No, I have been too busy with other things and now that days are warming up a bit again (and my shed is not so cold), I am planning to get back to finishing some other projects before tackling the Aleph 2.
I have looked at the option to replace the transformers, but I have perfectly fine 1kW Talema transformers and don't have anything planned for those, so I think I will keep them.
Reading through the many, many post I realised that moving the transformers out of the amp boxes would probably be a good idea. Leave the last C of the LCL psu in the amp box and add a large motor run cap (removing the other stuff will give me the space I need), and put the transformer plus the first C and L in a dedicated psu box. And then I realised the extra 10V I have may actually be a blessing. To burn off the 10V I could just add a power resistor somewhere in the psu chain, but a regulated psu might be a better option. In either case, I will need a box with heatsinks!
I plan to use the Aleph 2s on 4 Ohm speakers that are quite current hungry, so am not completely sure yet. The link above to the spreadsheet (I have it in my Aleph directory but had not looked at it for a while) has made me rethink my needs. In my case, the 12 fets per channel are already there, so I don't think I will increase them. When I redo the front end using the boards I got from @rhthatcher last year, I will have better/easier control over bias current and AC current gain.
I have an original Aleph 3 and like it more than the Aleph 2 in their current state. The 3 is of course highly biased, so perhaps I need to increase the bias to get the sound I would like. I have a variac, so can play around with the rail voltage in the current (unregulated) setup before deciding.
As for the regulation, I a still thinking of options. I remembered the Aleph P1.7 uses a simple regulation circuit that should be easy to copy, although I am not sure if it would work (upscaled of course) for an Aleph power amp.
A rough calculation would be that I need to drop 10V per rail. Assuming 3 or 4 amps of bias, that would result in 30-40W heat per rail, so at least 2 parallel 240s for the positive rail (instead of the 610 used in the P1.7).
I am not sure if I need to mirror the setup (using PNP rather than NPN) for the negative rail, or can get away with using the same circuit. I remember a similar issue with an LM317 base psu, where two LM317 circuits could be used to make a symmetrical (+ and -) psu.
Also, I had noted some time ago that in the current LCL setup I had a 100mV ripple at the first 47000uF C and only 5mV at the second. Adding a power resistor will bring that down even more, so perhaps full regulation is overkill?
So, long answer to say that No, I haven't solved the issue yet and, even worse, have not yet decided what way I will go ahead. I have the luxury of not being in a hurry because I already have too many amps
Regulated PSU on Aleph 2?!?!? I'd say that's mega overkill, and will be pretty tricky to implement.
If you have 5mV ripple with CLC you're doing well. If you want to go crazy, move that the CLC to a 2nd box, use a ROBUST umbilical cable and then add either RC or just C in the amp box. That could be the motor run caps and/or 'lytics.
Also, your new front end boards have a set of caps onboard, too (C20/21). That's the final C for clean power.
If you have 5mV ripple with CLC you're doing well. If you want to go crazy, move that the CLC to a 2nd box, use a ROBUST umbilical cable and then add either RC or just C in the amp box. That could be the motor run caps and/or 'lytics.
Also, your new front end boards have a set of caps onboard, too (C20/21). That's the final C for clean power.