Improving on frequencycentral's Red Star Drive aka Red Star Caster

[DaveLT]

Having ran through tagboardeffects the past few days (thanks everyone on that site!) I got interested in the red star drive that rick holt aka frequencycentral designed but with a unusual tube even by low voltage pedal standards : a very low voltage submini DIRECT heating pentode submini tube

I will do a updated schematic later but the gist of it is :
1) Ability to use another tube like.... 6AH6? I have tons of the A-series tubes!
2) Cathode bypass caps. Very important. You can even shape the frequency response with cathode bypass capacitors
3) Triode operation rather than pentode operation. The reason are that triodes have a better transfer characteristic leading to a more natural sound... not that you will care when it's distorted right? Well I'll try it both ways of operating a pentode in low voltage vs a triode in low voltage. And there's the lower anode resistance as well...
4) 24V operation! Technically any tube run at 24v as long as it's not some stupid huge triode.. I know, I have a chinese 6N7P that has a common cathode and have 2 triodes in it but the plate resistance is way too high to be of any use in low voltage effects. But it's cheap.
5) Double stages! So it will need 2 valves and will become a pseudo Valvecaster...
6) Use a DC-DC buck converter for the heaters. I am doing the same for my 5670 tube equipped valvecaster.

[DaveLT]

I have had one of those tiny palm sized headphone amps with 6AH6 tubes sticking out the top and even though it's running at 18V it still manages to sound fine although to be fair it's only operating as a buffer so it only needs to feed minimal current to a NE5532 so it's doing literally nothing.

The oddity that I see with it is that Grid 3 is tied to cathode... Usually a triode strapped pentode has Grid 2 and Grid 3 tied together with the anode (plate) however Grid 2 is tied together with the anode here as well. I will find answers to that topology and if it's a reasonable topology I will implement i here  but then again the traditional way of triode strapping a pentode is simply tying the plate with the screen grid as the suppressor is often internally tied to the cathode


EDIT : A cathode tied suppressor grid works fine in preamp tubes because of the general tiny currents involved and both ways are the same result. But the currents involved in a power tube as an argument is void as well... Strapping all grids to anode improves the transfer curves and will work better as a tube at lower voltages.

[DaveLT]

Got around to doing it faster than I expected. 
R2 is cathode resistor and its a very rough estimation following the red star drive just missing the LM317 that's regulating the heaters but I don't imagine that it's an insensible value for it  Of course, the heater is indirectly heated so you can do whatever you want... LM317? Yes. DC-DC buck converter? Yes. LM7805? Yes. But be aware that the tube will not run at optimal temperature!

Right now it's just 1-1 with the red star drive without most of the modifications I outlined apart from no.2 100uF is fat enough that you don't have to worry about cathode degeneration well past 20hz  Maybe the cathode resistor can be replaced with a 10K pot, it just reduces the gain from a dead short or if you want to limit max gain you should leave the 100 ohm resistor in. 100K plate resistor and 1200ohm cathode resistance is about 20 gain so fwiw you would want to lower the cathode resistor if you want huge gain.
Play around with the cathode bypass caps as every configuration is different. as a general rule of thumb, 100nF and smaller (1.8k cathode resistance not sure if it makes any difference) will make it sound like a treble boost so if you like huge treble boost this is for you  It also applies to JFETs!
VR1 is B100K btw and perhaps it is wise to add a 470uF cap on the 24v rail to decouple the power rail.
A 220uF cap (Intentionally small to reduce the load on the power supply preceding as this doesn't really to do much work but deal with the tiny ripple/noise which is possibly too high for it anyway) with a 1 ohm resistor going to ground and a following 470uF cap is plenty for most SMPS (But use it not on the power terminals! Closer to the tubes.). Should you still have issues with hum you can tweak the cap value to 1000uF but 470uF is a low pass filter at 338hz so that seems unlikely to be a issue unless there is a hum issue elsewhere. Or parallel a 0.1uF film/ceramic capacitor across both caps
A 1 ohm resistor shouldn't really drop any voltage at all even with an 50mA load (which is quite unlikely here). probably about 0.05V which is minuscule so 1/4W resistors will work very fine

Also another mod you can do is add a switch before or after the clip pot to engage/disengage the diode clippers and I believe you don't need a perfboard for this circuit, you can p2p everything on the socket itself 

[DaveLT]

Version 2. Starting to look more like a valvecaster than it is a red star drive but then again they aren't really that much different  I call this the Red Star Caster although its more like a StarCaster

[PRR]

Rg2 is "usually" 2X to 10X Rp. (Because Ig2 is smaller than Ip.) You have them equal.

Since you are not looking for "maximum" or "clean", this may be fine.

However a lower Vg2 (bigger Rg2) makes earlier breakup, which is sometimes an issue.

[DaveLT]

Rg2 is "usually" 2X to 10X Rp. (Because Ig2 is smaller than Ip.) You have them equal.

Since you are not looking for "maximum" or "clean", this may be fine.

However a lower Vg2 (bigger Rg2) makes earlier breakup, which is sometimes an issue.

I will create another version i.e these two will go untested while I make a V3 which turns it into triode mode meaning the Rg2 will not be an issue. As it is, it's in pentode mode whos Rp is 0.5Mohms 

[PRR]

I mean your R3 R4.

[DaveLT]

I mean your R3 R4.

No worries, it's running in pentode mode after all. I much prefer triode mode 😂