Simple Tube Boost Circuit

[Jim Hagerman]

I should have done this years ago! Finally took accurate measurements of a 12AU7 operating both starved plate and heater. Both samples are done with heater at 9V (instead of 12.6V). First plot here is a very "hot" high gain sample:



I hand select tubes for my pedals, not all work as expected under these severe conditions. The "hot" ones go into my VALVE pedal. This particular tube pushes 24dB of gain in a VALVE! Typical gain tubes go into TubeKlon and other pedals that run at 15V. Here is one of those samples:



So yeah, starved operation really does work! You just have to adjust circuit parameters to accommodate. Not shown is that grid current is much higher than normal (relative to plate current). This is actual measured data!

[PRR]

Thanks for data.

To really wet your pants, set up a LOG current scale and go below 1uA. (Here you find which bottles have gone a trace gassy.) The lines look more linear (an optical illusion) and "cutoff" seems to vanish (pushed far below 1uA).

I have no experience but I suspect that undercooking the cathode is just shifting all voltages higher. Or adding virtual resistance in the cathode circuit.

[Jim Hagerman]

I assembled everything into a short article...

Starved Pedal Article

[printer2]

It is still early, what is the bottom scale?

[Jim Hagerman]

It is still early, what is the bottom scale?

Plate voltage. Just like every other set of triode curves out there.

[merlinb]

Not shown is that grid current is much higher than normal.

This is the fly in the ointment of starved operation, the input impedance of the grid is typically less than 20k even when the valve bias is slightly negative.

[printer2]

It is still early, what is the bottom scale?

Plate voltage. Just like every other set of triode curves out there.

First day of the year and my brain could not relate to the numbers (at least the current scale is listed). Today it makes perfect sense.

[amptramp]

Maybe the 12U7 would be a good substitute for a 12AU7 in this application.  There were about 30 tubes with 12.6 volt heaters that ran on 12.6 volt plate supplies that were made for car radios in the late 1950's.  The output of these radios was a single-ended germanium power transistor but the rest of the circuitry was low-voltage tube and this eliminated the vibrator and vibrator transformer that had previously been standard in car radios.

[merlinb]

Maybe the 12U7 would be a good substitute for a 12AU7 in this application.

That tube is in demand precisely for tube pedals, which makes it cost even more than a 12AU7. Also they pretty much don't exist outside the US

[merlinb]

I hand select tubes for my pedals, not all work as expected under these severe conditions. The "hot" ones go into my VALVE pedal. This particular tube pushes 24dB of gain in a VALVE! Typical gain tubes go into TubeKlon and other pedals that run at 15V.

Does your 'half watt' run at 15V or does it use a voltage booster inside?

[Jim Hagerman]

Does your 'half watt' run at 15V or does it use a voltage booster inside?

Most of my pedals operate not starved, the HALF WATT runs single-ended power stage at 200V from internal booster. It also consumes 13 watts from 15V supply just to get that clean 1/2 watt output into 8 ohms. It also does 1.5 watts dirty.

[PRR]

Does your 'half watt' run at 15V or does it

I assume you know Jim's site. The HalfWatt manual is on there and has a PDF-fuzzed schematic showing "200V", though not how it's made.