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

i try to compare the pcb pics by Harry (post 411) with the Lt spice file.
so maybe it is not a new information 😉

GND and out at the bootstrap adapter pcb is at the Resistor R3, R4 on the end of the LT1010 - so your M2OPS.
pic 1
1730839052106.png


Diodes D1, D2 are not existing BZX84C15L -15v Zener diodes. Rfb1 (1k) and Rfb(19k) must be on your Whammy pcb.
pic 2
1730839198966.png


chris
 
Was the LTSpice bootstrap circuit originally for the Whammy?

Assume for the M2OPS a 3K3 to earth on the -ve input would be used and 47K global feedback with 470K local on the OpAmp.

I'd model/ validate this myself with an 828, but I don't have a laptop that will run LTSpice or a Spectrum Analyser to test the complete amp
 
Was the LTSpice bootstrap circuit originally for the Whammy?

Assume for the M2OPS a 3K3 to earth on the -ve input would be used and 47K global feedback with 470K local on the OpAmp.

I'd model/ validate this myself with an 828, but I don't have a laptop that will run LTSpice or a Spectrum Analyser to test the complete amp
Hi Dan

sorry, but what are you talking about?
do you have a M2OPS?
have done the complete amp?

AFAIK the bootstrap adapter pcb should help to use a higher rail for an existing op amp - without damage the opamp - to us a bigger output swing and get a better output power over the output MOSFETs.
 
551 can take a constant ±30V, 604 is rated at ±24V. To these ears 604 is miles better, YMMV, etc.

828 was released in 2017 (I think). It is the absolute SOTA but will only tolerate up to ±18V, hence the desire for that little Bootstrap daughter board.
 
  • Thank You
Reactions: BRN
Does anyone know why it's necessary to change the feedback layout with a bootstrapped front end? (Removal of local/ mixed feedback)

I'm also in the dark as to why there's no series or shunt (input loading R) at the input pin. I'm guessing that a 1K series and a 47K shunt would be appropriate for an OPA828, maybe a 100K shunt?

I did find an older version of LTSpice that I could run. Both the mixed feedback and the new proposed layouts appear to work, the mixed feedback layout gives less gain and does require the 3k3 input series resistor. If I reduced to below 2k then I got clipping.
 
551 can take a constant ±30V, 604 is rated at ±24V. To these ears 604 is miles better, YMMV, etc.

828 was released in 2017 (I think). It is the absolute SOTA but will only tolerate up to ±18V, hence the desire for that little Bootstrap daughter board.


First Datahseet entry of AD828 is 1993 -Analog device
operation voltage +/-15V

OPA828 is from TI --> first entry December 2022
oparation Voltage +/- 18V