Thicker PL400 faceplate with new analog meters...

[Gepetto]

Sorry, I'd just like to pull on this thread a bit. Which transistors are being subject to the full + and - rail voltages? Worst case (I see) is either the + rail voltage or the - rail voltage. Not both added together.

Secondly, dirty power - isn't that dirty power just coming through the transformer attenuated a bit by the inductance and capacitance of the transformer, but surely the rectification and the bulk capacitors will remove that to whatever extent you want. You could even add line filters (better be big ones) to reduce the incoming noise.

So, here's another try (with toroidal transformers). Instead of one transformer for the right channel and one for the left, use one to generate the positive rail voltage and another transformer to make the negative rail voltage. These should work: Toroidal Transformer 75v
So it's rated at 2x35 volts at 115vac, which would be 2x38 volts at 125vac, producing between 70 and 76 volts for each leg. There are other transformers that would allow even greater rail voltages (but with decreasing current capacity).

Outputs and drivers are subjected to the sum of the positive and negative rail voltage...both added together

 

[BlueCrab]

Ok, I think I see it now. For example, suppose the output transistors on the positive rail are in (or near) saturation. Little voltage across them, but that puts almost the total rail voltages across the output transistors on the negative rail, which are cutoff. Output signal is near positive maximum, so total voltage across the transistors in cutoff is nearly positive rail to negative rail. Is that correct?

 

[Gepetto]

Ok, I think I see it now. For example, suppose the output transistors on the positive rail are in (or near) saturation. Little voltage across them, but that puts almost the total rail voltages across the output transistors on the negative rail, which are cutoff. Output signal is near positive maximum, so total voltage across the transistors in cutoff is nearly positive rail to negative rail. Is that correct?

Yes but they are not in cutoff

 

[grapplesaw]

Sorry, I'd just like to pull on this thread a bit. Which transistors are being subject to the full + and - rail voltages? Worst case (I see) is either the + rail voltage or the - rail voltage. Not both added together.

Secondly, dirty power - isn't that dirty power just coming through the transformer attenuated a bit by the inductance and capacitance of the transformer, but surely the rectification and the bulk capacitors will remove that to whatever extent you want. You could even add line filters (better be big ones) to reduce the incoming noise.

So, here's another try (with toroidal transformers). Instead of one transformer for the right channel and one for the left, use one to generate the positive rail voltage and another transformer to make the negative rail voltage. These should work: Toroidal Transformer 75v
So it's rated at 2x35 volts at 115vac, which would be 2x38 volts at 125vac, producing between 70 and 76 volts for each leg. There are other transformers that would allow even greater rail voltages (but with decreasing current capacity).

Toroidal will would work if you can find 75-0-75 then need about 1600va to 1800va with how ever many of then need to get there