Kleuck, "would you believe"..almost minimum... LEDs icon_eek.gif Hmmmm???? how does it sound???
Just as good as with a resistor/capacitor bias, BUT, you need to provide a minimum idle current for the led to really behave as a 0 ohms (say a few ohms) component, my own tests seems to show a 5mA current as a good starting point..And you need to find a suitable bias point on a 1,6 volts line minimum (red led).
The bias point of the regular first stage of a 18 watts, and actually the 12AX7 are not really well suited for that, a 12AT7 or 12AY7 would be easier to use.
But the 12AY7 will provide less gain, and the 12AT7 according to ZP and others a somewhat "arsh" sound.
I think it would be a great idea to test with a 12AT7, as most of the time, when tested, these tubes are biased with the same idle current as are the 12AX7, though they are built to work with a lot more.
I'm not sure the harshness would still be an issue with a 2,5 mA or 5 mA idle current per triode.
I tested this trick with Fets, and my amp will use 6N3P.
Kleuck, "would you believe".
Looks like a joke, but i don't understand
