Musical Fidelity A1 modification

[ilimzn]

here is a little modification for the classic MF A1 amp!
A few notes:
Other mods on the circuit could be done, keeping the basic topology (and basic sonic signature) the same, but what was done was within the limits of the existing PCB (to avoid cutting and flying wires) and the existing transistor complement.
The diodes between the inputs of the LTPs were left in place (just not whown on the modded schematic).
C* are compensation caps for ringing that can appear uppon exit from clipping.
The rest of the schematic, i think, is self explanatory.
The schematic of the original, along with other interesting info, can be found at http://www.mhennessy1.f9.co.uk/mf_a1/technical.htm
The point of the mod was to completely eliminate the most obvious weak point in the A1, it's preamp. The input pot was, in light of this, re-connected in the 'conventional' way with DC block in front, and the gain increased for full output at about 250mV RMS.
The circuit has two potential 'snags' that would, with an instant appearance of both power supplies, potentially compromise the SOA of the outputs, see if you can spot them
As always, comments are highly welcome!


Moderator's edit: Incorrect schematic removed as per author's request.

[moamps]

Hi!

I'm a bit tired so I hope you don't mind if I can't really tell whether this is supposed to be an amp or an oscillator. CLG looks a bit high.

Regards,
Milan

[ilimzn]

The CLG (55x = 34.8dB) was chosen as a compromise, to let older sources with 315mVeff 'full scale' and potentially softer recordings to reach full power output, plus a bit to spare, using the existing parts to make up a passive front end.
If you think that is high, the original implementation (with preamp) had a maximum total gain of 137.5x (42.7dB)

It's not an oscilator (simulated, tested and re-tested ) although the 'compensation caps' actually do some harm in this regard, were it not for the ringing when the signal 'un-sticks' from the rail at clipping, it would be better not to have them. Surprisingly, the sim quite accurately predicted this - except in reality it happens at a slightly lower frequency. In fact, simulating the original design gives results remarkably similar to the published specs!

Just to make things clear, this is not just an idea - it was done, tested, and quite extensively listened to on various speakers.

[moamps]

Originally posted by ilimzn
Just to make things clear, this is not just an idea - it was done, tested, and quite extensively listened to on various speakers.

I've just had another look and I am pretty sure that you have drawn your schematic incorrectly. The error is quite obvious. The signal for the first driver transistor (BC550/560) should be coming from the left transistor in the differential pair. IMO.

Regards,
Milan

[ilimzn]

Ah, yes - you are right, of course, stupid mistake. Here's the corrected version, direct from the simulator (hopefully it has remained sufficiently legible). Too bad I can't edit my first post, how embarrasing I would appreciate it if you could remove the picture lest someone actually try to connect the thing that way!
R27 in the schematic, is of course the load. The actual situation is somewhat more complex at the input as the A1 board already contains an input RF filter, and potentiometer which are not shown.

[moamps]

Post edited. Everything should be fine now.

Regards,
Milan