I have been working on the tube section in the ART Tube MP and this is what I have so far. The stock circuit uses one of the triodes as a cathode follower but for an over drive I wanted both triodes to be common cathode gain stages. When operating the 12AX7 in starved plate there is 5uA of control grid current so the grid resistor value needs to be approx 50k ohms or less. Also the plate impedance of the tube is probably approx 120k ohms or better and with the high value of plate resistance used I connected an emitter follower to the plate as a high impedance to low impedance buffer to maintain high gain. I used a KSP44 as the transistor buffer the preamp comes with a MPSA29 which may be an even better choice being a 100 volt Darlington pair. Each tube stage has a gain of approx 40 so it is better then what I was getting using the low voltage 6GM8 tube. In the attached schematic I have 2k emitter resistors with 1uf bypass capacitors to reduce the low frequency gain. I also needed to load the plate of both tubes with 30pF to stop oscillations. If possible the collectors of the emitter followers should be connected to a supply greater then the plate supply so they will not go into saturation before the tube. You can operate the tube with no cathode resistors with a grid resistor of approx 39k. The heater supply in the ART is only 5.7 volts DC with the low plate current you do not need to heat the cathode as much as in normal operation. I can not be sure but the low heater voltage may also reduce the grid current. The input impedance of the tube is low due to the grid resistor value so if you were to make a preamp from scratch you should install a jfet op amp or jfet buffer before the input to the tube for a high input impedance.
Just to add trying the circuit with another amp it is a bit bright so may change the first bypass back to 100uF.
Here is the schematic of the tube section
http://i1020.photobucket.com/albums/...tetwostage.jpg