Design considerations for high voltage, high power, tube output section/PSU

[nsubulysses]

I have been thinking of trying to build an amp to suit my needs. It would be pretty fun and hopefully pulverizing. I think it could turn out good

My idea is to run 4x6CA7 at high voltage for an amp I would hope to do 120-130W at clipping. I can get a PT from Heyboer to accomodate this and I was thinking a loaded B+ of 550-600V could work good. Heyboer recommended an OT that is a Hiwatt style 150W so it could handle 120W or so without too much hassle

According to JJ datasheet 6CA7 have max Va 800V and max Vg2 of 500V. I am not sure if I should worry about exceeding the screen voltage because I typically run JJ 6L6GC in Ampeg V4s Va and Vg2 in excess of about 90V of datasheet limits. Datasheet lists JJ 6L6GC 450V max for Va and Vg2 and Ampeg V4 runs both at about 530-540V. If you bias cold (around 50%) there is no crossover notch and no blowout IME. I assume I could do something similar but with even a bit more power with 6CA7 and higher PSU voltages. I invite you to tip me off on my future headaches if I decide to build this amp.

My favorite amps to play are Sound City 120, Ampeg V4, and Sunn Model T (reissue and original), so I wanted to build something as powerful or so, but with a high gain preamp similar to reissue Model T or Soldano Avenger. That's kind of an aside though because what I really want to know is if there are shortcomings to my thoughts on output section/PSU design considering this high voltage 4x6CA7 arrangement and if other considerations need to be made.

I am not a veteran amp builder by any means but I have made a few. This is the first time I'll try to make something for myself though, rather than building something that is 90% pre-existing designs. Of course it's all based on pre-existing designs anyway, so whatever.

Schematics I'm researching are attached if anyone wants to see them.

Soldano Avenger preamp
Soldano Avenger power

Sunn Model T preamp (reissue)
Sunn Model T power (reissue)

[KevinOConnor]

Hi Guys

In the Hammond transformer line, the 278CX PT and 1650TA OT will easily yield 160Wrms with any four large-bottle power tubes. This circuit is shown in TUT3 as the Custom Special project, based on Traynor's original amps (which were based on these Hammond transformers).

Va=560V or so, unloaded can be closer to 600V, so caps in series for Va and Vs.

The circuit is essentially just a plexi and is loud and clean.

Like i said, you can use any large bottle tubes in this circuit, including 6L6, 6CA7, EL-34, 6550, KT-66/77/88/90/100 - even mixed tube types. The key is to use 1k-5W individual screen resistors and provide individual bias adjusts and meter jacks.

Of course, you can use any preamp circuit you wish to, add an effects loop, reverb, whatever suits your taste. There is about 30-lbs of iron here.

Have fun

[loudthud]

In the Hammond transformer line, the 278CX PT and 1650TA OT will easily yield 160Wrms with any four large-bottle power tubes. This circuit is shown in TUT3 as the Custom Special project, based on Traynor's original amps (which were based on these Hammond transformers).

Va=560V or so, unloaded can be closer to 600V, so caps in series for Va and Vs.

The circuit is essentially just a plexi and is loud and clean.

Kevin, Of the original Custom Specials I have seen in the US, none had a plexi type preamp. None had more than two inputs and I've never seen a Traynor schematic of a YBA-3 with a plexi preamp like the YBA-1 or 1A. Perhaps plexi preamp examples only shipped to Canada?

[KevinOConnor]

Hi Guys

The Custom Special (160W, 4-tube) and Bass Master (80W, 2-tube high-V; 40W low-V) both began as plexi circuits - like the high-voltage Marshalls of the day - nothing special. The only variation was the suppressor connection and the shared screen resistor, both hifi methods. The CS and the later Super CS had more elaborate preamps with passive and active EQ. TUT3 shows all the variations in schematic form, and these schemos are available on the web, too.

Regardless of your Traynor experience, the transformer set cited above will deliver 160W with no strain from the tubes. See the technical Article on my site of the expanded Hammond data, which includes a list of OT-PT pairs that work together..

have fun

[nsubulysses]

Quick question, with a PT that has center tapped high voltage winding is it necessary to use it or can the center tap be left disconnected and used with a full wave bridge rectifier configuration?

Or is it best to just do a two phase rectifier, which I guess it seems to be designed for, since it has the center tap?

Hammond 278CX

[KevinOConnor]

Hi Guys

Should be 1650T in my post above. Oop!

The CTed plate windings you see in the "classic" line from Hammond are from an era when a tube rectifier was typically used. However, you get oddities in the line that do not have a winding for the rectifier heater... so what to do? You can use a separate PT for the heater or go to solid-state rectifiers.

With any such winding, you can use one end and the CT and tape off the other end, then use a full bridge rectifier. This has an advantage in that the rectifier only requires half the voltage rating that the half-bridge would need using the whole winding and the CT. With the full bridge, it generally works out that the voltage ratio from AC to DC seems to follow root-2 as expected.

Either way, diodes are inexpensive and placing two in series to withstand the whole winding voltage as in the half-bridge setup or going to a full bridge uses four diodes either way. It's now worth the expense to go to Hexfreds or similar. Just use UF4007s or UF5408s if you want a reasonably fast diode.

Have fun

[J M Fahey]

This circuit is shown in TUT3 as the Custom Special project
Have fun

Hi Guys
The Custom Special (160W, 4-tube) and Bass Master (80W, 2-tube high-V; 40W low-V) both began as plexi circuits - like the high-voltage Marshalls of the day - nothing special. The only variation was the suppressor connection and the shared screen resistor, both hifi methods. The CS and the later Super CS had more elaborate preamps with passive and active EQ. TUT3 shows all the variations have fun

Mmmmhhhhh, not easy to follow such advice since it lies elsewhere and presumably has to be ordered and paid for.

I might be wrong, of course, and such information might be freely and easily accesed, in which case I would stand corrected.

It would be kind and appreciated if the relevant info were cut and pasted here or at least some clickable direct link were provided.

[KevinOConnor]

Hi Guys

Juan: You seem to have a serious reading comprehension impairment. My posts have info even if they have a reference to the books, where much more detail already exists, and also there is a reference to free info on the web for those who wish to look for it. Please read more carefully.

You don't seem to have read post-6 where I corrected the typo from post-4. You have your own reasons for doing that, I suppose. If one wants to extract maximum power from the 278CX, the 1650T or TA version is the match that will achieve that end. These are the prototype transformers Pete used when designing the Custom Special. Where he elected not to use the UL taps and then had custom OTs made without them, the OP here does want those taps. The CS circuit can be used with UL-connected OT just as easily and with just as good results as with the pentode-wired output stage.

Have fun

[nsubulysses]

Sadly that amp has a terrible low frequency oscillation I have not been able to locate the origin of. It is I believe anywhere from about 150Hz to 300Hz depending on control settings. The transformer buzzes audibly at about 180Hz when you turn the power on.

I tried switching OT leads and switching NFB input already and it oscillates differently and worse with these switched.

I can silence the noise by shunting the .022 coupling cap after the first gain stage to ground. If I shunt the grid of the input of the stage 1 to ground the problem remains. I tried shielding tubes and transformers with other chassis panels and/or aluminum foil with an alligator clipped to chassis and there was no change also. trying to shield preamp and power tubes, and transformers.

Not really sure what to do next and I really wanted to use this for band practice tonight



Edit: Kevin was right too. This transformer combination does do 160W at clipping with four 6CA7. Even more than I hoped for, which is great. I wanted something that would hang with my SOund City 120 so this will have no problem.

[nsubulysses]

I got a pre-fabbed chassis from a friend for really cheap to do up the prototype but the way it was fabbed kind of dictated that I put the PT and PS on one side, and I would have rather had it on the other side. This makes my run to the phase inverter long and some other parts of the layout a little weird. Perhaps this is part of the problem since V1 is on the PT side

Here is a labeled pic

[catalin gramada]

Hallo. just at quick first sight. Take off the input jack and move it in the hole between gain pot and power switches. It is no ideal but just can give a suggestion if your problem gone. If is so further you can move away two holes to the right those switches in far empty holes and screen the input jack with a shield.

[nsubulysses]

I tested that by removing the input jack completely from the circuit and grounding the grid of V1 and it still did it so I think that's not the problem. I wondered that too. Yeah this layout is a little weird I kind of regret using this chassis but it was cheap, pre drilled and just down the street

V1 is parallel but unparalleling it made no difference either.

[KevinOConnor]

Hi nsub

With the chassis punching already done determining which end the mains wiring will be on, it would have been sensible to just build the circuit in the order already dictated, placing the sensitive input on the opposite end even if you think this is a "backwards" control order. Personally, I dislike the backwards arrangement Marshall used, which happened naturally by flipping over a forward-control-order Fender chassis. A lot of work to make a half-corrected layout.

The input jack should not have a line to a star ground - which itself might be the cause of the oscillation. Star grounds are the exact opposite of what is needed in a tube circuit and especially one with high gain. The star assures that clean signals mingle with noisy ones and that in-phase signals mix together, all of which contributes to a noisy amp and potentially to oscillation.

Since you have a scope, you can isolate circuit sections and see where the oscillation is coming from. This is entirely a layout issue and some of your nice work will have to be redone. You could begin with the power amp or the preamp - I tend to start with the power amp since it is a simple block unto itself. Isolate it from the preamp and see if it is stable on its own. If it is not, break the feedback loop first and see if it is still unstable. If it is, there is likely an issue with the splitter wiring and/or its proximity to the OT wiring. This is one reason why I don't like Schmitt splitter - there are both phases of signal at the input and the output, so it is easy to have the splitter oscillate unto itself. Once the PA is stable open-loop, try closing the loop and see what happens. You might have to change the compensation cap in the splitter and/or change coupling cap values to achieve stability. The oscillation is likely there with a resistive bench load, so no need to have a speaker connected until you get rid of the oscillation quietly.

Those orange-drops are a bad choice as their bulk and inductive winding can couple signals from one cap to another. The only advantage their bulk has is to force an airy layout that is likely to sound good due to low parasitic capacitance.

An actual schematic of the preamp would be good. What is a .psd file? my computer can't open those.

Have fun

[Dave H]

What is a .psd file? my computer can't open those.

Photoshop

Edit:
First two schematics from post #1

[nickb]

The star assures that clean signals mingle with noisy ones and that in-phase signals mix together, all of which contributes to a noisy amp and potentially to oscillation.

I'm having some difficulty in imaging a situation where that would be true. Perhaps you could illustrate how this could be so with a schematic example?