On Wednesday, 25 April 2018 17:21:29 UTC+1, wrote:
NT

"Picking the first random ES design I see 475v dc bias. "

According to RCA's RC30 the peak pulse plate voltage it rated 5,000 volts.. Without reading all the details this is of course with negative voltage on G1, it is the retrace pulse for the yoke and rectified for HV. So a pair stacked SESAPP would yield 2,000 volts peak at best. Bridged, of course 5,000 volts. I'm fairly sure most ESLs need more than that. Without going to transmitter tubes I think that is about the highest voltage to be had except in a HV shub regulator such as a 6BK4. Those are good for 27,000 volts but at 5,000 volts will only pass 750 uA. I think that only works out to about 16 watts. Unless ESLs lose a ton of power in the internal transformer that will not do. The book doesn't give plate dissipation but the tube s designed with the anode away from the glass and they do get red hot and work just fine. It is not impossible though, first of all with that topology, triodes should be easier to deal with. It would be a matter of a bunch of them in parallel which would cost more, but a true audiophile doesn't care.

you already know that some ESes run at under 1kV, and the once mildly popular ES tweeters ran direct off ordinary audio amp valve anodes. But you can build a megavolter if you want.


It doesn't give the maximum heater cathode voltage but I am assuming any design will need separate floating filament supplies. Not too hard with modern SMPS technology. However there is also the issue of frequency response. Since it is designed for HV shunt regulator use I imagine its frequency response is not stellar. I'm not sure now much can be done with feedback when there is that much open loop voltage gain. Plus the capacitive load doesn't help.

"2nd smaller cap in series with the ac HV feed, shorted for full power. "

Makes sense. I red that the power id chiefly determined by the cap value, the transformers are not all that different, except the one I mentioned that actually had an extra tap for lower power.

In light of that, when I replace the cap in this 1992 Litton, what about using a higher value to squeeze a few more watts out of it ? Is that too dangerous or otherwise not feasible ? To me, it is damn hard to have too much power, you just use less time, and since this doesn't ave real power control as far as I know, it just cycles anyway. Of course it might already have the biggest value on practice because it is a fairly large unit and I remember it had alot of power when it was new. We had to get used to it, it wasn't like the elcheapos. I would up the cp value for decent results but if it is going to burn something up because of it I'll leave it as designed.

"Nuke magnetrons use thoriated tungsten bright emitters. "

AHA, so that filament is already burning pretty hot then right ? I see.

Nuke magnetrons run red hot. I wouldn't want to heat them even further. I used to have a nuke that had its magnetron on show, no cover at all between it & the cooking cavity.

The HV cap doesn't affect filament power of course.


NT