Hi Anatolij, I thought a lot about FET solutions, but in this circuit, we try to run Vds very low (1.5 .. 2V) to keep the heatsinks cool. The devices work OK, but the capacitance is huge (400pF++ even at little IRF510 level) at low Vds, and changes very rapidly with Vds change (supply ripple).
For best quality DHT heaters, we are trying to make the (ac, small-signal) impedance as large as possible, to keep cathode current from leaking out of the circuit - and from one end of filament to the other. This leakage has a very bad effect on the sound quality, and is one reason that ordinary dc heating made with voltage regulation sounds bad. In fact voltage-controlled dc sounds even worse than ac-heat - no lustre, no sparkle. With voltage regulation, the big capacitor across voltage regulator output (even if 0.1uF) presents direct ac-short-circuit across the filament, and is enough to ruin the quality of DHT-sound.
Anyway, I think you're teasing me about having B+ present before heater can switch ON!!
For best quality DHT heaters, we are trying to make the (ac, small-signal) impedance as large as possible, to keep cathode current from leaking out of the circuit - and from one end of filament to the other. This leakage has a very bad effect on the sound quality, and is one reason that ordinary dc heating made with voltage regulation sounds bad. In fact voltage-controlled dc sounds even worse than ac-heat - no lustre, no sparkle. With voltage regulation, the big capacitor across voltage regulator output (even if 0.1uF) presents direct ac-short-circuit across the filament, and is enough to ruin the quality of DHT-sound.
Anyway, I think you're teasing me about having B+ present before heater can switch ON!!
Last edited:
Anyway, I think you're teasing me about having B+ present before heater can switch ON!!
No, not at all!
If B+ before a stabilizer present instantly, that does not mean it does not rise up gradually on stabilizer's output!
I mean low voltage drop on saturated MOSFETs. More headroom on the same power loss, at least theoretically.
One more alternative, by Alex Zaslavsky from Israel. For my taste it is too complex; Rod's version is more elegant and optimal.
An externally hosted image should be here but it was not working when we last tested it.
