ON-CHIP VACUUM MICROTRIODE

Folks have actually been building MEMS tubes for a while now. You build them small and you can get away with cold cathode since the interlectrode spacing is so small. And you can run them at low voltages for the same reason. But you don't get a whole lot of current.

My guess is that these things will turn out to be most useful at RF, and that the next step is for someone to develop a MEMS magnetron.

--scott

The problem with goodies like this is the fact that sputtering will take place. As the electron winds moves from cathode to anode, it brings along a little bit of the cathode. The transfer will result in degradation. In vaccum tubes, it was not a real problem. In the nanoworld it may be. Reliabily studies need to be done.

Al

Which reminds me of a direct plug in replacement for some thermionic valves made using FET technology a few years ago.

Whatever happened to them?

Nanotubes are erratic and noisy electron emitters, and tend to wear out fast. This sounds like another parlor-game "breakthrough" that are too common these days.

When there's an actual application, let us know.

John

I still have kits to build several kinds, intended for amateur radio gear with now-unavailable tubes.

I keep saying I'm going to design something like this for TV sweep tubes too, but that's a non-trivial problem.

--scott

Good luck with that.

The ones I had in mind were something like an ECC83 replacement, only needing (in some circuits) the heater pins connecting with a resistor.

It would have been handy at the time to be able to replace an EL84, as I had a few pairs of those in output stages. I guess times have moved on past the need for them, though.

That was the Fetron.

Developed for (by?) the phone company, to replace vacuum tubes in equipment. They weren't robust when it came to HV spikes (lightning strikes) like tubes were, so Ma Bell shitcanned 'em. They were offered by one of the major ham radio companies (Collins?) for a short time, and also turned up in early Mesa-Boogie guitar amps, where they were feared and hated by all. ;-) Mesa, fortunately, included a switch which allowed the guitarist to replace the Fetron with a 12AX7. I still run into a Fetron every couple of years or so.

Here's a pic:

Lord Valve Geetah Amp Fixer

Are these naturally rad-hard like the glass ones?

LV

totojepast wrote:

Such a magnetron would probably operate at close to optical wavelengths. Might as well get an LED

They're the ones :-)

Shame they didn't work out. I vaguely remember more than one sort being available on this side of the Atlantic.

On a sunny day (Tue, 11 Dec 2007 14:59:14 +0000) it happened John Williamson wrote in :

In Europe the lumince amplifier tube PL802:

John Larkin hath wroth:

The audiophile market will love them. Tube amps and pre-amps have been hot sellers in the high end audio market.

For example, this CD player has 3ea 12AX7's and one 12AU7

and sells for about $1500. There are plenty other audio products on that site with tubes inside (search by keyword).

If you could still get high voltage JFETs, you could pretty much drop one in place of an ECC83 without any problem.

But you can't. If you want that kind of voltage and that kind of current, you are limited to MOSFETs. You can drop one in place of an ECC83 with some additional bias resistors added, but it WILL take some additional tinkering to set the bias and that may differ somewhat from circuit to circuit.

Thing is, if you replace a power tube with a solid state tube of similar transconductance, you still have the same heat issues that you did with the tube. And if you use a higher transconductance device, you're going to need to change the circuit a little.

Things like the EL84 and ECC83 work just fine and don't cost all that much, so the demand to replace them with solid state equivalents just isn't there. You could do it, if you really wanted to.

--scott

I recall integrated "micro" vacuum "tubes" being touted 30 years and more ago for use in instruments dropped into wells -- deep wells where the temperature gets too hot for most solid state devices, and can be used to excite emission from the tubes' cathodes.

Cheers, Tom

Accepted, which is why the replacements had a PCB with a number of components in them. A picture & schematic I've been pointed at on the web shows 2 FETs, a couple of resistors & a capacitor on the board. That's per triode, the picture was of a single triode, while the ECC83 was, IIRC, a double triode with a single heater.

They got pretty close, though, or so I heard.

It was more of a vibration sensitivity problem I had, to be honest.

I don't need to, now. :-)

Solid state's now better sounding (IMHO) than valve technology. The problem I had with the EL84 amps was the fragility of a valve in the back of a van on the way to a gig as well as the weight of the output transformer. The solid state version might have got round the fragility, but not the weight. Now I just need a 1U sized rackmount power amp that weighs a fraction of the amount which does an equivalent job.

As I said, technology's moved on...

I suppose in the sound field, the new triodes could be used as a compact valve replacement, but how close to the classic valve transconductance & other specs could they be made? How many volts can you use with a gap of

As has been said elsewhere in this thread, they seem to be a solution looking for a problem. A bit like lasers were to start with, I think. Uses will no doubt appear that nobody's even thinking of yet.

Nuvistor, 13CW4, perhaps?

You can. Back then, you could get higher voltage JFETs from Hitachi, but those have all been discontinued. Getting discrete semiconductors today is getting to be a lot harder than getting tubes.

Hmm. IGBTs would be interesting. Take some additional biasing too.

The EL84 was never really intended for high vibration environments, but the loctal tubes that were never really made it in the audio world.

Actually VUDU Tube was trying to build some high voltage JFETs whose curve carefully matched that of some standard tubes, even up in the clipping region. They did one production run of them and they were very impressive, but they could never get the financial backing to get production fab done.

--scott

Red Rhodes claimed he'd done that and demoed it but his employers then (a certain tube remarketing company founded by a guy with zero tech knowledge named after a tree) would have nothing to do with it. He refused to say much about it but said it consisted of one or two FETs and a horizontal output transistor.

anode

Similar devices worked great in Collins S-line equipment, but each was custom designed for a particular circuit application. You also had to replace all of the tubes called for replacement by the kit with the solid state plug-ins, and get them in the right places, because the series-parallel heater wiring in these boxes would otherwise not work. Another point was that the S-Line ran small signal tubes at 150 volts. The solid state plugins would only tolerate about 200 volts so would die in other equipment.

Direct substitution of a N-channel FET for a small signal triode is the only nontrivial "transistorization" of existing circuits.