• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

KT88 PP - Dotting the Is, crossing the Ts

Jump to Latest
#204
This is the status quo of the prototype, I'm wondering what else I could possibly do to further reduce distortion? I've been going over old schematics from the '50s and '60s trying to find more tweaks to that effect, but other than the 'Schade' approach, which I already tried and which didn't yield a positive effect, I've not found anything yet? Open to suggestions if anyone has any?

Proto 051
proto_051.jpg
 
#205
SSassen said:
My take is that the two triodes inside the 6922 are not perfectly balanced, which makes sense, and by changing the anode resistors AC balance is restored so now the loop gain that was available has less distortion to work with which nets lower overall distortion.
Click to expand...
P.s. this seems to line up with my findings?

https://www.aikenamps.com/index.php/the-long-tail-pair

When only one signal input is used (ignoring feedback inputs) R1 is usually made 10% - 20% lower than R2 to compensate the unbalanced gains of the two tube sections and make the two output amplitudes equal.
Click to expand...
 
#207
When only one signal input is used (ignoring feedback inputs) R1 is usually made 10% - 20% lower than R2 to compensate the unbalanced gains of the two tube sections and make the two output amplitudes equal.
Click to expand...

That's when you don't have a CCS in the tail. When you do the AC balance is near perfect as the the current in one tube is CCS-the other. The plate resistors will affect AC balance hence you added the pot. The remaining 2nd harmonic is maybe mismatches in the KT88. LT spice of course will consider them identical.
 
#209
That's when you don't have a CCS in the tail. When you do the AC balance is near perfect as the the current in one tube is CCS-the other. The plate resistors will affect AC balance hence you added the pot. The remaining 2nd harmonic is maybe mismatches in the KT88. LT spice of course will consider them identical.
Click to expand...

Could very well be, my KT88s need a different bias voltage to draw the same amount of current.
 
#213
OK when you do watch for a differential hum loop. That's when a current runs between the L and R screens of the inputs and through the 0V induced by the magnetic field of the mains transformer. If you place the input stage tubes as close as possible rather than having two amp spaced apart. Just a trap to fall into.
 
#215
Start with a pair of reasonably well matched good 6922 triodes (and in the same glass envelope).
Use a perfect CCS (Constant Current Sink) that is connected directly to both triode's cathodes.
Use perfectly matched perfectly accurate plate load resistors of 33,000 Ohms.
Use a signal generator to drive a 1k grid stopper resistor to the first grid.
The signal amplitude needs to be in the linear and non-clipping range of the phase inverter.
Ground the second grid through a second 1k grid stopper resistor.

Suppose the signal drive to the first grid causes its plate to increase by 0.500mA.
The CCS will take away 0.500mA from the other tubes cathode, so its plate current will reduce by 0.500 mA.

First plate voltage will reduce by -0.000500A x 33,000 Ohms = plate voltage reduction of -16.500 Volts
Second plate voltage will increase by +0.000500A x 33,000 Ohms = plate voltage increase by -16.500 Volts

Do you see how that works?
The balance is perfect.
No current goes into either grid. No current comes out of either grid.
The plate current changes are equal and in opposite directions.
The total CCS current is Constant.

Do you have to adjust the signal amplitude balance of the two plates?
No.
The balance is Intrinsic due to the circuit topology.

Simulation Experts:
Do you want to simulate what can happen with an unmatched pair of triodes?
u = Gm x rp
In your tube models, Increase Gm by 2% and decrease rp by 2% of one triode versus the other triode (u will be unchanged).
Run the simulation with the 2% unmatched Gm and 2% unmatched rp of one triode versus the other triode.
Find the resultant difference of the two plate signal amplitude voltages.
That is the only cause of un-balanced phase inverter amplitudes.
Excellent tube matching will eliminate that error.
 
Last edited:
  • Like
Reactions: jcalvarez and oemcar
#216
Yes, in a perfect world this applies, and when you whip up a LTP on the bench as you suggested. The issue here is that it is in circuit, with either triode not driving the exact same load, no doubt not even purely resistive, so when asked to drive a different impedance the AC amplitude will differ, simply because of Ohms law. The fact that some tube amplifiers that have an LTP also have a AC balance pot probably wasn’t a mistake?
 
You must log in or register to reply here.
Top Bottom