2 tube micro head

[PeanutNore]

In figuring out the output transformer situation for a 12AU7 push pull power stage, I've lost any certainty that I had about the design of the rest of the amp.

Basically, my goal is to build a super low power head using parts that I have on hand. The limiting factors of the size of the aluminium enclosure I am using plus the fact that I only have 2 spare 9 pin sockets mean that I have only 2 tubes to work with. I basically need to pick one of the following configurations and I'm looking for input:

2 x standard 12AX7 gain stages + self split 12AU7 (Firefly amp)
EF86 + self split 12AU7
12AX7 cascode + self split 12AU7
1 standard 12AX7 gain stage + paraphase PI + push pull 12AU7
1 standard 12AX7 gain stage + cathodyne PI + push pull 12AU7

Or possibly the 2 gain stages, the EF86, or the cascode with a center tapped matching transformer serving as the PI (I have a few small 1:1 units that I could try)

My goal is to have a preamp that will not distort before the power amp does, but to have enough gain to be able to push the power amp into distortion. Some back of the envelope math leads me to believe that a single 12AX7 stage with a cathodyne or paraphase PI will not give enough gain to even get full clean power out of the 12AU7 so I am leaning towards skipping the PI and going with the self split design. However, if that is going to compromise the sound or limit power itself I can always use a boost pedal (I just finished an EF86 preamp pedal with 300v B+, sounds pretty sweet).

[J M Fahey]

Why not a single pentode Class A + 2 cascaded 12AX7 gain stages?

Output tube may be an EL84 if you wish or any "Radio/TV" classic pentode, there's lots of them , many OEM gathering dust and sold for peanuts simply because "Leo didn't use them" .

Dozens in the 1W to 2W range.

Best is that proper OTs are available, just grab that old radio in the basement or "buy" one for a couple dollars at Salvation Army .

A PI will "eat" 2 triodes by itself, plus all those ludicrous 2 signal triode power amps never impressed me ... to say it politely.

[potatofarmer]

Well, the Firefly is a design that's known to work. Build that first, and if you don't like it, tweak it until you do. Or rip the dozen or so parts out and build something else.

[eschertron]

You are correct that 2 12AX7 stages (or about any 12A_7 tube) cascaded will give more gain (50x50=2500) than a single pentode or cascode arrangement.

A couple of years ago I saw some 12AT7 micro amps on youtube, so I built me one. I used a 12AX7 for stages 1 and 2, followed by the 12AT7 in a parallel SE arrangement. I've thought about the self-split, but never went there. More recently I got a box of 12AU7s at a swap meet, and modified the amp to use them. Basically all I did was add a switch to go between two bias resistors (on the combined output triodes) to allow for using either. The AU7 has about 3x the input headroom of the AT7 in this rig. Different, but still nice.
I have a lot of gain reduction between stages (tone scuplting) to keep from running out of headroom too early. If there's a tone stack between the last gain stage and the power tube, it can't be 'too lossy'. I can't wait to hear some soundclips from your amp!

[PeanutNore]

I realized I had a couple JJ 12AT7 tubes sitting around doing nothing. I can use one in push pull instead of a 12AU7 with the same transformer and it will give me a power stage with a lot more gain. I should be able wire up a single 12AX7 gain stage and a cathodyne phase inverter and have enough signal to get a decent amount of overdrive. I'm not planning on using any tone stack, just a volume knob between the single gain stage and the phase inverter.

[wozt]

I have built several push pull micro heads using the ecll800 compactron. Two pentodes and a phase inverter in one envelope, ie 1 noval socket for all your output. Superb fast easy build.

[eschertron]

I realized I had a couple JJ 12AT7 tubes sitting around doing nothing. I can use one in push pull instead of a 12AU7 with the same transformer and it will give me a power stage with a lot more gain. I should be able wire up a single 12AX7 gain stage and a cathodyne phase inverter and have enough signal to get a decent amount of overdrive. I'm not planning on using any tone stack, just a volume knob between the single gain stage and the phase inverter.

+1. you can definitely drive the 12AT7 to clipping earlier than the 12AU7. Being able to swap between the two is like having 2 amps (almost). I should measure what I'm getting to the grids of the power section under 'normal' playing conditions. Since the 12AU7 stays cleaner, I'm guessing about 20vpp. Enough to make the 12AT7 start to sing!

[leadfootdriver]

Randall has a version with 2-12AX7's and a 12AU7 for the power. This may or may not be what you're after. Fast forward to 1:51 for the amp only demo. The tonestack switch is pretty cool.

It was speculated on another forum that they might use some diode clipping to get this level of distortion. But, you could get 3 gain stages with a plate driven tone stack, and then a totem pole phase inverter. Maybe you could get that amount of distortion out of it, but I don't know.

[PeanutNore]

Eschertron, what value were you using with the 12AT7 for the cathode resistor (and is that for a single triode or both), and what is your B+? It looks to me like 470R would be good for each triode (or 220 for the pair) where a 12AU7 would do better with more like 1K each.

[eschertron]

Eschertron, what value were you using with the 12AT7 for the cathode resistor (and is that for a single triode or both), and what is your B+? It looks to me like 470R would be good for each triode (or 220 for the pair) where a 12AU7 would do better with more like 1K each.

I will consult my diagrams, but IIRC I have 200R for the pair of AT7 triodes (close to your estimate of 220!) and 560R for the pair of AU7s. 560R is an odd number, yeah, but my switch added a 360R in series with the 200R. So again, 560R is close to half of 1K.
My B+ is around 270vdc. Also, since I have 2 triodes in parallel, the 'effective' primary impedance, when doing my load lines, is twice what the actual OT primary is set for. Again, IIRC, it is 20k for the primary (might be only 10k though, I fussed with it until I got what I liked).

[PeanutNore]

This is what I'm looking at for the output stage. It seems to me that it should work pretty well. I might use a single 470R resistor for a pair of 12AU7 triodes and include a switch to add a second one in parallel for 12AT7. I expect to mainly use it with a 12AT7 though, considering that the entire preamp will only have a gain of 80, making the higher gain of the 12AT7 more important if I want to get any OD.

[eschertron]

What gain structure did you decide on for the preamp?

[PeanutNore]

For the preamp, I am going to use a single 12AX7 triode configured for high gain - 270k plate, 2.2k cathode, 22uF bypass, a 1M gain pot, and the other triode as a cathodyne phase inverter. I'm expecting that this amount of gain will keep the output fairly clean through most of the knob but be able to overdrive the 12AT7 when turned up far enough.

[eschertron]

For the preamp, I am going to use a single 12AX7 triode configured for high gain - 270k plate, 2.2k cathode, 22uF bypass, a 1M gain pot, and the other triode as a cathodyne phase inverter. I'm expecting that this amount of gain will keep the output fairly clean through most of the knob but be able to overdrive the 12AT7 when turned up far enough.

I referred to Merlin's preamp book to confirm my suspicion about the cathodyne. you will get 1/2 of "very nearly" unity gain from each of the output taps (plate and cathode) of the PI. You're not going to gain anything (pun intended) with that over a self-split arrangement. Same gain structure. Half of the preamp signal goes to each tube in the push-pull. I haven't tried a self-split arrangement, but I'd certainly be tempted to do so here, to save the second triode for other purposes.

IMHO you will be throwing a gain stage away. If you're looking for an 'at the edge of breakup' sound out of the AT7s, you'll get that. But nothing near clipping unless you slam the front end of the amp. JM2C

[PeanutNore]

I couldn't get the phase splitter working at all anyway, so I ended up going with the self split and a cascode preamp because I wanted to be able to play the amp today. I fired it up and it works pretty well - no hum or audible noise, clean until about two thirds of the way up, nice amount of overdrive cranked. The only thing I need to fix still is the B+. The tiny transformer I am using to step up the B+ voltage from a 12v AC wall wart is regulating a lot and not giving me the voltage I assumed, so I'm running on 202 volts when I planned for 260. I have a slightly different transformer from the same series that I'm going to wire up differently to get a higher B+. They are dual primary, dual secondary transformers for 115/230 volts stepped down to 12.6/6.3 volts that I have wired backwards. Right now I have the primaries and secondaries both in series (230v, 12.6v). I have a 16/8v secondary model that I am going to wire the low voltage side in parallel on (230v, 8v). 12.6v on the 8v side should give me closer to what I'm looking for. They are 2.5VA units and only supply the B+ since the heaters run off the wall wart. Playing it for a half hour earlier the PT didn't even get warm.