Low Power from Cathode Biased 2x6V6 Magnatone

[Helmholtz]

I think you should be able to get around 10VAC

Yes, but we don't know how much grid drive the PI is able to provide. The fact that the output clips before the grid signal means that the clipping is not caused by the PI or by grid conduction.

[Helmholtz]

B+ voltage at the OT center tap at the start of clipping are essentially the same as at idle, maybe 1 volt higher. From that point on the voltage goes up as clipping increases.

That's very unusual. In a class AB amp current demand increases with output causing B+ to drop.
There is a possibility that the amp actually runs in class A, which is very inefficient and not really desirable (from a technical POV) for PP stages. If the amp runs in class A will be determined by the idle currents and OT primary impedance (being very high). In class A current demand essentially stays constant with output.

What is known about the OT, is it original? Does that 150R cathode resistor look original?

Still I doubt that B+ could increase as clipping increases.

[Helmholtz]

@ hasserl: Sorry, why a thumbs down?

[hasserl]

@ hasserl: Sorry, why a thumbs down?

I'm sorry, I hit the wrong button! Bad eyesight. My apologies. I better go back and see if I gave anyone else a thumbs down, definitely not intended! I appreciate each and every response. Sorry guys.

[Helmholtz]

Do you have a chance to try a different OT with a primary impedance of 8k?

[hasserl]

That's very unusual. In a class AB amp current demand increases with output causing B+ to drop.
There is a possibility that the amp actually runs in class A, which is very inefficient and not really desirable (from a technical POV) for PP stages. If the amp runs in class A will be determined by the idle currents and OT primary impedance (being very high). In class A current demand essentially stays constant with output.

What is known about the OT, is it original? Does that 150R cathode resistor look original?

Still I doubt that B+ could increase as clipping increases.

I believe the OT to be original. The cathode resistor is not original, and apparently is not the same value as stock (according to the schematic the original value is 250 ohms, modified resistor is 150 ohms).

B+ definitely increases as clipping occurs. I assume current flow through tubes decreases as clipping occurs, causing the increase in B+.

[hasserl]

Do you have a chance to try a different OT with a primary impedance of 8k?

If I had one on hand I could try that. I don't.

[g1]

Have you tried it with a 250 ohm instead of the 150 ?

[hasserl]

Have you tried it with a 250 ohm instead of the 150 ?

NO, I haven't yet. I don't think I have one in stock, and I'm not looking forward to going to a local store to try to purchase one, just trying to avoid going anywhere with human interaction right now. I need to place an order for some other components since I'm running low of commonly used items, I'll try to get a resistor with that order.

[g1]

Anything you can stick in series with that one to test?

[Helmholtz]

B+ definitely increases as clipping occurs. I assume current flow through tubes decreases as clipping occurs, causing the increase in B+

As the total supply current drawn by power stage must flow through the cathode resistor, cathode voltage is an indicator of actual supply current.

If B+ increases with clipping, cathode voltage should decrease.

[hasserl]

As the total supply current drawn by power stage must flow through the cathode resistor, cathode voltage is an indicator of actual supply current.

If B+ increases with clipping, cathode voltage should decrease.

It does. See next post

[hasserl]

Looky looky what I found hidden in the bottom of a drawer:
(ETA: well that's not a very clear picture. It is a 250 ohm 5 watt resistor)

So, with the new cathode resistor I rechecked all voltages (note mains are slightly higher today)

A/C HT at the rectifier = 315/317
B+ = 332
C+ = 326
Power tube plates = 325
Screens = 323
Cathode = 20.3
A/C at 4 ohm load resistor at start of clipping = 4.43
B+ at start of clipping = 332

As I increase the clipping beyond that point the B+ immediately starts to climb, up to a high point of 342. At the same time K voltage decreases, to a low point of 17.8 (at the same time as max B+).

Could it just be that the OT is too small to produce any more output, that it's just saturating, causing the clipping in the output?

[g1]

Have you scoped the ripple on B+ at full power and at clipping?

[Helmholtz]

As I increase the clipping beyond that point the B+ immediately starts to climb, up to a high point of 342. At the same time K voltage decreases, to a low point of 17.8 (at the same time as max B+).

Ok, so measurements seem to be consistent.

I think it's like this: The power stage operates in full class A (very unusual with PP amps). In cathode biased class A amps cathode voltage essentially stays constant in the linear range. As clipping occurs and the output signal gets squared, tube dissipation reduces because the tubes are now essentially switching between 2 states of low dissipation. Consequently B+ rises.

The class A operation is due to the high idle current (is the 20.3V with a 250R cathode resistor?) and most probably an unusually high OT primary impedance (>>10k?, actual primary impedance can be calculated from voltage measurements at the disconnected OT). The reason for the high (reflected) primary impedance could be that the OT was rated for a load impedance below 4 Ohm. A high load impedance would also explain the high voltage gain of the power tubes showing in the low grid drive required for output clipping.

Output clipping is no sign of core saturation. Rather saturation typically shows as something like crossover distortion at zero crossings without limiting voltage peaks.

Could you post a scope picture of the clipped output?

What is screen voltage at full output?
I would also check grid coupling caps for leakage.

It should be possible to get around 12W of clean power from this amp with an appropriate OT.