GLASS-SAND CLASS A TUBE MOSFET HYBRID HEADPHONE AMP

[kenpeter]

The current sink was used to check the sensitivity to plate current to sound quality. I backed it down to 100uA and it still sounded good and then raised it up to 6mA and still sounded good. I have it set a 3mA. Original thought was 2mA/side 4ma total tail current.
If I could set the plate current by cathode to grid voltage your idea cold work. What I'm concerned about is the variation found in the tubes when subjected to this low voltage operating point. They are less predictable that when used at high voltage.

You assume to better set plate current by absolute cathode current.
Yet have unaccountable currents that vary from tube to tube, leaking
to the grids. It obviously works. More predictable than a resistor, are
you sure under these leaky conditions, that statement is valid math?

 

[roline]

I replaced the CCS with 1k tail resistors, 2.4-2.6 mA plate current. It simplifies the design, new schematic and PCB layout attached..
That puts the cathode at -7.4 volts @ 2.4mA vs the CCS at -8V @ 3mA.

 

[kenpeter]

Ideas to chew on... Just noodling, not making any claims...


Might this fix plate current regardless what leaks out the grids?

Does R13 stiffen the perceived impedance of the tail resistor?

Should both grid leaks want to see same resistance to GND?

 

[kenpeter]

Maybe we can get both plates to idle at +8 or +9 volts ?
Just sayin' (+12V) - pVGS is probably > (GND) + nVGS

M1 is low pass shunt that adjusts for the desired plate
current, hopefully correcting for both grid leak currents.
Tail resistor wants to be a little smaller than otherwise
to cover the expected variation of grid leaks together
with the shunt action of M1...

Then you got 4.7x Schaded voltage gain at M2 assures
the plate voltage won't need to swing more than +/-2.5V
Reduced plate swing might help with the grid leak too.

VGS multiplier action is fixing filament voltage till the
main feedback loop gets warm enough to take over.
Might want to tweak R10 depending the actual VGS...

R13 is stiffening the tail with a positive feedback <1.
It would be sure disaster if R13 / R14 was > R6 / R5
Definitely don't want a resistor bigger than 22 here.
18 Ohms was merely the next lower standard value.

A CCS here could be even stiffer without any threat
of positive feedback. But its a few more parts, and
wouldn't be ever really be "constant" or "stiff" on
account of the leaks. Worth it? Maybe, maybe not...

R8 par R7 probably wants to be same like R6 par R5,
else the grid leaks may create different gird voltages.

Not built, not tested, not even simm'd. Just noodling...

 

[roline]

LITTLE AMP STILL SOUNDS GREAT. built a millet starving student 24volt to compare to. prefer the sand in my glass version...