PA-940 - Mc Crypt, Better, Sun Key, Stage Audio, JB Systems...

R10 is technically incorrect. Implementation without a bootstrap is wrong from a purely technical standpoint, although it MAY result in something that subjectively “sounds good”
I'm probably not going to implement a bootstrap. Does that mean there's anything worth doing with R10?

I'm sorry for my ignorance, but I don't exactly understand how a long-tailed pair input stage works exactly. If I have balanced DAC/pre output, can I feed that into the LTP after breaking the negative feedback circuit (removing C4, C5 and R5)?

Other than that the only thing I’d jack with is the ratio of R3 to R11 to ensure differential balance, after ensuring the match between the transistors is 5% or better. Because the folded cascode is resistive loaded...
I will measure the output transistors if I decide to change them, for the time being I do not know how well they are matched. Does that mean I have anything to do with the ratio of R3 to R11? I thought that the biasing network already takes care of that.
I don't understand what is folded cascode, nor what it means that it's resistive loaded, but I'm pretty sure I will figure out at least the folded cascode part. Is it just when npn drives pnp perhaps (and vice reversa)?

I ordered some 5W metal film resistors to replace the keramic load resistors.
 
Last edited:
Q3 replaced.
Again less dull, better high frequency performance. Relatively big improvement. ABing around 10s between transistors. Do what I did in pic and listen for yourself.

Also it may not sound like important step, but do move the speaker ground and the PSU ground to the amp section of the board instead of the protection circuit area. There is significant enough difference fo warrant it in my opinion.
 

Attachments

  • IMG_20250319_212334.jpg
    IMG_20250319_212334.jpg
    288.2 KB · Views: 19