The biggest issue with this circuit is the grid will never reach 0V and maximum power. Inserting a low mu driver between the PNP and output tube complicates circuit but gives full potential power. (I am a flea watt fanatic and will happy live with simple circuits sacrificing maximum potential, but if...)
Btw I have standard LTspice lib so the PNP shown here is just one found there, not the one to use for sure...
Of and I swapped 300B for 6AS7 b/c I have many of those... To get max symmetric output teh 6AS7 is partially auto biased, which in the case of those tubes is usually a good idea anyways.
If Vp of driver tube is too high, a voltage dropping resistor and cap in sereis with the plate could be used.
Btw I have standard LTspice lib so the PNP shown here is just one found there, not the one to use for sure...
Of and I swapped 300B for 6AS7 b/c I have many of those... To get max symmetric output teh 6AS7 is partially auto biased, which in the case of those tubes is usually a good idea anyways.
If Vp of driver tube is too high, a voltage dropping resistor and cap in sereis with the plate could be used.
Another problem with the circuit: The DC bias to the 300B is not stable with time or temperature.
The small changes in Vbe are multiplied by the DC gain, so that normal warm-up will shift the bias voltage by 5V or more. That's a big shift in anode current for a 300B.
It can be greatly improved by adopting a stable biasing scheme for the base of the transistor, shown in the circuit below. Now the bias voltage is a large portion of the supply, and is less sensitive to drifts; meanwhile there is a separate (ac-coupled) resistor to set the gain (and feedback ratio).
The bias changes have the additional benefit of greatly increasing the input impedance, allowing lighter load for the source, and smaller capacitors.
Driving to 0V can be accommodated by adding a few volts of positive supply.
This is not just a matter of achieving full power - we also need to prevent the driver from clipping at the same time as the output.
If the driver has a little extra margin, the audibility of gentle clipping in a well-designed 300B-SE is remarkably low.
Example of the circuit, converted to NPN KSC3503 for the reasons indicated; LTS thinks the 124Vpp THD is ca. 0.4% with a good cadence of harmonics. The circuit will work with unregulated supplies.
It's not going to rival my Shunt Cascode driver, but it is easy and fun to build & try.
The small changes in Vbe are multiplied by the DC gain, so that normal warm-up will shift the bias voltage by 5V or more. That's a big shift in anode current for a 300B.
It can be greatly improved by adopting a stable biasing scheme for the base of the transistor, shown in the circuit below. Now the bias voltage is a large portion of the supply, and is less sensitive to drifts; meanwhile there is a separate (ac-coupled) resistor to set the gain (and feedback ratio).
The bias changes have the additional benefit of greatly increasing the input impedance, allowing lighter load for the source, and smaller capacitors.
Driving to 0V can be accommodated by adding a few volts of positive supply.
This is not just a matter of achieving full power - we also need to prevent the driver from clipping at the same time as the output.
If the driver has a little extra margin, the audibility of gentle clipping in a well-designed 300B-SE is remarkably low.
Example of the circuit, converted to NPN KSC3503 for the reasons indicated; LTS thinks the 124Vpp THD is ca. 0.4% with a good cadence of harmonics. The circuit will work with unregulated supplies.
It's not going to rival my Shunt Cascode driver, but it is easy and fun to build & try.
Yes well how bout using a 6922 triode for Q1? ...
LOL now we suddenly depart from the beauty of the simplicity in the first circuit. My fault too in trying to extract 10% more power in my suggestion.
The bias shift is benign, at least in the temp range 10-50C.
I have most of my tube amps DC coupled via mirrors or folded cascodes. Never ever had problems with bias drifting... and I never ever use fuses or other tube-protection.
LOL now we suddenly depart from the beauty of the simplicity in the first circuit. My fault too in trying to extract 10% more power in my suggestion.
The bias shift is benign, at least in the temp range 10-50C.
I have most of my tube amps DC coupled via mirrors or folded cascodes. Never ever had problems with bias drifting... and I never ever use fuses or other tube-protection.
Fixing the lack of positive swing is essential.
Not so much to get full power - but to prevent the driver hard-clipping before the power stage clips. If it does that, the sound is the same as a 7W transistor amp: as sweet as breaking glass, with a real speaker, and any kind of dynamic music.
It's not benign in the original circuit: for this temperature range, Vbe rises by 2.2mV/ °C; now multiply by 40. Then multiply by the DC gain (60) = 5.3V. In the standard 300B curves, that takes it from 350V/60mA to 80mA+.
This is without mentioning that the initial predictability of the bias voltage would be nowhere near accurate enough to get -74V without a trimmer pot.
The NPN circuit limits this rise to ca. 5-7mA over 40 °C.
The NPN circuit improves the input impedance hugely too...
Adding two parts and a 25V unregulated supply is good value for fixing all these faults - some circuits are TOO simple!
Dang... Valid points and your suggestion is much better, but then it's a far cry from the original simple circuit.
I always assume it's expected to have to tweak and tune circuits like the first post. Of course the PNP will need adjustment to get bias correct in a real life circuit, with a pot or different resistor values.... I would also do some mods to get the temp drift down. The gain is high enough so some local feedback from collector to base to minimize tempco. And having just a little resistance in the cathode circuit to give a little auto bias is a good idea, but giving even less amplitude before PNP clips.
I still think it's a fun circuit. I wouldnt use this on a 300B, but will have no problems breadboarding a simple amp using a 6W6 or 6AS7 for a low voltage(+-250V-ish) fleawatter. It's a nice weekend project just for fun. I realize weekends are not all free and silly BS projects will seem like a waste of a good weekend. But then again doing silly things can be liberating.
I always assume it's expected to have to tweak and tune circuits like the first post. Of course the PNP will need adjustment to get bias correct in a real life circuit, with a pot or different resistor values.... I would also do some mods to get the temp drift down. The gain is high enough so some local feedback from collector to base to minimize tempco. And having just a little resistance in the cathode circuit to give a little auto bias is a good idea, but giving even less amplitude before PNP clips.
I still think it's a fun circuit. I wouldnt use this on a 300B, but will have no problems breadboarding a simple amp using a 6W6 or 6AS7 for a low voltage(+-250V-ish) fleawatter. It's a nice weekend project just for fun. I realize weekends are not all free and silly BS projects will seem like a waste of a good weekend. But then again doing silly things can be liberating.