Note: groups.io will be down for maintenance this evening, starting at 2AM Pacific Time (9AM Friday 10/17/2019 UTC), for up to two hours.
Thank you all for your kind words and valuable feedback.
I will then leave the idle PA current near 500mA per device despite the current drain.
I have contemplated adding some series capacitor/resistors in parallel to the emitter resistors in the driving stage or pre-driving stage either at R911/R941/R942/R96 or R87/R88 to increase the gain in the upper frequencies.
If someone with an oscilloscope could determine where the saturation occurs along the chain after RV1 when this pot is turned up, it would help decide where to put the frequency compensation.
But based on data on the web and this group's feedback it seems that extracting 16 to 20W from a pair of RD16HHF1s at 13.8V is pretty much it, so the uBitx design with 13.5W out is not that sub-optimal from that regard when we substitute the finals IMHO.
I have an SSM2167 board on order hoping that the compression will deliver the 3 to 6dB average power gain we seem to see with good compressors. I think it beats doubling or quadrupling the power on SSB, especially with a flatter power curve like the one I got.
All the best,
73, John (VK2ETA)