MOSFET Vbe multiplier ???

[Andy Le Blanc]

Can a MOSFET be used as an amplified diode or a Vbe multiplier?

I have an application where I desire to provide a DC offset, but with high dc voltages.

Its a question out of my experience, I've be so focused on tubes for too long.

[JMFahey]

Yes, you can.

Basically same as with a bipolar transistor , with the difference that Vbe in a bipolar is around 0.65 to 0.7V and for a regular Mosfet Vgs to just turn it on will be around 3.5 to 3.8V .

[Andy Le Blanc]

Thanks so much, I didn't want to fall victim to assumption.

All the examples I've found were bipolar's, wasn't sure if I was mixing metaphors.

I'll rig the notion and try it out in my application and post the results.

its such a simple circuit idea, I figured that if it works, it could be usable for DIY tube applications.

[JMFahey]

Do you want to build some voltage regulator thingy, or a self biased gain stage?

FWIW some MosFet amplified diodes are used as termal compensation in MosFet output power amps, bolted to the main heatsink.

[Cliff Schecht]

I use high power FET's to do these pretty often. I have a large stash of high power TO-3 FET's laying around so they're what I use. As JM mentioned the threshold voltage is much larger than that of a BJT so you're dealing with a few volts less that you can reduce the B+ by. However this is easily overcome by using a reference with a slightly higher voltage rating to overcome whatever losses you incur in the FET.

[Andy Le Blanc]

I've rigged a mosfet with a 10k pot in the simplest way possible to try the Idea.

I've used this to set the screens voltage in a single ended amp... a champ

My application is to set the difference between the screens and the plate...

It works, even in this crude simple set up, I can swing the screens voltage +2/ - 100 v relative to the plate.

both the pot. and the mosfet have survived the experience. the rail voltage is 400+, the mosfet is a 200v drain to source, I was leary...

the only issue with this unrefined attempt is that I can just slightly hear a switching whine, I'd be very interested in learning more and finding a better result.

It works great for a first stab at it, I can set various combinations of valve bias and screen voltage, with the screen being slightly positive to very negative.

pretty excited to refine it into a push-pull circuit.

[R.G.]

both the pot. and the mosfet have survived the experience. the rail voltage is 400+, the mosfet is a 200v drain to source, I was leary...

If I were you, I'd go order a higher voltage MOSFET, just because there will come a day when the voltage on a 200V one will be exceeded by a transient. Might kill it, might not. But 800V and 1kV MOSFETs are not terribly expensive.

the only issue with this unrefined attempt is that I can just slightly hear a switching whine, I'd be very interested in learning more and finding a better result.

MOSFETs are very funny about source follower setups, particularly when they drive a capacitor directly. They want to oscillate, and often do. The standard practice is to put a 470R to 1K resistor in series with and as close to the gate lead of the MOSFET as possible. This adds some damping and tends to stop the oscillation the same way that a grid-stopper does on power tubes. For hard cases, you might want to put a 10 to 100R in series from the source to the capacitor.

You also should put a 12V zener, cathode to gate, anode to source, on the MOSFET. This protects the delicate gate against voltages over the 20V that will punch through the gate insulation.

pretty excited to refine it into a push-pull circuit.

Why would push-pull be good?

[Andy Le Blanc]

appreciate the reply...

ya, I've already sourced higher voltage devices, the one I scabbed up the circuit
was what I had on hand, out of me bone pile.

I haven't had much time at all with the device, I'll most certainly try the suggested mod, I try the simplest approach first as a proof of concept.
It has to be uber simple, I been doing point to point for so long that it takes
a real effort to remember to not cook a mosfet by trying to measure the voltages with my meter probes.

might as well be the 3 stooges

So... the application is to be able to set the potential between the plate and screen grids of the power tubes at the same time as I'm setting the bias.

I want it to be easy and simple, as both effect the other, dial it in...

I want to explore balancing the voltages over a push-pull pair.
Treating the screens voltage as a controllable voltage offset looks do-able.
I also looks to be a way to eliminate screen grid resistors.
Nothing new there, some ampegs and magnatone went that way.