The M2 Output Stage in Class A/B, and maybe a Power WHAMMY?

211226 M2OPS.JPG


Many here have jumped onto the M2 band wagen, but few have paid attention to the opto-biased output stage in detail.
We have, for example, discussed the use of different MOSFET pairs.
https://www.diyaudio.com/community/threads/complementary-power-mosfets.378024/post-6889027

Nelson has previously dropped some hints that it works well also in Class A/B.
But how ?

This is an attempt to analyse precisely that, using LT Spice simulations.
It is a waste, IMHO, not to use this clever OPS design in more applications.
For example, how about in a WHAMMY Power amp ?


Patrick

.
 

Attachments

  • The M2 Output Stage Analysed 220925 Public.pdf
    733.7 KB · Views: 1,859
Last edited:
  • Like
Reactions: 12 users
Member
Joined 2011
Paid Member
I think I recall reading in Nelson's thread "Official M2 Schematic," that he said his intended purpose of the three 1N4148 diodes was to give a smoother recovery from output voltage clipping when driving normal 8 ohm loads. The Pchannel side has a different number of diodes than the Nchannel side, intentionally.
 
Last edited:
Assuming Nelson would agree, and someone has a serious interest :

I shall make available the Gerber of the above PCB with an integrated DIP8 opamp powered by the same supply.
So then you can build a WHAMMY power amp and start playing with opamp and MOSFET types.

But Nelson has to approve first.


Patrick
 
  • Like
Reactions: 8 users
What opamp tp choose from other than OPA551 for a power WHAMMY ?

For +/-24V rails, OPA604 is well known.
ADA4098 will just make +/-25V.
Above +/-25V, choice is rather limited.
I found ADA4700, which is quite similar to OPA55x but can go as high as +/-50V.
Other candidates -- LTC2057HV, and perhaps ADA4522.


Patrick
 
Last edited:
Member
Joined 2011
Paid Member
Maybe Dreadnought in the Store might be a useful starting point for a low-complexity, higher voltage, discrete front end?

btw, if anyone's interested: to hand-calculate bias currents, assume (a) LM285 voltage is 1.235V per datasheet ; (b) Vbe of KSA1220A is 0.60V

_
 

Attachments

  • discrete_dread.PNG
    discrete_dread.PNG
    23.4 KB · Views: 969
The M2 OPS is a complmentary push-pull MOSFET output buffer.
The original M2 has a JFET frontend with no global feedback.

So, for me, a fitting discrete frontend is either an all-FET VAS with no global feedback, e.g our own HAGS :
https://www.diyaudio.com/community/...a-zgf-headphone-amplifier.225577/post-3395149
or if it has to come from Pass, an F5-Pre (single ended version).

But a power WHAMMY is an easy starting point.


Cheers,
Patrick