this when I'm using the Mid 2012 MacBook Pro (early Intel Core-i7)  running KDE Neon (Ubuntu 24.04)  

I don't have a different laptop to try running Linux on so I am thinking I can blame it on the hardware , but I would like to try it with a different computer running Linux to verify that!

I've been noticing this for quite some time.   Where the SNR at the other end would fluctuate  from 10-15 down to 0  and then asking the other station if my actual received signal level was doing the same thing and he always told me that my signal doesn't fluctuate.   

 I always thought it was due to propagation effects (multi-path, NVIS phase distortion or maybe even a problem in my auto-tuner not making good contact in the roller inductor, or an antenna  etc) but it's always a one-way thing.

This is the first time where I have switched computers to see if it does it with a different computer.

I might try swapping the HD in the newer ACER Core-i7 laptop and installing Linux to see if it still does it.....That'll probably tell me right away if that old MacBook Pro has some other problem.

[Mooneer Salem]

Hi Rick,

Out of curiosity, does building the latest from "master" help at all? (Or alternatively, you can try the latest AppImage, but I know you've had problems with those before: https://github.com/drowe67/freedv-gui/actions/runs/23220637451/artifacts/5976028919) This is an updated version that no longer needs Python, so there should be a performance improvement.

Thanks,

-Mooneer K6AQ

To view this discussion visit https://groups.google.com/d/msgid/digitalvoice/697fb5d2-f189-4af6-ba80-0c9544be0934n%40googlegroups.com.

[Rick, W4XA]

Mooneer, 

I have never "built it" at all and I haven't tried the updated  image.  I don't have a Github login but I'll establish one and give it a try!

I am guessing that maybe the updated one might also work with my older ASROCK mobo and AMD 3 processor too?  Either way, I'll give it a try and let you know how it works!

Tomorrow morning, I am meeting with a small Radio Club locally to do a FreeDV demo at their meeting.  I'm going to bring my MacBook Pro running the MAC OS and my ACER Win11 laptop plus a Bluetooth speaker.

A couple of the Club  members that I have worked on 6m using FreeDV live within a few blocks and will run home so we can use 6m.   I have a small 6m wire dipole that I'll tape to the wall for the demo.  With the IC705 and a couple of watts it should work well using the internal battery!

Thanks again, 

Rick

[Rick, W4XA]

Mooneer, 

*Update*

I tried the 2.3.xxx version you suggested and #1, it now runs on my old AMD 3 Asrock computer!     I still notice some SNR fluctuation but I am pretty sure it was due to my running only 10W and late afternoon propagation!

I'll try it on the Linux/MAC laptop too but I suspect it's going to work well!

I also noticed that it launches MUCH faster and pressing "Start Modem" starts the modem a LOT faster!

 I didn't look too closely but, is that version also built for the MAC and Windows too?

[Mooneer Salem]

Hi Rick,

Glad it's working better for you so far! Yes, we're also automatically building for Windows and macOS. For example, here are the latest builds for both:

Windows: https://github.com/drowe67/freedv-gui/actions/runs/23267063947/artifacts/5995330121

macOS: https://github.com/drowe67/freedv-gui/actions/runs/23267063963/artifacts/5995308507

(Note: you'll need to run "xattr -d com.apple.quarantine" in Terminal on the ZIP file before decompressing as it's not signed.)

Thanks,

-Mooneer K6AQ

To view this discussion visit https://groups.google.com/d/msgid/digitalvoice/91b96539-e8ad-493f-bd95-7812829db350n%40googlegroups.com.

[Rick, W4XA]

Mooneer, 

Thanks!  I recently got an email telling me that if my CPU (in any computer) doesn't support avx2 CPU extended instructions, RADE wouldn't work.

So I did a little test with my current desktop (ASROCK AMD 3/3+  MOBO + AMD Phenom II 1090T 6-core CPU (which is 2013 vintage) 

https://cpu.userbenchmark.com/SpeedTest/417/AMD-Phenomtm-II-X6-1090T-Processor

Then I tested my MacBook Pro (Mid 2012 with an Intel Core-i7 mobile)

With this desktop: 

rick@rick-W4XA:~$ lscpu | grep -i avx
rick@rick-W4XA:~$

No output.  no support for avx or avx2

On the MacBook Pro, I did show support for avx but NOT avx2

But I was told that if I had no support for avx2, RADE wouldn't function (AT ALL)

Using that 2.3.x dev version you suggested , RADE works actually quite well on both computers!

So was the 2.3.x dev  version rewritten (in C=/C++? ) so as to allow RADE to work with a CPU lacking avx/avx2 support or was it never really required anyway?

Sorry for my ignorance of this stuff but I now have conflicting information!!

Bottom line,  BOTH computers (One of which worked and one that didn't before....) now work just great with RADE.

AND the program launches almost immediately, where before,  it took several seconds to start!

Thanks again for the help!

73/Rick

[Mooneer Salem]

Hi Rick,

I think AVX/AVX2 were needed before because of the Python libraries that the original RADE implementation used. Since we don't enable those as part of the build process for this version, that would explain why you're able to run it on older systems.

Thanks,

-Mooneer K6AQ

To view this discussion visit https://groups.google.com/d/msgid/digitalvoice/7bcb7d00-1142-43e6-ad62-9567cd565fa3n%40googlegroups.com.

[Rick, W4XA]

Mooneer, 

Just an update.    

The Win, MAC, and Linux DEV versions you  suggested that I try (all 3),  worked flawlessly on my ACER Win11, Macbook Pro (Monterey), MacBook Pro (KDE Neon) and desktop (ASRock AMD-3 Phenom II 1090T -- Linux Mint) computers.

The only additional thing I would ask for is distance filtering in the Reporter (("show stations @ < xxxx miles/KM etc)   I saw that this was already requested in a "Feature Request" 

Thanks again!!

73/Rick

[Dieter Pelz (Faustdownunder)]

Intermittend noise peaks ? For example, some electric kitchen tools can produce that. Other machinery as well. 

73

Dieter VK3FFB

[Rick, W4XA]

Hi Dieter, 

I thought that at first,  but I know the station in Spokane, WA pretty well and he is one of the very lucky ones. .... His noise level is pretty much ZERO at all times!  I with I had it that good!  I could understand if the signal was fluctuating going both ways but the actual signal level (S-meter reading) doesn't change a lot at all.

The only other thing I could think of is that I might have an antenna problem or a tuner problem (bad contact in the roller inductor etc) but the power output and VSWR doesn't fluctuate much at all.

[Dieter Pelz]

Understood. I did forget to mention the ongoing solar wind we currently have. Wind gusts can perhaps do that  although these are not usually of short term nature.

73

Dieter VK3FFB 

--
You received this message because you are subscribed to a topic in the Google Groups "digitalvoice" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/digitalvoice/VwoGmCmiHmA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to digitalvoice...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/digitalvoice/f17b132c-b0ae-4e22-b5ca-9a67836258d1n%40googlegroups.com.

[Rick, W4XA]

When this was happening before, we weren't being affected all that much due to Solar activity.

What I discovered was that the receiving-station  fluctuation was when I was transmitting and only when I was using my Mid 2012 MacBook Pro running the Linux operating system and the FreeDV version for Linux (2.2.1 linux Appimage) 

Using my fairly new Win11 ACER laptop or my other (early 2015) Macbook Pro running MACOS (12.x  "Monterey")  the receiving station wasn't experiencing  large excursions (10dB to -2.0dB ) of received SNR.

Now that I am running the 2.3.x dev version (I suppose it's something like a "nightly-build)  on my Linux-based MacBook Pro,   I am not seeing those large changes in  (other station's received) SNR

Seems to work better now.....but I am not sure what to make of this.