AGC still useless in slow mode in 1.3

  • 2
  • Problem
  • Updated 5 years ago
  • Not a Problem
  • (Edited)
AGC STILL NOT WORKING CORRECTLY IN SLOW MODE BEEN LIKE THIS FOR LAST TWO RELEASES
Photo of steve smith ka1sdr

steve smith ka1sdr

  • 38 Posts
  • 4 Reply Likes

Posted 5 years ago

  • 2
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9154 Posts
  • 3507 Reply Likes
Steve - please do not post in all caps.  If you have a complaint, please use good posting etiquette and describe your issue in detail.  This is like calling the doctor and saying "I hurt" without any additional information and expecting the doctor to cure you.

So  describe what is not working.  Is the the attack, the hold or the decay and how is it not working?
Photo of steve smith ka1sdr

steve smith ka1sdr

  • 38 Posts
  • 4 Reply Likes
when you are in a group of say three having conversation when one of the three stops talking and unkeys and one of the other three then keys up and speaks  you will miss the first couple words spoken untill the receive comes back up hope this is enough info for the doctor
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9154 Posts
  • 3507 Reply Likes
That helps.  What band(s) are you using?  What are the band conditions from a noise perspective?
Photo of steve smith ka1sdr

steve smith ka1sdr

  • 38 Posts
  • 4 Reply Likes
80 meters band is quiet listen to a group of guys talking put agc on slow you will hear it
Photo of steve smith ka1sdr

steve smith ka1sdr

  • 38 Posts
  • 4 Reply Likes
tim just checked it on 40 meters and it works fine there no problems but 80 very noticeable
Photo of Jim K4JAF

Jim K4JAF

  • 361 Posts
  • 105 Reply Likes

If the Slow mode is not the AGC action you desire, why not use the Med mode?

I am on 80 meters every evening and don't have a problem but I do keep the AGC-T down low and even some attenuation. 

Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 4149 Posts
  • 1325 Reply Likes
I have found it better than the last version, but it still needs a little work.  
There seems to be some kind of overshoot where noise pulses cut the gain down drastically.  I have not spent a lot of time analyzing it.

I wonder if the problem is not the length of the hang time, but the speed of the decay?
or the fact that even a very quick noise spike will fully engage the AGC, requiring the full length of time to reset?

I remember reading an article in Ham Radio Magazine (in the late 70's?) by Ulrich Rhode about a "new" hang AGC that allows the AGC to reset quickly after short noise spikes, but once a signal is present for a defined length of time then the longer "hang" is activated.  Then, instead of merely a long slow gradual decay, the "hang" would last for .5, 1, 2 or more seconds, then return quickly to full gain.  

I think the "custom" setting on PowerSDR allowed settings similar to this - Threshold, Hang, & Decay, though I saw no adjustment of the "activation time threshold." 

I would love to be able to adjust several of these timing parameters myself, perhaps merely a few entry boxes on the set-up tab.  I wouldn't need pull downs or sliders, just some manual entry number boxes.

Ken - NM9P
Photo of Jim K4JAF

Jim K4JAF

  • 361 Posts
  • 105 Reply Likes

The effects of the lightning spikes can be mitigated with adjustment of the AGC and attenuation.  I agree a "custom" setting would be a worthwhile feature.  Even is just the delay time is adjustable, it would be a great feature.  Most all of the newer higher end radios allow the agc delay time to be adjusted, usually on all AGC values, slow, medium and fast.  I do like the slow agc with long delay time as it lessens hearing the background noise as the AGC decays on a noisy band in general ragchews on 80 or 160.  But you do have to use a faster decay if you have very weak stations in the mix.  Jim K4JAF

(Edited)
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 9154 Posts
  • 3507 Reply Likes
Official Response
Steve - if the AGC is working for you on 40m and not on 80m, then the difference is the band noise characteristics interacting with the AGC and not the AGC algorithm itself.  In these cases where pulse type (fast rise time) band noise or simply higher levels of noise in general (which I think is what you are experiencing) is capturing the AGC, you can adjust the threshold setting (AGC-T) to a lower setting or use the medium (MED) setting for faster recovery.  I did some testing last night and based on the changes we made, the AGC is working as designed.
(Edited)
Photo of David Decoons wo2x

David Decoons wo2x, Elmer

  • 1297 Posts
  • 276 Reply Likes

I found on 80 during static crashes (like right now) that SSDR 1.2.1 SLOW mode was better than 1.2.17 and 1.3. While 1.3 has improved, the issue seems to be the hang before starting to release is too long for SSB and causes audio dropouts. This is with AGC-T at default 55 , 40 or 65. Trying to copy a signal below S9 results in audio degrading or dropping out some syllables.

1.21. slow would mitigate static crashes but not eliminate them. It did not degrade the audio.

I would suggest that they incorporate 5 settings, like PSDR off, fast, med, slow, and long.

Improvements I would suggest are to have a little faster release with a more gradual decay rather than long release with fast decay.

I can do a recording of how 1.3 degrades the audio for those that are interested.

Dave, wo2x


Photo of Steve N4LQ

Steve N4LQ

  • 568 Posts
  • 92 Reply Likes
Just curious....Does the 6000 series have any hardware AGC or is it all in DSP?
Photo of Gerald - K5SDR

Gerald - K5SDR, Employee

  • 825 Posts
  • 1503 Reply Likes
It is 100% DSP.