I am pleased to let you know that after quite a long time and several experiments I
found a circuit configuration that suits beautifully your it103 transformer. I tried
several driver tubes with wide-ranging results but this beats them all.
I am sending the circuit schematic in attachment because I suppose it will interest you, since it is certainly not common.
The tube I use is the noval (almost exact) equivalent to the octal 6EM7. It features a very good voltage amplifier close to a 6SL7 in the first section and a small power triode in the second section (Mu=6; Ri= 925 Ohms; Gm=6500; dissipation 7 Watts). The 6EM7 has an identical first triode, but the second has slightly different (perhaps even better) characteristics (Mu=5.4; Ri=750 Ohms; Gm=7200; dissipation 10 Watts).
In my circuit, the first triode has local reaction, and there's not much to say about it, except that it produces almost zero distortion.
The cathode-follower/IT transformer combination is quite amazing, because contrary to the usual CFs, the DC potential of the cathode equals the polarisation (+20 Volts), and the charge is the input impedance of the following stage, thus very high, which still improves linearity. Furthermore, it works with almost constant current. The output impedance of the stage is even smaller than that which would normally be obtained with a conventional cathode follower (difficult to calculate, but I guess it is inferior to 100 Ohms). Also, it is interesting to note that the best results I have achieved with this circuit didn't require a very high current: only 12 mA. This means that it would probably work as well with your smaller it 102 transformer. The circuit is capable of accepting signals up to 100 volts with practically zero distortion, with an HT between 150 and 200 Volts. Actually, it works in a way very similar to a step-down transformer.
I have used the same circuit on another amplifier with a single power supply, RC coupled (a 18 Kohm resistor in the place of the transformer). It souds very good, but I prefer the version with a separate power supply for the driver stage and the IT transformer, which features much lower intermodulation distortion.
The reason why it sounds so good is obviously the impedance relationships between the different stages and the absence of distortion from the first and second stages. This is interesting too, for in the past I've tried to achieve harmonic cancellation between stages with clearly inferior results.
The way to calculate the CF transformer-coupled stage is quite simple, if you follow the instructions in Langford-Smith's "Radio Designer's Handbook" for the derivation of the CF curves. Then, you'll just have to choose the current you need, which will be set by polarising the stage. You'll have to take account of the transformer's primary DC resistance, of course.
I'm convinced that this is the ultimate driver stage: it's very much reliable than a direct-coupled stage, and you could use a variety of tube combinations - 6SL7/6BXT for instance, or even a power tube as a CF.
As for treble and bass, I believe the frequency extension is quite remarkable. The dynamics and transient response are quite good, too.
I'm sending a photograph in which the driver tubes are still the EC8010s I've used in the first version. It's not a very good one, I'm afraid. Thank you very much for your interest.
Jorge Silva Marques - Belgium