Open marcusmueller opened 1 year ago
I have been learning and testing about agc for a few weekends. @argilo asked for a real world test signal so, combining a fm mod block and a delay i created a periodic signal whose amplitude bounces from 1.0 to 0.08. This would make agc to stress trying to set the correct gain.
To solve the problems with 512 bit machines, i skipped the first period of the signal and and took the second one to perform the test. i absolutely cant understand this issue but the solution is similar to the proposed in test_006_003_agc3
I would love to hear some comments about the changes.
What happened?
What happened? Bugs happen, that's what :) Thus, we need good tests. Currently, and even after #6274, the AGC3 tests are so rudimentary it constitutes a bug.
System Information
n/a
GNU Radio Version
3.11-git (main)
Specific Version
v3.11.0.0git-317-g5bbe7bd1b
Steps to Reproduce the Problem
Relevant log output
No response