TI power stage for selfoscillating amplifier

Ptero power mono
TAS5162 PBTL, 280W@4Ohm
 

Attachments

  • 20240512_164202.jpg
    20240512_164202.jpg
    365.1 KB · Views: 25
  • 20240512_164153.jpg
    20240512_164153.jpg
    381.5 KB · Views: 26
I need to ask this question because it appears to be unexplored.

Why does the amplifier have an adjustable frequency? In my recent simulations, I observed that randomly sweeping frequencies—or in your case, toggling between two frequencies—significantly influences the "system's overall transfer response" control theory domain. By selecting an "idle target frequency" and designing the poles and zeros around this point for different operational states, the "final set point idle frequency" is established. However, this necessitates readjusting the LC double pole at the output. Given these complexities, I'm curious why these frequency sweeps are conducted as a production "user settable" option, as they seem to complicate the predicted behavior of the control loop.
 
In this case, I didn't notice an impact in the audio range.
The frequency boost option activates the feedback before the filter, as it is realized in the original ET400 modules.
But since the delay in my circuit is much lower than in the ET 400, even without the additional boosting OS, the frequency is 600-640kHz
 
Interesting, I think the influence of the loop are during "modes" and NOT audiophile listening evaluation.
I read the paper and noticed some areas of interest like clip recovery and current trip under operations such as loud playback or even shorting the output.

- 600-640kHz I cannot seem to get an IRS20957 (due an already internal 350ns delay + 10ns comp delay) using an NCORE modulator to go that high, not without 100p/220p lead/lag networks, I don't like 10^-9 pico farad values in loops they seem to be sensitive to finger touching drift.