My screen supply is the old fashioned approach with a 5R4 rectifier. Choke input
with negative lead filtering, and a 6X5 to rectify the ripple across the choke,
thus providing bias.
Bias supply regulator is a shunt regulator, using an MJE340, and a couple of NPN
transistors of some sort as an long tailed pair amplifier driving the MJE340.
Screen regulator is a 6L6G driven by a 6SJ7, with a VR150 reference.
Works fine for 4-400s or 4-250's. They don't draw negative screen current, at
least not that I've ever seen. I suspect they were before the days of aligned
grids in big tubes - the design is older than I (and I suspect a fair number of
us here) am. For AB1 operation, you need about 600 volts on the screens; I find
AB2 operation with shunt regulated bias and series regulated acreen and cathode
resistors for negative feedback gives significantly less distortion. (Despite
what Rich says - the HP spectrum analyser is reasonably truthful if used with
care). For that, you need about 300 on the screen and around 60 volts of bias.
I have a 6 volt 15Amp switcher. It's very quiet electrically, and the plan is
to use it for the 4CX1000 amp. However, it definitely won't start up on cold
heaters, and of course, the thermal time constant of the heater cathode assembly
makes the electrolytic dodge a bit difficult. I'm going to use some PNP
germanium pass transistors in a constant current source for the filament.
Germanium because I have them, the beta in the ones I have doesn't collapse as
quickly at high collector current as 2N3055s, and the lower Vce sat means that
with the switcher delivering 6.8 volts, there's still enough headroom to
guarantee constant current.
I note that my 4-400 and the 4-250s glow quite brightly at rated dissipation.
Following Tom's comments, I presume that helps to keep the vacuum clean.
Still, real radios glow in the dark..................even if it's just the over
dissipation on the plates!
FAQ on WWW: http://www.contesting.com/FAQ/amps
Administrative requests: amps-REQUEST@contesting.com