Any examples of a 12AU7 LTP phase inverter?

[Raezzor]

Spitballing a PP guitar amp using a pair of 6EH5/6CA5 for power tubes. Little suckers only need like 4.5 volts peak to get to full power so I don't necessarily want to use a high gain PI like a 12AX7 LTP and have to attenuate the signal either before or after the PI. So I got to thinking about using a lower mu twin triode like the 12AU7 but for the life of me I can't find any examples of them being used as a PI let alone as an LTP. Any advice?

[sluckey]

The Magnatone M10A uses a 12AU7 cathodyne PI.

https://sluckeyamps.com/magnatone/Magnatone_M10A.pdf

[Stevem]

A 200 watt Marshall Major uses that for a PI.

[sluckey]

A 200 watt Marshall Major uses that for a PI.

No. V3 is just a driver. The triode just prior to V3 is the cathodyne PI. That triode is V2 and it's a 12AX7. There is an error on that cathodyne schematic.

[johnnyreece]

There is an error on that cathodyne schematic.

Is it where the 1M off the grid is shown jumping the (I think it says) 2.7k/100k junction coming off the cathode, rather than connecting there?

Edit: Added picture, as my word salad above barely made sense to me, and I wrote it...

[sluckey]

That's it. Your word salad made perfect sense to me.
Mod Edit: I think this is in response to a deleted user/comment so it probably no longer makes sense

[Raezzor]

Dang, some weird designs using cathodyne PIs lol. Guess everyone wants more juice with an LTP and so just naturally use a 12AX7.

Oh well, time to throw something together and tinker until it works I guess! Shouldn't be too hard to get in the ballpark I guess using the standard values for a 12AX7 LTP PI and just sub in a 12AU7 right?

[johnnyreece]

I did exactly that in my 4-4-0. 12ax7 drives those little triodes too hard. I was too lazy to modify the circuit, so I swapped that in. Much better.

[cdemike]

The Vox AC100 used a 12AU7 in a LTPI:

[sluckey]

The Vox AC100 used a 12AU7 in a LTPI:

That's not a LTP PI. It's called a floating paraphase. Look at the last paragraph of this short page...

https://www.valvewizard.co.uk/paraphase.html

[cdemike]

The Vox AC100 used a 12AU7 in a LTPI:

That's not a LTP PI. It's called a floating paraphase. Look at the last paragraph of this short page...

https://www.valvewizard.co.uk/paraphase.html

Whoops! Thank you for the correction, sluckey! Sloppy of me not to look closer or read the entire page -- my apologies, Raezzor. FWIW, assuming you have adequate B+ voltage, you likely could make a long tailed phase inverter borrowing the same operating points -- 47k feeding each plate, 1k5 bias resistor, tail resistor to taste.

Is there a reason you're not considering a cathodyne over the LTPI to reduce gain?

[Roe]

The Vox AC100 used a 12AU7 in a LTPI:

one of the greatest amps of all time. but there are many variations. later versions, introduced ecc83 and LTP (some are dubbed ac120 and ac125)

[B Ingram]

... PP guitar amp using a pair of 6EH5/6CA5 for power tubes. Little suckers only need like 4.5 volts peak to get to full power ... lower mu twin triode like the 12AU7 but for the life of me I can't find any examples of them being used as a PI let alone as an LTP. ...

You probably can't find many examples of push-pull 6EH5/6CA5, where 3-5 volts peak is enough to drive the snot out of the power section. So you're already venturing in less-charted waters.

Do observe a few things in RCA's Application Note on this output tube: sloppy wiring will get it in trouble, and the supply voltage can't be over about 140v.

A "long-tail" inverter wastes a fair amount of supply voltage across the "tail" in order to achieve output-balance. Do you really have the voltage available to "waste" in that type of circuit?

I would probably use a paraphase inverter, akin to what Sluckey noted was used in the AC100. Easy to think-through, easy to wire up. It's only limitation compared to a long-tail inverter is a ready-spot to inject negative feedback around the power section (but you haven't said that's a design element you'll be adding).

[Raezzor]

With 5 volts peak and I was figuring about 120 volts on the plates of the PI that should leave me with plenty of voltage swing at the outputs of the PI even with a relatively large 45k tail, but to be fair I haven't done the maths yet since I was still settling on what tube to use for the PI, and what topology.

At the least I do have the PP info from the RCA sheet on the 6EH5 so I don't have to calculate all those numbers from scratch.

But ya, was hoping there was someone out there who happened to have done something similar with which to base my figures off of. Oh well, time to dig in to my books/papers I guess!

[Raezzor]

I did exactly that in my 4-4-0. 12ax7 drives those little triodes too hard. I was too lazy to modify the circuit, so I swapped that in. Much better.

Cheers! I noticed on the 12AU7 data sheet that even with a single triode they showed using 100k plate resistors and like 2.2k cathode. So I figured a good starting point would be to copy a 12AX7 LTP and go from there! Glad to hear I wasn't far off on that assumption.