Tube Power Buffers

[Ben N]

Thinking about a buffer to follow vigilante397's submini F2B clone and provide (a) a bit of power amp mojo and (b) a low-Z output, I came across this: Tube Power Buffers, including this:


Any thoughts about whether this might be the basis for a useful addition and/or how to come up with values to fill in, beyond the grid R and grid stopper?

I have some Russian submini pentodes, although I don't have the specific part number handy--I'll try to check them later tonight.

[merlinb]

You mean build a submini power output stage? You could copy the Muder One...
https://www.diystompboxes.com/smfforum/index.php?topic=73222.520

[Ben N]

Not really. I don't care about being able to drive a speaker, just a line - and possibly get a bit of pentode saturation in the deal. Of course, all that NFB could mean there is no mojo to be had...

[merlinb]

Not really. I don't care about being able to drive a speaker, just a line

Well that's not a power buffer so we're into a different question. If you want some submini pentode distorton then just build a pentode gain stage and buffer it with a simple emitter follower to drive the line. Or with another tube if you have the money.

[Ben N]

That makes sense, but I figure I will have enough current to power one more tube, not two, so I was hoping to find a way to get low impedance out of a pentode--and a transformer seemed like a natural way to do that, especially with zero voltage gain and good noise performance--and maybe even get some transformer saturation in the bargain. But I don't have the design chops to know if this is realistic or how to achieve it. Maybe the ticket is to scale down something like a Champ output section; or your original suggestion, Merlin, of the Murder 1 output.

[Ben N]

Would something like this work to turn a pentode to a line driver? 600:20k Audio Isolation Transformer

[Phoenix]

especially with zero voltage gain and good noise performance--and maybe even get some transformer saturation in the bargain.

Unfortunately those are conflicting goals. To get unity gain necessitates negative feedback (or attenuation before or after gain which degrades noise performance), which results in very different overdrive characteristics than I suspect you are after - power amp "saturation" right? That would normally imply little or no feedback. Large amounts of feedback required for unity gain will result in very "harsh" and "sudden" onset of distortion, as the stage switches from closed-loop to open-loop gain.

I'm with Merlin, pentode gain stage followed by a buffer. If you really wanted you could choke load the stage which would give you a little more "power amp" cred, or a small transformer (reverb transformer?) with appropriate load then attenuated/buffered as appropriate.

[PRR]

>>      just a line
> Well that's not a power buffer

Depends what you call "power". Good line-level can be +20dBm which is 0.1 Watts. This is near the limit of most opamp chips (good studio gear often uses two opamps to drive a main line out). 0.1 Watts is a LOT for small transistors, or high-gain tubes. We do move from 12AX7 or 6SJ5 to some "power" tube. Over about +18dBm the ancient studio gear used 6F6/6V6. (Or for short runs, ran < zero VU so a 6J5/6SJ7's 50mW was adequate.)

> a bit of power amp mojo

There is very little "mojo" if a power amp is working far below maximum.

Take a Fender Champ. It will put 5 Volts into 4 Ohms. It *needs* a <10r load to avoid bursting the OT when clipped. If you have a dummy load and feed a 0.5V "line level" signal out, you are at 1/10th maximum output, have <0.5% distortion, real clean (by guitar standards), no mojo.

FWIW, a Champ-amp OVER-driven to >5V output and then padded-down to go to the main guitar amp is a thing. "Herzog", "American Woman" tone. But that's not what you are asking.

Steal some small "power amp". 100V tubes from US AC/DC kitchen radios made about 1 Watt max. This is 2V in their customary 4-Ohm speaker. Now drop the B+ from 110V to say 48V. (Which conveniently matches the 50V heater on many of these bottles.) It will "have some mojo" at 0.5V out, clip at 1V out.

Alternatively there are beach-radio and sub-mini tubes good for 0.1-0.3 Watts. One I like is 1D8, "power" pentode with volt-amp triode. Very thrifty of filament current. Working near 67V B+ and a Fender Reverb OT it figures like 0.53V on the secondary at clipping, and probably 10% mostly 2nd THD at that point. The pentode needs 6Vpk to get there, but the triode can give gain like 15, so around 0.28V into the bottle for max output. The bottle is *large* but very cool looking.
https://www.radiomuseum.org/images/tubephoto_klein/1d8g_a.jpg
Note: no "glow". The filaments run so cool they can hardly be seen in room-light.
http://vintageradio.me.uk/radconnav/valvetrf/1valvemidget0944pw/page1.jpg
http://flipperhome.x10host.com/BatmanProd.htm

[printer2]

A 6AK5 with a mosfet driven off the cathode.

[Ben N]

Good ideas, all, but not the form factor I'm after. Maybe just leave well enough alone, or, as Merlin suggested, add a triode configured as a cathode follower (or an NE5534) and call it good.

[merlinb]

>>      just a line
> Well that's not a power buffer
Depends what you call "power". Good line-level can be +20dBm which is 0.1 Watts.

No, no, you're right, it's important to be precise. I guess we'd better ask the OP. Do you need your pedal to put 20dBu into 600 ohms? And if so, can you hear me back there in the 1950s?

Besides, a power buffer is one that has low input impedance and consumes the power from a smaller power amp at its input, but delivers even more power at its output. A power-to-power converter, if you will. A regular buffer (even a couple of opamps) isn't really a power buffer since it doesn't consume power at its input, it has high input impedance. It's more of a voltage-to-power converter. ;P

[Ben N]

Now boys, let's all get along.
LOL, no. I don't think I need my pedal to put 20dBu into 600 ohms--just a lo-z output adequate to drive a line--something I could easily get just by tacking on a simple closed loop non-inverting op-amp buffer using something like a 5532.

I was just trying to think of a way to use a submini pentode in there to good effect, as a 0-voltage-gain (because the signal level of the F2B is too high as it is) buffer with mojo, but it seems the game is not really worth the candle, since 0-gain and mojo are incompatible objectives in a single high output-Z device like a pentode. l I do seem to recall somewhere that the classic Marshall preamp cathode follower does compress in an interesting way when hit hard by the preceding common-cathode stages, but that is a very different animal from any kind of pentode buffer I guess. After all, the power buffer article and schematic I linked to are for hifi use.

[PRR]

I have to wonder if you are proposing a combination meat-grinder and tire-wrench.

A "color" stage is level-sensitive. Too low and it is boring. Too high is dirt.

A line-driver is normally expected to take "any" level and pass it out cleanly.

To me, that's two tools. Maybe in one box, but with two level controls to set "color" to taste and "output level" to suit the destination.