What Tubes Can I Use?

[pdf64]

...The screen voltage limit for UL is often higher than for standard operation as the transformer coupling forces the instantaneous screen voltage to follow the plate voltage, both changing in-phase. Consequently plate voltage never gets lower than screen voltage. In standard PP designs high instantaneous plate current causes the plate voltage to be lower than the screen voltage making the screen an attractive target for the electrons. The consequent increase of screen current and dissipation is the reason for the lower screen voltage limit...

I think you may have got that bit about UL (plate voltage never gets lower than screen voltage) wrong, and you may have mixed up UL with triode mode?
Rather for both UL and regular pentode p-p operation, with signal, the instantaneous plate voltage is lower than that of g2.
But with UL, g2 voltage, though still closer to HT voltage than plate voltage, is somewhere between the 2.
ie the primary tap it's connected to effectively forms a reactive potential divider.
eg so if the HT is 400V and the plate has been pulled down to 100V, a 43% tap would result in g2 being at 279V.

That reduced a - g2 differential reduces g2 current and hence dissipation, so allowing a higher g2 voltage limit in UL than regular pentode mode.

Yikes, hope I've got that right

[pdf64]

...I also noticed last night while looking at the layout I used ,and some pictures I had taken ,that I may have left out a connection point in the amp. The amp works with little to no noise ,so I'm not sure exactly how this would affect the amp? That cap goes to ground ,and the connection between that resistor ,and cap may be missing. I'm going to look to be sure I didn't come back ,and add that at some point. The part in the circled are is what I may have left out.

Yikes, are you sure that the link isn't underboard?
If not, then there would be no HT supply to V2, so yes, I can't see how the amp could pass signal
A simple voltage survey (even just at the cap and V2 socket terminals, idle Vdc at each circuit node) would prove it one way or another.
See the 'valve voltage chart' for an example of a voltage survey https://drtube.com/schematics/marshall/jtm45tr.gif

[gtrplayr1976]

Yikes, are you sure that the link isn't underboard?
If not, then there would be no HT supply to V2, so yes, I can't see how the amp could pass signal
A simple voltage survey (even just at the cap and V2 socket terminals, idle Vdc at each circuit node) would prove it one way or another.
See the 'valve voltage chart' for an example of a voltage survey https://drtube.com/schematics/marshall/jtm45tr.gif

I'm going to check it. I don't think I put any links under the board. I wanted to be able to see everything for troubleshooting. I think it's working off PFM.

[gtrplayr1976]

Yikes, are you sure that the link isn't underboard?
If not, then there would be no HT supply to V2, so yes, I can't see how the amp could pass signal
A simple voltage survey (even just at the cap and V2 socket terminals, idle Vdc at each circuit node) would prove it one way or another.
See the 'valve voltage chart' for an example of a voltage survey https://drtube.com/schematics/marshall/jtm45tr.gif

It HAS to be there. I have voltages on the V2 terminals. Pin 1-197 , pin 6-337 ,pin 7-197.

[gtrplayr1976]

Just to note that I too find the JTM45 circuit way too bassy, but dislike the response from its tone stack if the treble is turned much above halfway (linear pot). However, slightly reduced value coupling caps between the input stages and their vol controls (10nF normal, 3n3F bright) seems to reduce the extreme bass extension and clean up subsonic hash when it's overdriven. And a bigger bright cap, at least 220pF, on the channel vol controls. I didn't like the treble peaking cap across the mixer resistor that some version use.

But the big saving grace with the circuit is the extreme degree of treble boost available when the presence is turned up full, which is where I leave it, permanently. And I now prefer my JTM45 type amp over even my old Vox AC30.
Though with mine, compared to the OP's, I think that KT66 and the original 'shared+individual' screen grid resistor arrangement are key to that preference.

Pertinent to this thread is that I noticed the OP's circuit uses the original 27k/5k NFB ratio, but taken from the 8 ohm tap, rather than the 16ohm. So the degree of NFB, and hence presence boost available, is reduced somewhat (~30%?).
But more significant than that is the use of a dual track master volume between the LTP phase splitter and power tubes, as I understand things?
As that is turned down, the power amp's loop gain will reduce, and along with that the available degree of boost from the presence control.
Hence I feel that given the issue (amp way too bassy), such a master volume is unsuitable and perhaps may be part of the problem, as it disables a very necessary lifeline, that in the amp's stock form is essential to brighten up the amp's response.

And all this stuff with putting bright caps on to that master vol is just akin to trying to polish a turd of a mod
If a master volume is wanted, then start with a proven good design that uses one (eg 2204), rather than hammering a square peg into a round hole by trying to get good results from fitting a master vol in a classic non master vol circuit.
But that's only my opinion, whatever works for the OP is good.

Are these the caps that you would change?

[Helmholtz]

Rather for both UL and regular pentode p-p operation, with signal, the instantaneous plate voltage is lower than that of g2.

You are absolutely right!
Thanks for clarification and making me think about it again. My mistake was to only consider the AC circuit, which shows that screen signal amplitude must be lower than plate amplitude. But as both signals are actually referenced to HT (being an AC ground), we have Vp(t) = HT - plate signal < Vg2(t) = HT - screen signal, meaning that always Vg2(t) >= Vp(t).
But as for high plate current plate AND screen voltages are low(est), the most critical phase of standard PP operation with high screen voltage is mitigated.

[pdf64]

Are these the caps that you would change?

Yes, plus I forgot to mention that I changed the coupling caps between the LTP and power tube grids to 47nF (were 0.1uF).
Also, move the NFB wire from the 8 to the 16 ohm terminal of the impedance switch.
And leave the master vol up full so that presence do its thing properly.

If you must have a master vol, then use a type 4, regular 2204 style, between tone stack and LTP, as that won't interfere with the power amp. https://drtube.com/schematics/marshall/2204u.gif
But I suggest not, as it leads to the channel volumes being turned up, which defeats the benefit of a bright cap; if you use the channel vol much past 50% (10% audio taper pot), then I prefer 470pF bright cap with my Les Paul.

Obviously you now have to manage a rather loud amp.

[gtrplayr1976]

Yes, plus I forgot to mention that I changed the coupling caps between the LTP and power tube grids to 47nF (were 0.1uF).
Also, move the NFB wire from the 8 to the 16 ohm terminal of the impedance switch.
And leave the master vol up full so that presence do its thing properly.

If you must have a master vol, then use a type 4, regular 2204 style, between tone stack and LTP, as that won't interfere with the power amp. https://drtube.com/schematics/marshall/2204u.gif
But I suggest not, as it leads to the channel volumes being turned up, which defeats the benefit of a bright cap; if you use the channel vol much past 50% (10% audio taper pot), then I prefer 470pF bright cap with my Les Paul.

Obviously you now have to manage a rather loud amp.

I'll give that a go. The NFB is already on the 16 ohm terminal I believe. I'll check again to be sure.

I like loud.

[gtrplayr1976]

The only .033 I have are the square radial Nichicon ,or some axial I got from Mouser ,but only rated at 250V.

[gtrplayr1976]

Just to note that I too find the JTM45 circuit way too bassy, but dislike the response from its tone stack if the treble is turned much above halfway (linear pot). However, slightly reduced value coupling caps between the input stages and their vol controls (10nF normal, 3n3F bright) seems to reduce the extreme bass extension and clean up subsonic hash when it's overdriven. And a bigger bright cap, at least 220pF, on the channel vol controls. I didn't like the treble peaking cap across the mixer resistor that some version use.

But the big saving grace with the circuit is the extreme degree of treble boost available when the presence is turned up full, which is where I leave it, permanently. And I now prefer my JTM45 type amp over even my old Vox AC30.
Though with mine, compared to the OP's, I think that KT66 and the original 'shared+individual' screen grid resistor arrangement are key to that preference.

Pertinent to this thread is that I noticed the OP's circuit uses the original 27k/5k NFB ratio, but taken from the 8 ohm tap, rather than the 16ohm. So the degree of NFB, and hence presence boost available, is reduced somewhat (~30%?).
But more significant than that is the use of a dual track master volume between the LTP phase splitter and power tubes, as I understand things?
As that is turned down, the power amp's loop gain will reduce, and along with that the available degree of boost from the presence control.
Hence I feel that given the issue (amp way too bassy), such a master volume is unsuitable and perhaps may be part of the problem, as it disables a very necessary lifeline, that in the amp's stock form is essential to brighten up the amp's response.

And all this stuff with putting bright caps on to that master vol is just akin to trying to polish a turd of a mod
If a master volume is wanted, then start with a proven good design that uses one (eg 2204), rather than hammering a square peg into a round hole by trying to get good results from fitting a master vol in a classic non master vol circuit.
But that's only my opinion, whatever works for the OP is good.

Should I remove the treble peaking cap as well?

[pdf64]

Should I remove the treble peaking cap as well?

Don’t go on my say so please, try it with and without, to find out what works best for you.
That cap makes the bright channel brighter, but at the expense of making the normal channel muffled.
I use an AB box with both channels, so it’s no good for me.
But if just using the bright channel then you may like its effect.

[gtrplayr1976]

Don’t go on my say so please, try it with and without, to find out what works best for you.
That cap makes the bright channel brighter, but at the expense of making the normal channel muffled.
I use an AB box with both channels, so it’s no good for me.
But if just using the bright channel then you may like its effect.

I usually jump the channels. I like that effect. What do you think about tubes? If you read what I posted previously ,I can't use KT66s. I currently have 5881s in it ,and will be trying 6l6gc ,and KT77 again.

[Helmholtz]

I think that KT66 and the original 'shared+individual' screen grid resistor arrangement are key to that preference.

What do you use for the shared screen resistor? 1k?

[pdf64]

Yes, the standard 1k shared feeding 470 ohm individual. I tried the 1k shared feeding 100 ohm individual, and couldn’t see/hear/measure any benefit.

[gtrplayr1976]

What do you use for the shared screen resistor? 1k?

Mine is 1K as well as the individual ones.