Screen grid voltage at full B+?

[Rhodesplyr]

I've noticed that in some of the earliest output stage designs using beam tetrodes, most often with 6V6s, the screen grids were run at full B+ voltage from the same power supply leg as the center tap of the output transformer.

Then, as you get into the early 50s, this approach was generally abandoned in favor of dropping the screen voltage below plate voltage. I see this pattern in the first 6V6/6L6 designs by both Hammond and Leslie. They were all revised around '51-'52 to drop the screen voltage.

I can see that one advantage of running the screens at full B+ was build simplicity, but did it have any other advantages? On the other hand, if you're running the screens from the first filter stage, as is often the case with the OPT supply in guitar amps, you end up with a good bit of AC ripple on the screens. However, even the Hammond and Leslie designs that had effective choke-filtered power supplies were changed to lower the screen voltage.

Were there other design disadvantages to running the screens at full B+? I've heard various comments over the years that it caused reliability and stability problems, but I have no solid information on the topic, hence, my post.

[Merlinb]

How early are we talking when you say "some of the earliest"? And are we mainly talking guitar amps?
In the early 1930s, pentodes/tetrodes were pretty new and exciting, and many designers would have had little insight into what was best for the screen grid. And in any case, big capacitors were as expensive as they were unreliable...

[jazbo8]

"the screen grids were run at full B+ voltage from the same power supply leg as the center tap of the output transformer"

Is this not a pseudo triode connection? Is there a resistor in between, if so, how large?

[Rhodesplyr]

I'm talking about 1945-1951, but I've seen it in a few amps through the 1950s.

In all these I've seen, all the screens are directly connected to the B+ filter capacitor with no intervening resistors.

Someone just gave me the power amp from a 1957 Dumont phonograph console, and it uses 4 x 6V6 with the screens connected directly to the OPT center tap. Other examples include the Leslie 31H, 25, and 125 amps.

Here's a Hammond example, the Type G amp from ~1945:



Also, if you've ever looked at 6V6 datasheets, several of them show the screen voltage and plate voltage at the same value. Most of the circuits I'm talking about appear to follow these datasheet numbers.

[trobbins]

I have a 6V6 PP PA amp from late 1940's with same supply for OT CT and screens, and no grid or screen stoppers. RCA RC-18 manual from 1956, and Brimar 7 from circa 1957 both show same circuit configuration. Maybe some major users changed their views in early 1950's due to valve reliability, and/or a better hum appreciation and how to optimise filter cap placement and $. I haven't looked at RDH4 comments.

[potatofarmer]

"the screen grids were run at full B+ voltage from the same power supply leg as the center tap of the output transformer"

Is this not a pseudo triode connection? Is there a resistor in between, if so, how large?

Why would this not be a (poorly implemented) pentode connection? The screens are connected to a B+ node which is coincidentally also used to supply the OT, and thus the plates. The screen voltage does not follow the plate voltage.


To the original post: Seems to me a choke that can handle the current of the whole amp is a whole lot more expensive than a choke rated for just the screen + preamp current. Or a resistor and capacitor, for that matter.

Actually, could this have been related to the cost of capacitors?

Running the screens as close to their maximum voltage as possible is still the rule of thumb for maximizing power output. It just so happens that in 6V6, 6L6, and the other common tubes in guitar amps this is also close to the maximum plate voltage. Not the case for EL34, sure, but running their plates at around the maximum screen voltage still gets you piles of output. When you try this with bigger beam tetrodes, you start running into trouble. You have to drop the screen voltage (or, conversely, you can now run the plate voltage significantly higher than the screen voltage) to keep the screens from melting.

[jazbo8]

I see now... the schematic shows a pentode connection. Perhaps the tubes were more robust back then and the people didn't know better. I haven't come across any modern design using such connection probably for all the reasons already mentioned.

[g1]

Having the screens run off the OT CT with no screen resistors will pretty much guarantee they are at a higher voltage than the plates, will it not?

[Enzo]

They knew perfectly well "better", but they also knew this wasn;t rocket surgery and they could easily get away with it without adding the expense of more power supply stages etc.

[J M Fahey]

I look at it from the other end.
I don't find that after a certain date they started to lower screen voltages, but that they started a power race and started raising +V voltages beyond (some cases WELL beyond) tube manual recommendations.
They could pull that stunt successfully because they could, plates were stronger than expected and are large (relatively speaking) masses of metal, with big dissipation surface, while screens were (are) made out of thin flimsy wire.
I have witnessed bright yellow screens blinking in rhythm with the music, (which means a VERY high temperature) while plates only glow dull red or at most orange and only a part of them, keeping structural integrity.
Then adding limiting resistors became a necessity.
I bet those 40's amps were very conservative with their +V supply values.

[pdf64]

Screen grids get a lot of stress when the amp is overdriven.
As power output increases towards max clean they draw much more current, then when pushed into squarewave it ramps up again.
Hence the need for current limiting resistors.

Bear in mind that a Hammond set up procedure involved tweaking gain etc such that there was no overdrive, even with pedal full down and all stops out.
I seem to remember seeing that on a schematic somewhere, apologies if I dreamed it.

And it was probably never envisaged that amps would be operated heavily overdriven.

With guitar amps, the focus of Fender seemed to be to squeeze the max audio power out of the power tubes, even to the point of exceeding the limiting maxes.
That seems to have set the subsequent methodology.
Given that, they may not have seen the benefit of a fancy choke input supply for the screens (thereby prolonging tube life) when more output power could be obtained by running the screens up close to the main VB+.
Pete

[Rhodesplyr]

pdf64, Your comments jibe with my general impression of the situation. You're quite right that Hammond and Leslie did not intend for their amps to be overdriven. Still, both Hammond and Leslie dropped the practice of running 6V6 screens at full B+ after doing it in their first 6V6 (6L6 for Leslie) amps.

The 1957 Dumont power amp that someone gave me uses 6V6 screens at B+ voltage. It had been in a home phonograph console, and, with the original circuit design, you can't really overdrive the output section. It has four 6V6s in parallel push-pull, and the driver (12AX7) begins clipping before the output stage at about 15W. Tweaking the circuit slightly (dropping in a 12AT7), I can drive the output section to ~20 Watts. (The reason I can drive the output section harder with the 12AT7 is because its greater current draw produces a higher cathode voltage and allows for a higher voltage input signal).

So, what I'm concluding is that running screens at full B+ voltage was OK in amps that were never supposed to be overdriven. Following from that, if I were to convert this Dumont amp for guitar use, it might be a good idea to change the screen supply arrangement. (This amp is very simple, using no choke. B+ for output tubes and screens comes off the first filter stage.)
+
In contrast, I've measured a Leslie 44W amp, which runs four parallel push-pull 6V6s--and which drops the screens 40V below the plates--at 28 Watts before clipping. It has a robust 6SN7 driver circuit.

I can't find a Fender amp that uses this full B+ screen arrangement, even looking at the earliest ones, but I don't have an encyclopedic knowledge of those circuits.

[pdf64]

'I can't find a Fender amp that uses this full B+ screen arrangement'
See the 5F4 Tweed Super http://www.webphix.com/schematic%20h..._5f4_schem.pdf
Probably manageable due to the lowish, saggy B+.
Pete

[Jazz P Bass]

Hah.
Had a clown running a BF Bassman with some off the wall attenuater.

Wondered why the output tubes where failing when he played it on '10'.

I actually watched a screen sag & melt.

1K 5 watt cement screen resistors sent him on his way.

[Rhodesplyr]

'I can't find a Fender amp that uses this full B+ screen arrangement'
See the 5F4 Tweed Super http://www.webphix.com/schematic%20h..._5f4_schem.pdf
Probably manageable due to the lowish, saggy B+.
Pete

I stand corrected. The 6L6G datasheet I have lists 300V as the absolute screen limit, so Leo was playing roulette here :-)