ON-CHIP VACUUM MICROTRIODE

Linearity might be a problem. Most of the ones I can find available are designed for power switching. 50 amps at 1200 volts.

I never got that far into it. I just carried a couple of spares. At least they didn't need the bias adjusting after changing them, like power transistors did then. I could have a dead amp back up in seconds if it was just a dead valve.

That could have been useful. Come to that, it could still be useful in keeping the signal in the linear part of the curve, which is getting to be a problem with some stuff.

--
Tciao for Now!

John.

anode

Thank you but I'll take the WE396A if it is all the same to you. For what someone with brains and taste can do with a single agreeable little WE tube per channel, check out my Type 68bis "Minus Zero" 417A single tube micropower single ended amplifier, a glorious one-third watt:

Of course the T68bis "Minus Zero' with that huge one-third watt needs high sensitivity speakers, so if you don't have any old VOT, check out my Fidelio-pattern bicor horns, as in the T91 HWAF3 below: HWAF stands for High Wife Acceptance Factor, as the footprint is only

11in wide by 14in deep --- that's why I keep mine in my studio, for fear my wife likes them so much that she grabs them...

Those horrid Fetrons should not be inflicted even on axe murderers and other guitar players; it is not humane.

Andre Jute Visit Jute on Amps at

"wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review

Yup, but the curves on them are actually pretty linear. The problem is the bias points on them aren't anywhere near where you would want for a tube.

You see them for switching, because that's what there is a demand for. There are very few high power linear devices made out there, so everyone has to make do with semiconductors intended for switching. For the most part (well, except for SCRs and triacs) good linearity can be obtained with careful biasing.

All you need is to round up a venture capitalist....

--scott

--
"C\'est un Nagra.  C\'est suisse, et tres, tres precis."

Don't have the same bias problem with depletion mode DMOS, like thee one from supertex

On the other hand, you can just cascode a typical low voltage jFET into a high voltage transistor and I'm planning on doing just that for the first gain stage in a little guitar amp I'm building.

IMO the bigger problem is that you'd lose the soft overload characteristic of the tube.

There have been electron emitters (cathodes) made using IC fab techniques for at least 10 years or more; in fact since they were active devices, the emission rate was very controllable.

That's the hard part over here.

--
Tciao for Now!

John.

AFAIK both the AC and DC characteristics could not be made to match any of the popular tubes.

Only when presented with a good looking hottie...

Except the *narrow* bandwidth would be a major advantage...

How do you get a characteristic triode curve from a FET?

By tinkering with the shape of the channel, and by building composite devices with multiple channels on one substrate.

I believe the VUDU Tubes guys have a patent on one method.

It should be possible to do with off-the-shelf FETs as well, by using a bunch of parallelled devices at different bias points. Clumsy, though.

But it all depends on how closely you really need to match it. For the IF strip of an communications receiver, you don't need a super careful match and a single off-the-shelf FET would do.

--scott

--
"C\'est un Nagra.  C\'est suisse, et tres, tres precis."

Excellent! I have never seen these before! I just ordered some samples! Many thanks for introducing me!

--scott

--
"C\'est un Nagra.  C\'est suisse, et tres, tres precis."

Get a laser then.

--
Dirk

http://www.transcendence.me.uk/ - Transcendence UK
Remote Viewing classes in London

Plate resistance?

Tim

-- Deep Fryer: A very philosophical monk. Website @

It's easy to add more series resistance. It's hard to remove it.

By "drop one in place" I am assuming some additional resistors, and maybe some shunt caps too if you care about matching high frequency performance, but all on a module that you can stick into a tube socket when a tube is removed.

--scott

--
"C\'est un Nagra.  C\'est suisse, et tres, tres precis."

performance,

My comment refers to the internal feedback that triodes have and pentodes and transistors do not. I know of several circuits (though not commonly used in commercial products) where that would make a huge difference.

I saw a post somewhere below that suggested a manufacturer was onto compensation for this. I suppose a certain doping profile allows the JFET "linear region" to extend to high voltages? This would approximate a triode's characteristic better (roughly constant Rp), but not very linearly (a JFET's "mu" varies widely with Vg).

The 12AX7/ECC83 is a excellent example of highly linear tube characteristics, far in excess of any other device: mu is stone flat over a

200% variation in plate current. 6SL7 is somewhat better still. Transconductance varies as a 3/2 power of plate current, as I recall, which is less "curvey" than FETs or BJTs can claim. (For that matter, 12AX7 has just as much transconductance as 2N3819 is guaranteed to have, which is just silly. Not quite as much bandwidth though.)

Tim

-- Deep Fryer: A very philosophical monk. Website @

Well, if this does work out, we could have D/A converters with on chip analog "vacuum tube" buffers. Like the circuits I threw together using an R2R ladder DAC and triode buffers in CD players, see

Internal feedback? Between what and what?

a

And a 6SN7 will beat both of them, although at a lower mu. In fact, a 6SN7 is a remarkably clean device, operated open-loop.

Peace, Paul

Uh oh, here we go again with the perennial triode feedback debate.

Between the output (plate) and cathode.

Note that it doesn't occur with pentodes because plate voltage is isolated by the screen, which is why they have higher gain.