PI output swing voltage ?

[petemoore]

Would IRF510 mosfets running on ~40 volt supply have enough headroom as source followers to drive output tubes and accomodate the output voltages of an 18W's PI?

[Steve Conner]

Well, the 18W is cathode biased, right? So take the maximum voltage developed across the cathode resistor at full chat, and that's as much as you need to drive the grids positive. Maybe a little more if you want to do the "drive grids positive to get more power" thing, though I was always unsure about trying to do that with tubes that weren't designed for it. EL84s have large cathodes and close-spaced grids, so they'll draw a lot of current and fight your driver all the way, possibly overheating their grids in the process, as the grid already runs very hot due to being so close to the red-hot cathode.

That's the positive excursion dealt with. How far negative, though? The same as the positive excursion might be a good idea. Otherwise you'll have to either clip the PI output or blow your MOSFETS' gates. So you'll need both positive and negative rails for your MOSFET drivers. The 18 watter was such a nice simple amp, too! ;-)

Of course if you're making an 18 watter with 300Bs in the output stage, ignore my comments about positive grid drive and go ahead ;-)

[woodyc]

Just curious, what would be the ballpark output Z of a mosfet in that application?