attenuating USRP gain post-AGC processing

Patrick Perdue


I have a bridge set up using AGC with the following parameters:

usrpAudio = AUDIO_USE_AGC                 ; Digital -> Analog (AUDIO_UNITY, AUDIO_USE_GAIN, AUDIO_USE_AGC)
usrpGain = 0.50                         ; Gain factor when usrpAudio = AUDIO_USE_GAIN (0.0 to 5.0) (1.0 = AUDIO_UNITY)
usrpAgc = -19,10,200                    ; Set the agc threshold (db), slope (db) and decay (ms)
This sounds great, but I want to reduce the output of the processed audio by just a couple DB going into ASL. It doesn't appear that the usrpGain value does anything when set to usrpAudio = AUDIO_USE_AGC, which, I believe, is expected behavior.

Yes, I could raise the threshold to something closer to 0, but I want the processing as-is, just with a trim on the output. Actually, I would like to possibly make the processing even more aggressive (lower threshold and longer release) to comp for some lower levels and keep higher ones in check), but then reduce that output by, say, 3 to 6 db coming into ASL's USRP driver. Is this at all possible with one instance of A_B, or can I use a second instance of A_B just as an attenuater? With older versions of A_B, this could be done, but the AGC algorithm was practically useless.

Mike Zingman - N4IRR

I added the gain block after the AGC processing, so this should address your needs.  Steve will release soon as there are some other changes in the pipeline as well.