friedman post phase master design

mixn4him said:
220k inline from treble pot, 1M to ground…
Click to expand...
I think that’s the pre-phase inverter MV. The post-phase inverter MV is just a dual 250k pot but leave the 220k bias splitters. So each of the dual pots is in parallel with each bias splitter.

It’s just like other post-phase inverter MVs except it keeps the 220k bias splitters and uses 250k dual pot. Most designs used a combination of resistors/pots such that the bias splitter value remained close to the stock 220k, for example the LarMar used a dual 250k pot but used 2M2 resistors from wiper to ‘ground’ (‘ground’ is the bias supply here). Or you could use a dual 1M pot and leave the 220k.

For some reason the Friedman MV sounds a little better when turned down. Not sure why.
 
SpiderWars said:
I think that’s the pre-phase inverter MV. The post-phase inverter MV is just a dual 250k pot but leave the 220k bias splitters. So each of the dual pots is in parallel with each bias splitter.

It’s just like other post-phase inverter MVs except it keeps the 220k bias splitters and uses 250k dual pot. Most designs used a combination of resistors/pots such that the bias splitter value remained close to the stock 220k, for example the LarMar used a dual 250k pot but used 2M2 resistors from wiper to ‘ground’ (‘ground’ is the bias supply here). Or you could use a dual 1M pot and leave the 220k.

For some reason the Friedman MV sounds a little better when turned down. Not sure why.
Click to expand...

It’s because the shunt input impedance has a maximum tolerance for the power tubes and is typically 1M. The tubes are happier and there’s also less white and shot noise being amplified.
 
If these designs are messing with the bias supply voltage, how are they assuring no crossover distortion as you turn it up or down?
 
glpg80 said:
If these designs are messing with the bias supply voltage, how are they assuring no crossover distortion as you turn it up or down?
Click to expand...
I don't think they are messing with the -DC bias voltage, that stays pretty much the same. The bias splitter resistance is at both ends of the pot so it doesn't change. It's just the AC signal that is getting shunted to 'ground' (which is the filtered bias supply).
 
lespaul6 said:
I wish I knew… I thought it was messing with the nfb loop like that but not bias?
Click to expand...

Nope. The 220k are apart of the grid bias that sets idle plate current. Friedman’s design just turns your amp into a hard improperly biased bi-phase class B amplifier and gradually becomes class A/B as you turn it up and then depending on where bias is initially set, could become hard class A. Honestly I think that design is total shit if he is in fact adjusting the bias like that and calling it a volume pot.
 
SpiderWars said:
I don't think they are messing with the -DC bias voltage, that stays pretty much the same. The bias splitter resistance is at both ends of the pot so it doesn't change. It's just the AC signal that is getting shunted to 'ground' (which is the filtered bias supply).
Click to expand...

No DC current flows in grid bias but the pot shorts the bias splitter when the value is less than the the total resistance in parallel with the pot, which sends more negative voltage to the grid, biasing the amp hard class B, reducing volume. Unless I’m misunderstanding something here that’s exactly what’s going on and why he has to use a dual gang pot.
 
I was wondering if he has the PI coupling caps connected to the wipers of the dual gang pot.
That would keep the effective bias splitter value constant at 117k and the bias stable.
Dave has mentioned that when you turn the master down, it loads down the PI.
 
glpg80 said:
No DC current flows in grid bias but the pot shorts the bias splitter when the value is less than the the total resistance in parallel with the pot, which sends more negative voltage to the grid, biasing the amp hard class B, reducing volume. Unless I’m misunderstanding something here that’s exactly what’s going on and why he has to use a dual gang pot.
Click to expand...
I just assumed they were connecting the pot end lugs where the old bias splitters were and then running just the signal to the wipers. So the pot is kind of backwards than the typical schematic.

EDIT: yeah @fusedbrain said essentially same.
 
fusedbrain said:
I was wondering if he has the PI coupling caps connected to the wipers of the dual gang pot.
That would keep the effective bias splitter value constant at 117k and the bias stable.
Dave has mentioned that when you turn the master down, it loads down the PI.
Click to expand...
I thought it was something like this but, of course, I could be persuaded by either argument not knowing enough about it honestly... very good discussion
 
glpg80 said:
It’s because the shunt input impedance has a maximum tolerance for the power tubes and is typically 1M. The tubes are happier and there’s also less white and shot noise being amplified.
Click to expand...
9569.jpg
 
Honestly you’ll get better tube life out of his amps so my original statement was a bit extreme but I prefer not to be messing with the bias as a volume taper. The reason is because as tubes age, their bias drifts, and your volume could actually change as the amp warms up which doesn’t happen with normal designs. I don’t think it’s a completely awful design but I do think it leaves a lot of room for improvement.
 
You must log in or register to reply here.
Back
Top