please help me finding a mosfet

[JC@]

I would need to find a reference for a n-channel depletion mode mosfet.
50 - 75 watt or so.
I pretend to lower the B+ of an amp using the amplified zener approach. the amp is not mine and 50w zenners are way too expensive.

[R.G.]

I would need to find a reference for a n-channel depletion mode mosfet. 50 - 75 watt or so.
I pretend to lower the B+ of an amp using the amplified zener approach. the amp is not mine and 50w zenners are way too expensive.

I think you mean "enhancement mode". Depletion mode MOSFETs exist, but unless the market has changed a lot, I don't know of any that get up to 50-75W.

But the good news is, you don't want a depletion mode device for an amplified zener anyway. You want an enhancement mode device. The run of the mill MOSFETs from IRF and others will do the job nicely. You should be able to get a suitable MOSFET in the USA for $2.00. Other places may be higher.

You'll want a highish voltage zener just to eliminate device degradation on power on/off transients. You oughta be able to find a suitable N-channel device easily enough. Be sure to put the zener current resistor from gate to source, zener from drain to gate, and protection zener (12V is fine) from gate to source. A 100 ohm resistor directly in series with the gate is a good idea to stop UHF oscillation.

[loudthud]

The nice thing about using an N-channel Mosfet is that the Drain is grounded so you can just bolt the case to the chassis. Some people have reported blowing those expensive ($25USD) zeners in amps with solid state rectification.

It might be a good idea to connect a large, low-ESR, high Ripple current cap around the whole thing. Maybe 1000uF @50V. What does everybody think about that?

[JC@]

Thank you R.G.
>>But the good news is, you don't want a depletion mode device for an amplified zener anyway. You want an enhancement mode device. The run of the mill MOSFETs from IRF and others will do the job nicely.

Ah, this is good news then.

>>and protection zener (12V is fine) from gate to source.

cathode to gate, anode to source?

[loudthud]

It's going take a little finesse to make such a circuit withstand the inrush current when the B+ is turned on. You don't want a big current through the voltage drop determining zener AND the protection zener which could be enough to destroy them both. But, at the same time it takes some current to get the gate of the Mosfet turned on. Throw in the unknowns the like size of the filter caps, the transformer's resistance and it's saturation characteristic, and it is going to be a challenge to come with a circuit that won't fail. We're probably going to kill some Mosfet$ before a solution is found.

That's what got me thinking about the big cap across the whole thing. You would like to size the cap such that the inrush current divides properly between the zener and the filter cap(s).

Is this thing that much different than O'Connor's power scaling?

[JC@]

The nice thing about using an N-channel Mosfet is that the Drain is grounded so you can just bolt the case to the chassis.

Well it works!
Now the amp's B+ is just a bit over 450 volts dc.
I had my doubts about putting some heat conductive grease or not. I did, so to be sure about grounding I throwed a little cable from drain to the nearest ground point in the chassis.

I feel good with my work

[JC@]

It's going take a little finesse to make such a circuit withstand the inrush current when the B+ is turned on. You don't want a big current through the voltage drop determining zener AND the protection zener which could be enough to destroy them both. But, at the same time it takes some current to get the gate of the Mosfet turned on. Throw in the unknowns the like size of the filter caps, the transformer's resistance and it's saturation characteristic, and it is going to be a challenge to come with a circuit that won't fail. We're probably going to kill some Mosfet$ before a solution is found.

That's what got me thinking about the big cap across the whole thing. You would like to size the cap such that the inrush current divides properly between the zener and the filter cap(s).

Is this thing that much different than O'Connor's power scaling?

it could very well be the luck of the novice, but it is working.
I used an IRFZ30 (50V 30A 75W) I had recycled from some other board.
I turned the amp on a few times and all seems ok.

Oh, and I found this on a search, nothing new to you all, but it has been a good read for me.
http://www.semtech.com/pc/downloadDocument.do?id=525