Jfet/Mosfet cascode with OPT

[Fuling]

Hi

I recently bought a bag of 50 J310 Jfets on Ebay without any direct plans for their usage.
Some of them will probably end up in various linestage prototypes, but I have this other idea that I´m been contemplating for a while: a cascoded Jfet "power" amp with output transformers.

Jfets usually goes belly-up when the drain-source voltage exceeds 30V or so, but if we make a cascode with a high voltage mosfet on top of the jfet(s) they should be safe to use at (much) higher voltages.

Next issue is the low drain currents allowed by the jfets.
J310 seems to be able to swing about +-10mA which wouldn´t even tickle an 8 ohm loudspeaker.
Ok, I could buy a couple of hundred J310s and wire them all in // but I´m not sure it would be worth the effort.
The tubes guys have faced the same problem since the dawn of time (or at least for about 100 years), we have all the voltage swing we need but almost no current.
A transformer between the output device(s) and the load is the obvious solution.
I am well aware that output transformers are considered the worst bottlenecks in most tube designs and I completely agree with that but hey, I´m in the mood for some experimenting here...

Let´s say we take ...5x J310 and wire them up in // with adequate source resistors and gate stoppers in a common
source configuration.
Let´s set the total drain current to 55mA by chosing the right source resistors and then we put something like an IRF840 on top of the whole thing to deal with the high voltage and to dissipate most of the power. 10V across the jfets would probably be just about right.

Assuming that each jfet can swing from 2mA to 22mA (according to the datasheet) we can get a total current swing of 100mA p-p.
With a transformer forming a 2,5k load on the drain this current swing will render a voltage swing of 250V p-p giving an output power of... just above 3W!!!
About the same output power as the legendary 2A3 triode, but at much higher efficiency and without all the troubles related to the direct heated filament...

I guess the supply voltage would end up somewhere in the 150V region, easily available from an isolation transformer...

And, uh, one more thing: The complete absence of (voltage) feedback would leave us with a very low damping factor...

[Zen Mod]

well ........ be my guest ;
I have da toobz for that .

even my preamp is more powerful ..........


http://www.diyaudio.com/forums/showt...91#post1364291

[Fuling]

Oh, I should have put a " " after those " "s...

As I said, I´m just in the mood for some experiments. If this thing would work as expected (or anywhere near it) it would still not be a very versatile amp.
Perhaps it would sound good driving high eff fullrange drivers in open baffles but regular speakers with crossovers and tuned boxes would probably get very confused...

[wrenchone]

Fuling - Funny you should start this thread, but I have a cascode simlar to what you've described using J110 FETs. It's all built up and waiting for me to finsih it one of these days.

[Steve Eddy]

Why not flip it around?

Use step-up trannies up front for your voltage gain and then an all N type push-pull follower for the output, using Lovoltech power JFETs cascoded by MOSFETs with a center-tapped choke load?

se

[Fuling]

It's all built up and waiting for me to finsih it one of these days.

Let me know how it turns out!

step-up trannies up front

I like your output stage idea, but for gain I think I would prefer triodes rather than step-up transformers. Especially as my linestage already has step-down output transformers...

[Steve Eddy]

Originally posted by Fuling
I like your output stage idea, but for gain I think I would prefer triodes rather than step-up transformers.

Perhaps. Though I think you should give them a try sometime. As long as you avoid overloading them at the lowest frequencies, I think a good quality step-up transformer offers the cleanest and sweetest voltage gain I've ever heard.

Especially as my linestage already has step-down output transformers...

Well that's because the active devices in your linestage are being used as voltage gain devices. For the amplifier I'm suggesting flipping that around and using transformers for signal gain and the active devices for impedance transformation in the form of voltage followers.

se

[Fuling]

Didn´t Susan Parker use ordinary mains toroids as stepups in some version of her Zeus amp?
If it is managable to get full frequency response through such transformers it would be fun to try a pair with adequate input and output buffers.

[GRollins]

Might I suggest a preamp instead?
They would also do well in crossovers.

Grey

[Steve Eddy]

Originally posted by Fuling
Didn´t Susan Parker use ordinary mains toroids as stepups in some version of her Zeus amp?

Yes.

If it is managable to get full frequency response through such transformers it would be fun to try a pair with adequate input and output buffers.

Depending how much you're looking to step up, you may not need an input buffer.

How much gain are you actually needing for the power amplifier?

se