Hello all, first, let me wish you Happy Christmas Holidays! And then, a little back on-topic.
I would like to return to maximum available power from Purifi modules with respect to SMPS used, module used, and frequency. My plots are based on 1ET400A + SMPS400A180 measurements, as I have nothing else here, however, I am quite sure that the results can be extrapolated to another Purifi and SMPS modules as well.
As a start, we have datasheet plots of output power vs. supply voltage, which says:
With 2 x 45V PSU voltage, the graph says we shall get 100W into 8ohm at 1% THD, signal frequency is not specified. My SMPS400A180 gives 2 x 48V with Purifi module at idle and about 2 x 46V at maximum power with 1% THD.
First, measurements of distortion vs. output voltage into 8ohm at 20Hz, 1kHz and 10kHz with measuring bandwidth 45kHz:

As we can see, the maximum power varies with signal frequency, and is highest for 10kHz and lowest for 20Hz. Where is the reason, is it the SMPS or Purifi module? We need t investigate SMPS rail voltage with the module operating near its maximum power.
1) at 1kHz we get this record of SMPS rail voltage:
DC voltage is almost 46V and the rms ripple voltage is about 0.5Vrms.
2) at 20Hz we get this record of SMPS rail voltage:
DC average voltage is 46.4V and rms ripple is 1V. So yes, minimum SMPS voltage seems to drop lower than with the 1kHz signal.
Finally, the record from 1ET400A at 20Hz and approximately 1% THD:
We get 104W/8ohm/20Hz with 34.5 dB SINAD and 41.14 dB SFDR, so we are close to THD 1% and also close to datasheet plot here above. Regarding dependence of maximum power on signal frequency, in my opinion both the Purifi module and the SMPS play their game here. It is evident that we get most power at high frequencies (at 10kHz we get 32V/8ohm, i.e. 128W/8ohm with THD<1% !), though the difference is unimportant regarding perception.