New amp build design - lots of features

[paggerman]

Greetings all. I am about to embark on a new build and, prior to getting the soldering iron out, I thought I would ask for opinions on the potential of this buildand any pitfalls that I may not be aware of. This will be my 7th or 8th build so I have a reasonable amount of experience. I will be building this point to point using socket pins wherever possible.

the basis of the build is an unknown organ chassis that I acquired a few months ago. it came with a pair of 7591A valves and has another octal and a noval socket. there were two power transformers on the chassis, one of which seems to be the PT for other parts of the original organ. the PT that powered the power amp chassis has three secondaries @ 5V, 6.3V and 164V. the remains of the original circuitry had what I believe was a voltage doubler for B+.

so now I want to build an amp that puts together lots of featrures. a friend of mine said that he wanted an amp that has lots of 'weird stuff' in it so I am building this with that in mind, though it's not necessarily for him.

I have put in a number of features from Merlin's preamp book. Those familiar with it will recognise the switchable parallel/cascade stage, the EF86 morph dial & the mid lift. the power stage is quite generic.

Before I put the solder to joint, any review of the diagram and any suggestions would be welcome. I would particularly appreciate commentary on the voltage doubler stage as I have had to derive this from a number of diagrams. I just feel kind of uncomfortable having one end of the secondary windings directly connected to the chassis.

Thanks everyone - it's always a pleasure to read this board.

Jeff

[soundguruman]

I think the pentode should be the first stage.

[Steve Conner]

The voltage doubler circuit will work fine, but there's a slightly different version that gives 120Hz ripple instead of 60, and the caps only need to be rated for half the output voltage.

It's figure 4 in the Wikipedia article.
en.wikipedia.org/wiki/Voltage_doubler

[paggerman]

Thanks for that. I toyed with the idea of having the EF86 as the first gain stage as used in early Vox amps. But I figured that by having a bigger signal into the pentode would give more of a pentode feel to the sound if so desired. I could always rejig it if I don't like the outcome. Your comment is appreciated.

[paggerman]

The voltage doubler circuit will work fine, but there's a slightly different version that gives 120Hz ripple instead of 60, and the caps only need to be rated for half the output voltage.

It's figure 4 in the Wikipedia article.
en.wikipedia.org/wiki/Voltage_doubler

Thanks Steve.
I'm more likely to get 50/100 Hz ripple in Australia, but point taken. A for voltage ratings on the caps, I already have the 450v rated ones, but certainly useful info for future reference.

[mooreamps]

I think the pentode should be the first stage.

Yes, I agree...

[soundguruman]

What builders fail to realize is that a guitar amp must have very high input sensitivity. In other words, a 50 mv input (or, less) should be able to drive the amp to a full output. In older amps, 100mv would drive the amp to a full output (super lead). With a pentode as a first stage (Operadio), you find this outrageous sensitivity that a triode can never live up to. When we try to drive a triode this hard, it breaks into unstable operation. (What Randall calls: "non-linear" operation). The stages no longer exhibit the quality of inverting and non inverting. The stages are simply clipped. (mesa boogie) (welcome to my mud pie)
A pentode can be driven this hard all day long without ever breaking a sweat, and still maintain the linear operation (Dukane).
In this amp, with a single stage triode front end, the sensitivity desired by guitar players is absent. Essentially what you have is a very clean amp, like a Silvertone. The amp will sound like it has no guts, bland, but it would probably be a great hi fi amp. (and a great surf guitar amp)
The design does have something good going for it- no 12AX7s. At least the designer abandoned the mundane and used some higher quality, lower noise valves. That's the part I like. There are so many great audio valves out there, which have so much potential for guitar- and these are largely ignored.
But, what do you think the triode / pentode control is going to do? (not squat) Because the pentode is in the wrong position.

The 6J7 is dripping with overtones and fine harmonics, the ideal guitar / harp preamp tube, with the grid cap, allowing much greater physical separation in construction between the input and output nodes. Add to that, the nearly shrapnel proof metal envelope. I don't need to install a shield over the tube. JAN rules the Galaxy!
The pentode tradition continued with the newer design pentode preamp tubes. (what's with the glass you guys?)

[J M Fahey]

Whether that pentode is the first or last stage is irrelevant, the signal on its plate goes nowhere.
And its screen resistor is non decoupled.
Useful on power tubes, but ... on preamp ones?
And donīt get the idea of AC coupling the suppressor grid ... and nowhere providing a DC path for it, so it can't fulfill its main use.

[J M Fahey]

What builders fail to realize is that a guitar amp must have very high input sensitivity.

No kidding !!
Somebody should rush and tell Marshall, Soldano, Mesa, Engl, Bogner, Ampeg, Peavey, Diezel, ..... etc etc etc.

In other words, a 50 mv input (or, less) should be able to drive the amp to a full output. In older amps, 100mv would drive the amp to a full output (super lead). With a pentode as a first stage (Operadio), you find this outrageous sensitivity that a triode can never live up to.

Where have you been living the last 60 years?

When we try to drive a triode this hard, it breaks into unstable operation. (What Randall calls: "non-linear" operation). The stages no longer exhibit the quality of inverting and non inverting. The stages are simply clipped. (mesa boogie) (welcome to my mud pie)
A pentode can be driven this hard all day long without ever breaking a sweat, and still maintain the linear operation (Dukane).

Sheer nonsense and lack of basic Math skills .
Given the same +B supply, say 250/300V , a triode can have around 50X gain, a pentode will be around 200X gain.
The pentode will be 4 (four) times easier to overdrive with an input signal.

Add to that, the nearly shrapnel proof metal envelope.

One of my first projects (looooooooooooong ago) was to reform a surplus WW2 vintage Sherman tank radio transciever.
It had a Metal 6L6 output tube.
NOW I know it was to make it bullet proof!!
Dumb guy !!!

The pentode tradition continued with the newer design pentode preamp tubes. (what's with the glass you guys?)

Refer to earlier quote: "where have you been the last 60 years?"

[Steve Conner]

I think soundguruman has been at the hemp cones again.

When I built an amp with a pentode in it, I also used the EF86 as the first stage and 12AX7s for the later ones. I used the EF86 as a booster with a bass cut like a Rangemaster or a Vox Top Boost, in front of an otherwise stock circuit.

But I've seen it done with a triode first and the pentode as the overdriven stage too. It's purely down to taste. I don't doubt that you can make a good sounding amp entirely with 6J7s. Although ironically, the last schematic I saw had the 6J7s triode connected.

Yes, we have 50 and 100Hz ripple over here too.

[tubeswell]

With that load on the EF86 you will lose lots of highs. Better to have that TS after the first two stages. Then have the EF86 directly pushing the PI. (Or maybe better still, have the first two stages set up like a blackface amp with the TS in between the two and then have the TS recovery stage directly driving the EF86)

Also seeing as how you have the morph control, this effectively can make the EF86 act like a plain triode or a pentode, in which case you may find that using the driver/cathodyne setup is way too over-the-top in pentode mode. You would probably get a more useable range of distortion sounds if the EF86 was directly driving a LTP. That way you would also have optimal impedance bridging for the EF86 output impedance. If you decide to use a LTP, then maybe copy the 5G9 LTP circuit values (since you are driving 6V6s)

JM2CW

[Bruce / Mission Amps]

I guess I'm dumb too.. or just not seeing what the pentode is doing... maybe I should look at Merlin's pages too... is it not for any amplification?

[tubeswell]

I guess I'm dumb too.. or just not seeing what the pentode is doing... maybe I should look at Merlin's pages too... is it not for any amplification?

One end of the pot rotation (of that pot at the 'bottom' of the screen bypass cap) ensures the screen bypass cap goes straight to ground, making the pentode function as normal. The other end of that pot rotation couples the screen to the plate at AC, effectively making the pentode 'act' like a triode (the other cap is there to keep DC off the pot). And the in-between parts of the pot rotation provides something in between, with a bit of unbypassed screen compression going on nearer the 'fully screen bypassed' end of the pot rotation. Is that what you were asking? (So its effectively a fancy gain control).

[J M Fahey]

Dear Tubeswell.
Didnīt you notice that pentode's plate signal goes *nowhere*?
That the control grid (closest to cathode) is AC coupled and has no DC reference to ground? (So Bias is impossible).
That the screen grid (the middle one) is DC connected to ground, not acting as a screen grid at all?
That the suppressor grid (closest to plate) is connected to +B through a 560K resistor, instead of being close to ground potential?

Are we all looking at the same schematic?
I mean the one shown in post #1.

I fail to see the "screen bypass cap".
And if you find the grid drawn closest to the cathode *is* the screen (not the usual way to draw it, that place is usually reserved for the control grid), where does that "screen" get its high DC voltage?

I just don't get how People can discuss the sonic effect of an unconnected and unbiased tube.

[diagrammatiks]

What builders fail to realize is that a guitar amp must have very high input sensitivity. In other words, a 50 mv input (or, less) should be able to drive the amp to a full output. In older amps, 100mv would drive the amp to a full output (super lead). With a pentode as a first stage (Operadio), you find this outrageous sensitivity that a triode can never live up to. When we try to drive a triode this hard, it breaks into unstable operation. (What Randall calls: "non-linear" operation). The stages no longer exhibit the quality of inverting and non inverting. The stages are simply clipped. (mesa boogie) (welcome to my mud pie)
A pentode can be driven this hard all day long without ever breaking a sweat, and still maintain the linear operation (Dukane).
In this amp, with a single stage triode front end, the sensitivity desired by guitar players is absent. Essentially what you have is a very clean amp, like a Silvertone. The amp will sound like it has no guts, bland, but it would probably be a great hi fi amp. (and a great surf guitar amp)
The design does have something good going for it- no 12AX7s. At least the designer abandoned the mundane and used some higher quality, lower noise valves. That's the part I like. There are so many great audio valves out there, which have so much potential for guitar- and these are largely ignored.
But, what do you think the triode / pentode control is going to do? (not squat) Because the pentode is in the wrong position.

The 6J7 is dripping with overtones and fine harmonics, the ideal guitar / harp preamp tube, with the grid cap, allowing much greater physical separation in construction between the input and output nodes. Add to that, the nearly shrapnel proof metal envelope. I don't need to install a shield over the tube. JAN rules the Galaxy!
The pentode tradition continued with the newer design pentode preamp tubes. (what's with the glass you guys?)

facepalm.