02-05-2020, 02:49 PM
Dan
I think now that a blocking capacitor is essential. The original circuit applies something around -35V grid bias to the AC044 with the notional centre tap of its cathode at 0V. But your MOSFET is an enhancement mode N channel device which needs its gate to be more positive than its source to start to switch it on. A large negative voltage is therefore the opposite of what you need.
You need the FET's gate to be about 4V more positive than its source in order to start to bias it into the most linear part of its curve. Unfortunately, this characteristic of FETs has huge sample to sample variation, so it's not possible to give a precise number.
The other thing you need to do is ensure that the FET passes roughly the same quiescent current as the valve did i.e. about 50mA with no audio signal applied. Too low and the HT voltage will rise which may upset the rest of the radio's circuitry, too high and you risk damage to the output transformer, rectifier valve, speaker field and mains transformer.
I wouldn't try to set this up with a trimmer - there's every chance that a tiny touch will cause an enormous change in drain current and consequent destruction.
It would be worthwhile breadboarding to find the ideal Vgs to give a reasonably linear Vgs to Id curve. You can choose a suitable bias voltage and source degeneration resistor combination to give an approximation to the valve's characteristics.
Another thing to bear in mind is that, if you approximate the valve's characteristics then the MOSFET will be dissipating as much power as the valve did i.e. about 15W - it will need a heatsink and you won't want this heatsink to be at drain / anode potential.
Good luck, this might be a bit of a project.
John
I think now that a blocking capacitor is essential. The original circuit applies something around -35V grid bias to the AC044 with the notional centre tap of its cathode at 0V. But your MOSFET is an enhancement mode N channel device which needs its gate to be more positive than its source to start to switch it on. A large negative voltage is therefore the opposite of what you need.
You need the FET's gate to be about 4V more positive than its source in order to start to bias it into the most linear part of its curve. Unfortunately, this characteristic of FETs has huge sample to sample variation, so it's not possible to give a precise number.
The other thing you need to do is ensure that the FET passes roughly the same quiescent current as the valve did i.e. about 50mA with no audio signal applied. Too low and the HT voltage will rise which may upset the rest of the radio's circuitry, too high and you risk damage to the output transformer, rectifier valve, speaker field and mains transformer.
I wouldn't try to set this up with a trimmer - there's every chance that a tiny touch will cause an enormous change in drain current and consequent destruction.
It would be worthwhile breadboarding to find the ideal Vgs to give a reasonably linear Vgs to Id curve. You can choose a suitable bias voltage and source degeneration resistor combination to give an approximation to the valve's characteristics.
Another thing to bear in mind is that, if you approximate the valve's characteristics then the MOSFET will be dissipating as much power as the valve did i.e. about 15W - it will need a heatsink and you won't want this heatsink to be at drain / anode potential.
Good luck, this might be a bit of a project.
John
