... The 6D6 ... plate are only seeing around 40v, with the anode resistor at 250k.
With these values the valve seems to work fine and sound good. 
What is the relationship between B+ and plate voltage and the effect on the valve. ...
This is a bit of a challenge because it's a pentode. I presume the screen is sitting at ~40v, too? (I can't open sch files on this computer, but got the general idea from the image of your schematic in the other thread).
First off, it's an old amp with 2A3's, so the available supply voltage won't be huge in any case. That said, this is the input stage for a mic channel, which might have a signal level similar to, or lower than, your guitar pickup. So output voltage swing may still be quite small, even if gain is relatively high (relative, because these are somewhat early pentodes).
The key figure for a pentode is the Gm (transconductance) at the operating point. You can take voltage gain as being Gm * Rl. Looking at the last page of a
6D6 data sheet and knowing the predicted results may be a bit optimistic, 40v on the screen gives a Gm of 500µmhos, or 0.0005 mho. 250,000Ω * 0.0005 mho = 125. Apply a 100mV signal and you could get 12.5v output. That seems like it might be within reason for ~40v on the plate.
Pentode Gm rises with higher plate current, which rises if you raise the screen voltage. You'd think based on the formula above that this is the path to higher gain from your pentode. But the tradeoff is unless the overall supply voltage is quite large, your big plate load resistor will have the output voltage swing bumping into cutoff & saturation fast. The fix would seem like lowering the plate load resistor value, but that also reduces gain, which is why you were shooting for higher Gm in the first place.
Surprisingly, while the show-off Gm figures in the data sheet operating conditions are well above what you'll get in practice, and consequently the tube won't deliver as much gain as you'd expect from those show-off conditions, most real amps run pentodes with a fairly low screen voltage, and low plate current so that the plate load resistor can be raised. You're trading one part of the gain equation for another, and easing the requirements for the power supply. And as shown above, this may be perfectly fine when you're using that pentode as an input stage with relatively small input & output signals.
...6D6/6A6 :- If the max B+ supplied is 280v and can't be raised without reducing the up stream dropping resistors and risk decoupling.
And with the valves working well at these voltages has got me stuffed that they actually work.
6SJ7GT :- I'm sure if I could read "load lines" it would tell me the sweet spot the valve sits with the plate resistor value that is in the original circuit.
Otherwise I'll shoot for the "typical" voltage of 250v as per the datasheet. ...
I'm too brain-fried with it being late to look up the 6SJ7 and cross-figure. But I think you might ignore the "typical 250v condition from the data sheet".
PRR has long said when I asked about stuff like this that they're "show off" conditions. I am beginning to understand. In fact, for the 6D6 I think the 250v plate condition under "Class A Amplifier" typical operating conditions is either bunk derived from applying d.c. voltages to the tube without any load (to calculate/measure element currents & Gm) or is a suggestion for a choke/transformer-loaded stage.
Let's look at that condition with 250v plate & 100v screen. Gm is given as 1600µmhos, which by our formula should give a lot of gain, right? Let figure out the plate load & supply voltage necessary to get just the gain of 125 as in our 40v plate sample from above, while still having 250v on the plate. Working Gain = Gm * Rl backwards to solve for Rl gives Rl = Gain / Gm = 125/0.0016 mhos = ~78kΩ. Pretty good, right?
But look at the plate current in that condition: 8.2mA. (!) That's gonna be pretty draining on the power supply, but what about the drop across the plate load resistor? 78kΩ * 0.0082A = ~640v (!!!) plus the 250v left on the plate, for a needed supply voltage of 890vdc!
So there's no way that could be an R-C coupled gain stage with a plate load resistor. It just doesn't make any sense. And you shouldn't chase the data sheet condition in this case. Without peeping the 6SJ7 sheet, I suspect we'll find something similar over there.
It might be better to figure out how much supply voltage you reasonably have available, and what numerical gain you need to develop. Then you can pick out something reasonable for these tubes.
Topic: B+ and plate voltage relationship (Read 7136 times)