Screen supply voltage re-visited

[skyboltone]

Greetings Lads:
I have a PT with 320-240-0-240-320 that was used with a pair of 7591As in fixed bias operation and a pair of 6GW8s for reverb. I'm thinking of running a pair of EL-34, KT-66, or even a pair of KT-88s with it. I am contemplating using the 240-0-240 winding for a screen supply thereby putting the screens about 100V lower than the plates. This is standard "by the book" screen supply, but I'm wondering if it will work at all with a guitar amp?

Thoughts?

Dan

[Andy Le Blanc]

Recently I tried 6550 with 600v on the plates and a 300v G2 supply voltage.
It made the watts, predictably, and was great sounding before clipping, but the
natural distortion wasn't pleasant. The same amp and same tubes with a
typical supply (fender twin supply) had a much better tone when pushed to
clipping. If you have an intended program or fixed drive level that doesn't
push the amp into a hard clip, you should be fine, but you'll have to use your
best judgment listening if your after a natural distortion. I got a pair of kt88
and a pair of kt90 today for my amp. To me, the only reason to invest in
them was, to try and take advantage of PT rateing in milli-amps.
6gw8 have a 9w plate dissipation and 7591a are on par with a 6l6, it might
be worth while to stay within that family of tubes, or a plate dissipation of
around 19w max, if that was what your PT's intention was for.

[skyboltone]

Recently I tried 6550 with 600v on the plates and a 300v G2 supply voltage.
It made the watts, predictably, and was great sounding before clipping, but the
natural distortion wasn't pleasant. The same amp and same tubes with a
typical supply (fender twin supply) had a much better tone when pushed to
clipping. If you have an intended program or fixed drive level that doesn't
push the amp into a hard clip, you should be fine, but you'll have to use your
best judgment listening if your after a natural distortion. I got a pair of kt88
and a pair of kt90 today for my amp. To me, the only reason to invest in
them was, to try and take advantage of PT rateing in milli-amps.
6gw8 have a 9w plate dissipation and 7591a are on par with a 6l6, it might
be worth while to stay within that family of tubes, or a plate dissipation of
around 19w max, if that was what your PT's intention was for.

Thanks for the reply Andy. Like you said the 7591s are 19w and the 6GW8 are 9 so that's about 56 watts total of plate dissapation, looking at it that way. They both come off the same winding, just different taps. So I think the PT will run at least 6L6GC at AB1. The Leslie 122 amp uses a pair of 6550s at about 450VDC and (I think) around 300 on the screens. It's sweet distortion with a Hammond. Different dynamics with guitar. You are right that it's probably best to cut off the 240 taps and insulate them.

Anyway, I have 40 and 50 watt transformers around and this OT has a larger core (iron) than any of them. The tube data for 7591s say they are good for 40watts at 450V. Hard to imagine at 19 watts of plate dissipation. Anyway, on the PT I think I'm good for a pair of EL-34s maxed out. Or KT-77. The OT is 6300/8 ohm though. Sort of an Express transformer but I want to do something Marshallesque for a change. Maybe a 2550 Silver Jubilee. Two parallel 8 ohm loads and see if it smokes. The PT has a separate bias winding too. Really a nice set of transformers. Came out of an E-100.

[Andy Le Blanc]

Sounds like you've got everything in the ballpark. I built a superlead pre
for a kid years back, clipper tube as a second stage driving a follower,
made his eyes roll back into his head the first time he hit a power chord.
Even in their simplest form, marshal based circuits can sound very nice.

Don't give up on your original design intent for the power side, but give your self an out
if it doesn't meet your expectation. You could fab two component boards
for the PS that could reduce the time reverting the project to whichever works
best to your ear. I'm constantly trying Ideas, that don't always work well for
guitar, but you can't know until you explore the concept. I'm always finding
that an amp built for guitar comes down to only a handful of design choices.

tube type, operating condition and loading, power supply (with attention paid to the screen girds),
inverter type and feedback application(s).

By "two parallel 8 ohm loads", are you intending to provide a 3k plate to plate
load for your power tubes? The trade off is more distortion, less power.
I might go the other way, try a OPT with a 3k primary impedance and mismatch
the load so that the tubes "see" something in the neighborhood of 6k.

[skyboltone]

By "two parallel 8 ohm loads", are you intending to provide a 3k plate to plate
load for your power tubes? The trade off is more distortion, less power.
I might go the other way, try a OPT with a 3k primary impedance and mismatch
the load so that the tubes "see" something in the neighborhood of 6k.

Well, that's what I thought too! I've always looked at the EL-34 as being a 6600 at 450 kind of an animal. Biased somewhere around 30V. But Marshall uses a 3400 load on the Silver Jubilee. It just doesn't seem right to me but what do I know. The OT was designed as 6300 into 8 ohms and size wise is easily a 50 watter so I'm with you. Initially I'll plug it in for the higher load impedence and see how that sounds. This should be fun I think. My ultimate goal is a stereo amp more or less with this arrangement on one side and an Express on the other. Or maybe this on one side and an AB763 on the other. Two voices, two OT, and a balance ccontrol.

Thanks for the feedback Andy, as usual I think you're on top of things.

Dan

[Andy Le Blanc]

Thank you, it drives me nutz when I cant find a good answer to a problem.

Loading poses a good question, your looking to get something around a 3.4k
load out of a 6.3k tranny. With 8ohms on the secondary, this means an impedance
ratio of around 787:1 and a turns ratio of 28:1. The 3.4k marshal with a 8ohm
secondary should have a 425:1 impedance ratio and a 20.6:1 turns ratio.
You can put a 4ohm load on the secondary of your 6.3k OPT to get the
effective plate-to-plate load resistance your looking for but you cant change
your turns ratio.

What does this mean for the performance of the amp? Does it create or make
worse, issues relating to inductance leakage in the OPT? or is it the opposite case?

[swarty]

I would have thought a Marshall Silver Jubilee OT would be identical to same period JCM800s.

[Firestorm]

Without getting technical, you have to consider that the screens actually act as the anode of the tube. Yeah, the electrons blow past them to hit the plate, but voltage of the screen grids set all that in motion. The screens are much closer-in and control the transconductance of the tube much more effectively than the plates. So you have to adjust the bias voltage differently that you would if the screens and plates were close to the same potential.

[skyboltone]

I would have thought a Marshall Silver Jubilee OT would be identical to same period JCM800s.

Take a look.

[Andy Le Blanc]

I did some digging and found one good reference that might support an argument
to use a different OPT. Current flowing through the primary winding causes
a voltage drop, IR. This is the product of primary current I, and winding resistance R.
The winding also presents a reactance X, which causes an IX drop.
This reactance is caused by leakage flux that does not link to the secondary.
The primary turns link this leakage and introduce an inductance
into the winding producing a leakage reactance X, at the frequencies your
trying to pass. The larger the primary current, the greater the leakage, and the greater the IX drop.
So if you intend to really push the amp, you will be
better off with a lower turns ratio, which will have a smaller IR/IX in its primary.
The energy wasted to induction has to go somewhere. OPT sunny side up.