diamond coatings as electron source in vacuum tubes

[hoola]

hello, there have been built by MIT researchers and others, vacuum tubes without using heaters to activate a cathode of thorium coated metal to produce current. A seemingly simple crystalline coating of pure diamond on a metal substrate introduced into a standard amplifying tube seems to work and create a useable current flow with which to produce a "filamentless" tube, without the heat and power consumption of a filament, longer proposed lifespan, and of course a vast improvement in overall amplifiier efficiency....I read about this in S.A. magazine's October 1992 issue....would it be possible to get a diamond coated piece of metal, say a 1/8" copper rod, about an inch long, and put it into an evacuated glass vessel surrounded by a cylindrical piece of uncoated metal (a plate in tube-terms) and then see if current will flow from the coated rod (cathode) through the vacuum to the plate using a normal cathode to plate potential in the 300 volt range? This is what the article claims to have been done.....has anyone out there tried this? Is anyone interested in discussing the viability of a project to see if the claim can be verified.?...thanks, edd

[EdEarl]

Electron tubes generally have more noise and distortion than transistors. What purpose is there for electron tubes without filaments?

[hoola]

well. that is dependent upon the particular circuits involved. Many audiophiles prefer the tube sound in their home stereos, and are willing to build their own amps customized to their own needs....the amount of distortion/noise is essentially the same for either type circuit, to even critical examination with instruments. Tubes tend towards 2nd order harmonic distortions, and solid state with 3rd order harmonics, In either case, noise and distortions can held down to inaudible levels. I am into circlotron type amps, built with tubes, that do not need output transformers that regular tube amps require, This allow the tubes to be directly connected to the speakers, giving less restriction to the output power bandwidth. The frequency response and other factors are improved by this technique. One source of noise in tube circuits is hum, introduced by the AC voltage delivered to the heaters to activate the cathodes. Without these heaters, that source of noise is gone, but the main advantage is with the increased efficiency of tubes rivaling that of transistors.....and of course, the better stability, longer life, and instant on features that transistors have. the SA article goes on to say that the diamond cathoded tubes could outperform standard tubes and even solid state devices....a pretty tall order. Since they have never been developed to the point of being on the market, who knows....there may be some others who would like to do a fun project of building a few prototype diamond cathoded tubes just to see if the supposed effect works...thanks for your interest....edd

[Enthalpy]

You can supply the filaments with DC to remove the hum.

 

Having not compared with my own ears, but knowing how many crooks exploit the audiophiles, I won't risk an opinion about tube amplifiers nor their distortion.

 

Nanodiamonds are still a research topic for electron emission. To my limited understanding, their bulk properties are nothing special; diamond has the advantage of sharp edges and peaks that concentrate the field to favour emission, and nanosize puts more peaks per area unit. Tungsten carbide or silicon carbide might be as good, but we happen to have a working process for nanodiamond.

 

Don't expect a current density as good as a hot filament of thoria-coated tungsten. The present research goal is to make plasma screens, and people try hard to obtain locally the current needed by one pixel, while a hot filament provided the current for a whole screen.

 

For more current, you might replace the rod by a wound filament - as is done with hot tungsten. Apart the higher resistance, it also concentrates the electric field while keeping some area. Put the nanodiamonds on the thin wound wire.

 

The other limit of cold emission, as yet unsolved, is the sensitivity to absorption of residual gas. Field emitters in hard vacuum last for a quarter hour, then they absorb one gas molecule, and it's over. That's the very reason why electron microscopes use warm LaB₆ cathodes. Their current isn't as nicely concentrated as from field emitters, but far better than hot tungsten, and they last long.

 

An other use of emissive cathodes is to build micrometre-size vacuum tubes. Quick tubes have been demonstrated, integrated by hundreds on a single substrate, but here a cold cathode seems necessary, so the lifespan isn't solved.

 

Marc Schaefer, aka Enthalpy

Edited December 6, 2013 by Enthalpy

[EdEarl]

Most hum was removed when filaments and cathodes were made separate elements in tubes. IIRC electrons hitting the anode cause noise, especially when amplifying low level signals. I know that some guitar players prefer tube amplifiers because of the sound, but they tend to use distortion for special effects.

[hoola]

enthalpy...you are correct in your asessment that fraud is rampant in the audiophile arena...to the point of embarassment to me personally as I have been in the business my whole life....but, what enterprise in commerce/entertainment is free from such problems....caveat emptor...PT barnum is everywhere....If the article in scientific american is correct, the emission of a diamond coated cathode is limited by the current carrying capacity of the metal substrate......and that they would outperform regular tubes and solid state devices. Who knows if this is true, I presume SA did their homework and the article represents real research from MIT....it should be easily gotten online. I haven't tried that as I have my original Oct 1992 issue always with me. They did not use nano damonds, they applied a pure diamond crystalline coating to a cylindrical rod, I believe and got the results they claim. I have been in touch with a company in england that will coat anything you want with an amorphous diamond coating....but I don't want to do anything other than to encourage the tube manufacturers to check out the amorphous diamond coating to see if it will produce the results of the crystalline form, as the amorphous coating is a relatively inexpensive and simple procedure compared to going crystalline. There is a problem with the lack of heat from the filament I can forsee...the "getter" is a barium coated ring that absorbs stray outgassings and it's required in the construction and long life of a tube. During construction, the getter can be activated to purge the interior of the tube to a hard vacuum with microwave inductive heating....however, gasses tend to be emitted by the metal surfaces within the tubes a little at a time. The normally hot filament is used not only to activate the cathode, but keep the getter sequestering gasses over tube lifetime. Some of the gasses, I would conjecture, come from the thorium coated cathode and filament wires...so this source of potential gas would be gone. There is still the plate, grids and internal supports to deal with, so some way would need to be developed to allow the getter to function in at least a minor role as before. The plate of a tube runs hot, and some of the heat is not from the filament, but from plate current. In a diamond cathoded tube, a primary getter would be positioned as normal, and used as normal in the initial purge. A second getter could in theory be affixed to the outside surface of the plate of a tube and some of the normal gettering would take place. In a high current output tubes like a EL34, 6L6, 6550 (of which I have 12 in my DIY stereo circlotron amp), the plates get hot enough to activate a getter if the getter was placed on the plate exterior instead of near the side-wall or on top above the active assembly, I would be willing to venture, owing to simple plate current. A system could be developed to accomplish this goal, at least with getting around the getter problem. A small tube, such as 12AX7, 7025, etc, used in the voltage gain circuitry of a typical amp would not generate enough heat to use the plate as getter activator...so they would remain as normal. They use a small amount of heater current anyway (300ma), so the diamond cathode tube idea only make technological sense if high current output tubes are made filamentless, as they draw anywhere from about 1 amp and up per tube filament. My amp with it's 12-6550 collectively draw 19.2 amps@6.3 volts. In the winter it makes a handy space heater, which works out fine for now, as I heat with electric anyway. In the summer though it is just a pure waste of power. There are several medium sized tubes that might be hot enough for a filament-less plate only activation...in my amp those would be the two 12BH7s that I use as phase inverters and drivers to the outputs. They get pretty hot too. But the output tubes in any tube amp would be the first choice in developing a line of direct replacement output tubes for the legitimate audiophile market, of which there has been one since the late 1940s...and yes, I am one of those people who enjoy the sound of a tube amp, especially if is of the corclotron topology....thanks for you interest....edd

hello ed earl...the hum issue is minimal as there are several ways to minimize it. There is going DC, but which is cumbersome. A better way is by placing a small positive voltage on the AC driven filaments in respect to the cathodes, which is easy (takes 2 small resistors), hum in the voltage gain section of an amp is reduced to near inaudibility.....there is also a less elegant way by introducing a reverse phase of the hum, adjusted to afford cancellation. Yes, some noise is in a tube amp, but any amplifying device has some noise. The thing about distortion is that tubes, when overdriven, deliver an even-order harmonic distortion, which is much more appealing than the odd-order harmonics from a similarly overdriven SS unit....and some levels are distortion are alway present in either system with respective predominate orders of distortion. Guitar players like the even order, with the exception of the heavy metalists who like the harshness and discord of the odd orders...many SS guitar amps have a special overdrive circuit that "fakes" an even harmonic overdrive using a completely transitor/IC driven amp.....and they sound pretty good. I have one or two myself. Stomp boxes have taken the synthesis to an artform and actually sound very appealing, with their so called "tube modeling" creating various type of collectible amp sounds in a multi-select mode....edd

[EdEarl]

 

If the article in scientific american is correct, the emission of a diamond coated cathode is limited by the current carrying capacity of the metal substrate......and that they would outperform regular tubes and solid state devices.

I am curious: In what way do these tubes outperform solid state devices? I would think solid state devices would be smaller and less expensive. Perhaps they mean gain per device or some aesthetic quality. I know tubes have been made that are very large and capable of operating at high voltage and current, perhaps greater than solid state devices, maybe that is what they mean by outperform.

[hoola]

as a vintage electronics collector and repair-person, I like both tube and SS amps, and have both. I have built both and with not just the ascetics of tube architecture, but the actual sound quality, presence, sound stage, or whatever you might call it, seems different to the ear and is technically different. The even-order / odd-order distortion comes into play, but for me the question is the negative feedback used in amplifiers to achieve stability and a desired response curve. I build my own amps and incorporate local negative feedback within a stage, and not have what is termed "globlal" neg. feedback. This is especially important as, although there is endless debate as to if it can be heard or not, global NFB sends a reverse phase signal from the output back to the input in a sort of "race track" of the signal chasing itself around the circuit on an endless quest of keeping the amp parameters in check. This is more of a problem with SS units as the speed of electrons in a vacuum is less in a solid (transistor), than in a vacuum (tube). It also has been used as a crutch to make a basically unstable design rendered stable and reliable enough to be sold to the public. The feed-backs of various types are one type of way a company will have equipment with a characteristic sound quality that some people prefer, just as some people prefer listening to vinyl over CDs, though they are technically inferior. This is one debate I fall on the no "global" NFB side on. I figure, if you can build an amp without it, don't use it. It is largely used industry wide as a money saver as well as giving it a certain sonic character. Another aspect is the DIY aspect of someone who like odd and unusual electronic ideas. What I have built is a "circlotron" amp, requiring multiple output tubes, hence the desire for the filamentless ones. With certain arrangements the amp doesn't need an output transformer as used in all other amps you might see...by hooking the speakers directly to the output tubes there is a technical advantage as the output transformer is a major limitation of frequency response range and power capability, as well introducing it's own form of distortion. That form of distortion is another endless debate as to if it can be heard, called the "flywheel" effect. As a signal voltage waveform to the transformer primary rises and collapses, the energy is transformed into a magnetic field, which is picked up by the secondary, and converted back to a voltage signal. That causes an effect as a collapsing magnetic field tends to continue a reverse EMF for a brief period even after the input is gone....this effect is taken advantage in what is known as a "class c" type amp used in tansmitters, mostly by ham radio operators....in audio, the effect can actually be a small positive, as it "fills in" gaps in the D/A converter's output in a CD player, smoothing out the somewhat chopped up high end as the limitations of certain cheaply recorded CDs become apparent. So, the improvement in removing the transformer reveals the defects within the source material, which is a rule of thumb in the audiophile crowd..... As your overall sound system improves technically, the limitations of the source material become more apparent..... So, my answer is to improve the source, in this case, the SACDs which take a special player, but with a higher sampling rate. Supposedly they are clean in their high end up to over 40 KHZ. Too bad they didn't really take off, but even regular CDs sound pretty good. Some have "special bit mapping" or other listed tweaks, and one of the best CDs I have ever heard is Santana's Abraxas. It has the special bit mapping. I have a SACD player, and my friend brings over a few SACDs, but I tend to not buy on ebay, and that is the main source of them.....another source of high end regular CDs seems to be in the classical stuff put out by deutsche grammophon. The problem with some SACDs is that they are recorded with the intent of being played in surround sound, and don't sound right on an old school 2 channel stereo as I have. Other SACDs did ok, but I have only hear a few, listened to briefly when my friend left some with me for a while....and they seem to be a hit and miss proposition...I have heard talk that CD technology is already nearing extinction....moving to flash memory only....kinda sad....edd

[hoola]

ed earl ...here is few lines from the SA article in answer to your question: (in what way would these tubes outperform solid state devices?)...I quote the article......ARTICLE TAKEN DOWN PENDING APPROVAL OF MAGAZINE....edd

Edited December 12, 2013 by hoola

[EdEarl]

Yes, thank you. Would you reference that article, please.

[hoola]

the article is from Scientific American October 1992 (vol. 267, no. 4) titled Diamond Film Semiconductors, pages 84-89, authors Michael W. Geis and John C. Angus. And I have taken down the article as I have read the fine print on the head page of the magazine and they specifically prohibit reproduction of the article without permission....sorry. I will contact them for permission and then will re-post if they allow me to. For a complete article, I suggest going to their webpage and downloading the article. I will do this and see how this is done, and if there is a charge. I will get back to you with details. I do not wish to be in trouble with the S.A. magazine, even though I doubt they would be very upset with posting an article over 20 years old, but I will do the right thing and try to correct my error....edd

well, I just went to the SciAmDigital.com and they list their archives as starting in 1993.....3 months post the magazine article in question. I am hopeful that it is available from them through some extended search, and I will continue to persue the matter. There is phone number of 1-800-925-0788 for print copies listed, of which I will check out also. Sorry, I expected it to be quick and get the article back on tommorow....I will call the phone number and see what can be done....perhaps after 20 years the articles become public domain?..... I am wondering if a copy could be purchased on ebay? ......edd

just checked on ebay...one issue seems to be there for a "buy it now" $9.77 ....edd

Edited December 12, 2013 by hoola

[EdEarl]

tyvm

[Enthalpy]

From what I read, current density from nanocrystalline diamond films only outperforms other cold emitters but is very far from a hot filament. You can be pretty sure that a breakthrough in 1992 would result in widespread use in 2013.

 

As for sensitivity to residual gas: vacuum obtained by pumps in electron microscopes is excellent - no hope to outperform it in a tube with a getter. If this were the key to field emission, all users of microscopes would already have improved pumps, since field emission would bring so much to them: tiny emission zone and interesting current at the same time. Alas, it's unuseable.

 

Better devices with electron beams: academic research hunts more uses than audio amplifiers. First where electron beams are still not replaced: electron colliders, powerful radiowave transmitters, X rays sources... and then for uses where they might outperform semiconductors, say in the upper GHz frequency domain - maybe. One very strong incentive are displays with one electron beam per pixel and colour.

 

Many research papers use cold cathodes as these suffice for half an hour experimentation, but up to now, all commercial devices have a tungsten filament or a LaB₆ or SrB₆ warm spike.

[hoola]

the lack of development of the reputed efficiencies of the diamond cathode in industrial or medical uses as a cold cathode emitter seems evidence of an early overestimation of the technology's potential. Or of the traditional method of hot filament approach being so well understood, reliable and cost effective, that there is no incentive to develop an alternative. The article states a "crystalline diamond" coating was used in the MIT prototypes. Not a "nanocrystalline", for whatever that detail is worth.....and that the article stated a reduced output was seen in the prototype compared to standard emitters, but the theoretical efficiency seemed quite high given proper development....this is starting to appear to be similar to the "pebble-bed" type nuclear reactor talk I heard of years ago that seems to have not delivered the benefits early proponents were giving, that I had such hopes for....or even fusion reactors for that matter...alas indeed, sir.

Edited December 14, 2013 by hoola

[Enthalpy]

The incentive to replace hot filaments with cold spikes is huge! To make a display with one electron flow per pixel and colour, filaments are impractical while everyone hopes field emission would enable the option. This is a many-billion market. The only difficulty is that field emission isn't useable; the weak current would probably have solutions, but sensitivity to residual gas not, and this is a no-go.

[hoola]

seems the gas could be "gettered" out like any normal vacuum tube.....but I guess this is a no-go idea par excellence....edd

[Enthalpy]

Screens with cold spikes would already be there, if a getter did suffice... Pumps achieve a much harder vacuum, but not sufficient for long operation of cold spikes. They get destroyed by a single molecule stuck at the tip.

 

Researchers make the best possible vacuum, bake the spike for long in the vacuum, carry their experiment in a quarter hour and publish. Little to do with a screen whose million pixels must work for 10 years.

[hoola]

I am proposing using the vacuum pump, achieving a hard vacuum, then using the getter after the fact to suck up residual gasses that the elements from the tube, as in a regular tube....with an additional getter mounted on the hot plate to allow that to activate it instead of the filament heating it. I don't think one molecule of air in a tube, hot or cold cathode would hurt anything, or no tubes would work...there was a guy in S. Africa that tried to make a large flat screens using the diamond cathode idea years ago, but his efforts got eclipsed by the plasmas, then the LCDs, now the LEDs , so the only viable market would be the boutique high end audio market...which sadly seems terribly disinterested.....I sometimes think I am the only one who wants the idea to get picked up by Winged C of russia or JJ of slovokia, or the chinese guys.....edd