Balancing the Difference

[bruce egnater]

Are you making "audio" measurements to validate what you are hearing or just measuring DC voltages? Certainly one can assume no production PP output transformers will be balanced. This is interesting stuff.

[dorrisant]

If I am understanding this correctly, you would feed the screen node into the wiper of a "balance" pot to adjust the screen voltage from each side of the push pull. Kinda like a "screen-dinger". Then each side of this pot would be connected to another pot(s) used as a variable resistor (replacing the actual screen grid resistor), this would then be tied to the socket pin. Also tied to that pin would be the bleeder resistor to ground...

As I went back and forth, I noticed that when setting the voltage difference between plate and screen, that the difference would peak, and at this peak both sides would be in balance.

I'm having trouble with this part... difference between what? ?

I did Andy's screen voltage divider mod to a red knob twin a while back.. there was a nice difference. The owner still talks to me about how much he liked it.

I love this kinda stuff.. Great work Andy.

Tony

[Andy Le Blanc]

the difference between the plate a screen... on each leg of the P-P pair

and then the difference between these two measurements

this was a surprise to me, I had been varying the screens voltage with a series resistor, before a bleeder and screen grid resistors, the larger the ohmeric value the lower screens voltage, etc...

but the behavior change when I rigged a balance pot..
the pot is simply rigged before the screen grid resistor and bleeders

I'd swing the balance and the voltage difference on one side would increase,
the screen would become more negative to the plate, but then would go the other way and become less negative despite the direction of rotation on the balance pot.

and when the screen was the most negative, the peak reading, both sides were closely balanced with the same difference measured between the plate a screen.

I got to say again that it was a surprise. I didn't expect it.

[dorrisant]

Thanks Andy... I think I've got it now.

I will definitely try this out and I know just the guy to call... the dude with the red knob twin. He will be stoked.

Tony

[Andy Le Blanc]

So far its been simply comparing the DC measurement to what I'm hearing.

Actually getting some real THD and RTA measurements is next.

The hard part is keeping it stupid simple, and practical for DIY building.
Simply being able to get a measured reading out of the P-P dynamic circuit,
and then manipulating the circuit values towards your own aesthetic is really the goal.

There is a practical limit, jazbo8 isn't wrong, I have tried plenty of HI-FI approaches that didn't pan out for guitar amps.

I really like being able to set up an amp under static conditions with a good measure of control, IE: measuring bias over a 1r... and controlling the screens voltage with a minimum of effort and parts is great, makes one hell of a difference in the consistency of tone for tube set to tube set, and even tube type.

It gives you one more tool to manipulate your tone. just keep it simple.

I think I mentioned a couple post back, a simple link switch to tie the screens together post of the screen grid resistors, and modern/vintage mod thing.

[M Fowler]

I've noticed in my builds that PP EL84 PA cathode bias always have screen voltage at or higher then my plate voltage for some reason.

[Andy Le Blanc]

I first got on to this kick with fender circuit, a funny extra distortion that went away when I got the screens under control.

you can manipulate it with your bias point too.

just don't drag down your plate voltage by over biasing the amp.

Hard part is that the rig will probably sound "hard" if at the bias point that is friendly to the screens.

looking at older designs, if the screen are sourced thru a series resistor from the center tap of the OPT, the screen should track with the plate voltage as the increased current from a hotter bias drags down the plate voltage.

all kinds of aspects to explore.

[Andy Le Blanc]

chance came up fast, Just got a Gretch Electromatic thru the shop.

the rig had been converted from a field coil speaker, but hadn't had any attention paid to the screens.

the last tech had them dressed to the ct of the opt with no other screen resistors or series dropping resistor, which was the appropriate choice to keep the rig as close to vintage as possible, but there was no choke or series resistor to replace the old field coil speaker resistance.

so the screens were positive to the plate... and the client didn't like the response and tone when he pushed the rig.

The simple solution was to put in a series dropping resistor on the power rail and run the screens, still with no screen resistors, from there own tap.

This put the screen about 20 volts negative to the plate.
And is a very tweed solution... it a corker of a rig now, and with out the screen grid resistors the screens are tied at the same voltage.

you get that "compressed" drive when you push the rig

[Andy Le Blanc]

good little read, might seem like apples and oranges being about transmitting types...

but a tetrode is a tetrode... The conditions for screen failure and the role that series screen grid resistors play is pretty well defined.

have to ask just what the role of screen grid resistors are.

look at older designs, how a difference between the plate and screen is maintained with a series resistor on the power rail.

there are no screen grid resistors, the screen are negative to the plate, and the screens are tied at the same potential.

when you add screen grid resistors, you free the push-pull dynamic circuit to self balance around the tube characteristics, but you also introduce the
possibility of a runaway condition that can cook the tube.

you need to pay attention to loading, plate dissipation (bias) and screen supply, at the same time.

[dorrisant]

I think I will have to read that a few times till it sticks...
What values and ratings for the pots? I'm assuming the screen grid pots could be 1-1.5k to give a little wiggle room. I have two amps that I would like to test this on... Maybe three.

Tony

[Structo]

Interesting article.

So now I have to put a spark plug in my amp?

I didn't really understand the bleeder resistor part.

Maybe I'm making it too simple, but on the average 6L6 push/pull amp, if the amp has
screen grid resistors (470R or so) just putting another resistor in series to those two resistors will
help control screen grid current and in the process, sound better?

Next time I have the chassis out I will try some of this.

[matt h]

(deleted)

[dorrisant]

Great explanation Matt.

What values? Ratings?

Tony

[matt h]

(deleted)

[M Fowler]

Here a guy is using zener diodes to drop screen voltage or balance the screens.

http://el34world.com/Forum/index.php?topic=17991.0