12AU7 as output power tube

[AmberB]

I have a basic question about using a 12AU7 as a power output tube.  I see them used in small power amps like the Firefly and the Bassman micro, but I haven't been able to find the impedance for the output transformer.  It may be that I just missed it, but none of the schematics that I've looked at give the primary inpedance.
I'm guessing that a 6SN7 would use about the same impedance, with slightly more power, due to the plate ratings of the tube.  I saw on the Bassman micro schematic that they use the Hammond 125B OT, but it has a variety of different impedances...
I'm going to have to write it down so that I can find it in my notes...

[sluckey]

FireFly rev.3 has the impedance right on the schematic.

[PRR]

Single, parallel, push-pull??

Depending on operating point, a power triode can be loaded in 2X to 5X its plate resistance, not real fussy.

12AU7 for some operation point has rp near 6k.

So load in 12k to 30k.

That's not a good value for an audio transformer. That's why 12AU7 is not an audio POWER tube.

[Williamblake]

I have tried the selfsplit 12au7 Rob mentioned with a cheap small 100V transformer used for audio distribution as a reverb driver and really liked it. When the output was connected to a speaker for testing it sounded like an amp to me.

[tubeswell]

One way to think about a 12AU7 reactive load is as the equivalent (in impedance) to a typical resistive load. That’s not making an OT with a primary DC resistance of (say) 22k, but using an OT that delivers 22k AC impedance. On a tube characteristic chart, the load line looks the same but you use a cathode resistance (or bias supply voltage) that delivers the bias voltage needed to keep the tube within a happy 2.75W plate dissipation. If you happen to have a OT that delivers a 22k load plate-to-plate, you want B+ around 245V for Class AB1.


Or if you have a PT with a B+ in mind. you could be guided by ohms law:
Zout = Vp/(Pmax/Vp) for centre-bias Class A operation

E.g. for  300V B+;

300/(2.75/300) = 32,727R so 33k - this would be 33k plate to plate for a push pull amp

For a SE amp with both triodes in parallel, we’d use 2.5W per triode for Pmax, so 5W total

300/(5/300) = 18k

[tubeswell]

Here's some load lines for 245V 22k and 300V 33k

[AmberB]

Thanks for all the information!

[AmberB]

So, looking at some of the other twin triodes that I have that could possibly be used for a firefly amp, I come up with 2 different tube possibilities.  The 12BH7 and the 6FQ7/CG7 tubes are bolt low mu tubes with higher plate current capability than the 12AU7.
The 12BH7 has a max plate voltage of 300 volts, and a max plate current per plate of 3.5 watts, or 7 watts, both plates driven.
The 6FQ7 has a max plate voltage of 330 volts, and a max plate current of 4 watts each, but only 5.7 watts both plates driven.
There's also a couple of single triodes that might be useful, such as the 12B4 that can handle 5.5 watts on the plate with a class A amplifier maximum plate voltage of 550 volts.
Or the 6C4 that is rated at 3.5 watts max on the plate and max plate voltage of 300 volts.

The 12BH7A heaters will pull .6 amp in 6.3 volt parallel operation.  This tube seems to be about the max wattage from a twin triode of at least the twin triodes that I'm aware of and have in my tube stash, although it looks like the 6SN7 may have a slight edge on the 12BH7 if you don't mind using an octal socket.

[HotBluePlates]

... for a firefly amp, I come up with 2 different tube possibilities.  The 12BH7 and the 6FQ7/CG7 ... The 12BH7 has a max ... per plate of 3.5 watts, or 7 watts, both plates driven.
The 6FQ7 has a max ... of 4 watts each, but only 5.7 watts both plates driven. ...

You may want to think about whether it's desirable to make the Firefly louder: there's already many ways to build a 4+ watt amp.

A different way to think of the problem is, "How loud do I want the amp to be?".

Imperfect as it may be, the guys on That Pedal Show have a dB-meter in the right corner of the room.  The response is slow, it's off-axis from the speaker(s) and more than 1-meter away, but it does provide an indication of relative loudness.  30-40 watt amps just starting to compress & distort in their videos tend to register ~105-110dB, and would read higher if the meter were moved to the usual measuring position.  60-70dB SPL is on the quiet(er) end of things, depending on your living situation & preferred hour to practice.  See the graphic below.

The chosen speaker has as much or more impact than amp-power, and sensitivity ratings (dB SPL at 1w, measured 1m away) tend to vary from ~93dB to 100dB (or slightly more).

           dB SPL when Amp-Power is Known:
           dB SPL = 10 log (Amp Power / 1 watt) + Speaker Sensitivity for 1 watt

           Amp-Power needed when dB SPL & Speaker Sensitivity are Known:
           Amp Power = 10^{[(dB SPL - Speaker Sensitivity)/10]}

If I want my amp's power tube(s) to distort at 75dB SPL and have a 98dB speaker, then

           Amp Power = 10^{[(dB SPL - Speaker Sensitivity)/10]}
           Amp Power = 10^{[(75dB - 98dB)/10]} = 10^[-2.3] = 5 milliwatts (!)

           CHECK: dB SPL = 10 log (0.005 watts) + 98dB = -23 + 98dB = 75dB SPL

• Pipsqueak amps are louder than you'd expect with sensitive speakers.
• Amp-power reduction has less impact than you'd expect.
• Insensitive speakers for guitar aren't easy/cheap to find.

https://georgeantonios.com/wp-content/uploads/2017/05/volumelevels.jpg