Korg 'Nutube' - next generation vacuum tubes

[frequencycentral]

Yummy!

http://www.korg.com/us/news/2015/012212/#

[wavley]

That thing looks pretty awesome.  I thought it was going to be another one of those solid state tube replacement things.

I can't wait to see a datasheet and how much they cost.  I'm very curious about the sound, there's all sorts of pedal based tube stages that sound super fun in applications where it just wasn't cost or space effective to use tubes.

[Mike Burgundy]

Very curious as to how these work. I remember years ago reading about (TU Delft, I believe) a solid state tube that actually *did* use a cathode/grid/heater construction, but made on a chip-die with modern etch/mask procedures. The cathode wasn't heated, but sprayed electrons because it was shaped into a microscopic spike. Insulating layer on top, on top of that a conductive layer with a hole in it for the grid...you get the idea.
Never heard anything about that anymore...

[amptramp]

Very curious as to how these work. I remember years ago reading about (TU Delft, I believe) a solid state tube that actually *did* use a cathode/grid/heater construction, but made on a chip-die with modern etch/mask procedures. The cathode wasn't heated, but sprayed electrons because it was shaped into a microscopic spike. Insulating layer on top, on top of that a conductive layer with a hole in it for the grid...you get the idea.
Never heard anything about that anymore...

I saw some displays done like that where laser pulses were used to blast tiny diamond cones out of a carbon target.  They were field emission cathodes with no cathode heat, using the extremely high field intensity around a point to achieve several hundred cones per pixel.  The display had a grid and used a zinc oxide phosphor that ran at 60 volts.  They also used a grid for multiplexing.

The Korg tube appears to be small signal only, suitable for a preamp but nothing with more power than that.  I am looking for a future with carbon nanotube cathodes with a negative work function for small signal amplification.  There was work being done on low-temperature and room-temperature cathodes, but it all got swept away in the semiconductor juggernaut.  I would like to see the noise figures for the Korg tube.  If it is running in a temperature-limited rather than space-charge limited regime, the noise will be higher and a triode just might act like a pentode.

[vigilante397]

Yummy!

http://www.korg.com/us/news/2015/012212/#

So hang on .... am I understanding this right? This thing could change everything? We can run our tube amps at lower voltages and in more compact designs with less heat as well? 

Unless I misunderstood something (or they turn out to be super expensive) THIS COULD BE A HUGE DEAL!!!

[PRR]

> Very curious as to how these work. ... a microscopic spike
> blast tiny diamond cones out of a carbon target
> appears to be small signal only, suitable for a preamp but nothing with more power
> am I understanding this right? This thing could change everything? We can run our tube amps at lower voltages and in more compact designs with less heat as well?
> THIS COULD BE A HUGE DEAL!!!

Move on. Nothing to see here. This is all VERY old-old-stuff.

It is Noritake's vacuum fluorescent display (VFD). We all know VFDs. They are/were a favorite in microwaves, VCRs, DVDs, and car-audio 1980s-2010. They were expensive but it cost very little to have special words, symbols, logos built-in. My 1994 truck heater has a VFD with little hands/feet/windshield icons. The present DVD machine here has a VFD with the counter/time and symbols for the too-many types of DVDs/CDs it can play. The X-Plod radio in the car is an eye-popper with every type of EQ (jazz/classic/rock/easy), a blazing waterfall (I guess) for volume indication, etc.

A VFD _is_ a vacuum triode, plus some glow-stuff, and recently they put dozens or hundreds of triodes in the bottle.

The main difference between an amplifying vacuum triode and a VFD (aside from the glow-stuff) is that a VFD needs very little current or gain. The geometry is optimized for low-low cost and silk-screen design, not to be an amplifier.

The original VFD is from 1959, the DM-160 from Philips. That one made you diddle the control grid directly. The up-surge of many-object VFDs in the 1980s was due to integrating a digital controller to allow a few lines to control a hundred objects (you can't diddle the grids).

You don't have to get the memos to suspect the VFD business is on a decline. New toys use LCD/Plasma displays so they can put up ANY image, not just built-in images. Cheap new toys (my 2013 microwave) fall back to digit-only green LED (to make the fancier color LCD models look better). VFD is in the crack, and the crack must be narrowing.

So some guy from Korg was crying in his beer (sake?) about the high cost and poor life of 12AX7 and similar devices. And some guy from Noritake was under the same table crying about the fade-out of VFD production and his job. Each heard the other say "vacuum", and after swapping a few curves they realized that Korg could use the mature but going-out-of-style VFD foundry to make low-performance vacuum triodes at going-out-of-business prices.

While the digital decoder would be left out, VFDs usually have a "high"(ish) voltage converter inside which avoids one issue.

It is what it is. A small to very-small vacuum triode of poor perveyance (low conductivity) but "the right curves" and proven long life (with declining brightness/current). Small bent signals should be (are) very easy. The fact that the foundry machinery now has capacity to spare is a big economic bonus.

[vigilante397]

Fair enough, cross that one off the "things to be excited about" list

[PRR]

As said, modern VFD product does not have the grids available for play (you talk to an onboard chip which decodes to the right segments).

So I am forced to guess from the data on the DM-160 of 1959. (As the Korg puff says, very little new has happened in the West since then.)

The Mu(gp) of a DM-160 is about 30. The saturated plate resistance is around 120K. With 50V supply, practical plate resistor, and reasonable audio load, the in-circuit voltage gain is about 10 or 15, the maximum RMS output may be 3V to 5V.

Of course 3V-5V is more than enough to slam ~~1V amp inputs (or if not, we have 19 cent opamps to boost weak tubes).

The maximum operating zone of a DM-160 is much much less than a 12AX7. 65V plate, 0.75mA current. (12AX7 can stand 330+ V and around 3mA.)

DM-160 curves in pink plotted over 12AX7 curves:



The dot far-right is roughly the op-point of a 12AX7 in a Fender or copy-cat.

DM-160 curves are very much "triode". However we love our 12AX7, perhaps because its curves depart from classical triode: the zero-grid line bends up due to electron velocity.

Are modern VFD triodes different from DM-160? Sure, but IMHO maybe in a direction of "smaller", lower amplifier performance.

OTOH, you can get 69 or more segments in one VFD. If Korg is just asking for two "segments", they can be equivalent to many single segments. So I will not be shocked if it turns out that Korg's devices run dozen-K impedance at over 1mA each.

The VFD "cathode" is directly-heated oxide-coated filament. DH/OC filaments are darling among some audiophiles. OTOH, these filaments are more like the ones in battery power beach-radios than the robust cathode of a WE 300B. And for VFD use, the filaments are run as cool as possible so the constant glow does not detract from the clever segments. For non-display use the filaments could be run hotter, but that very soon shortens the life down more like a beach-radio tube; and long-life is one of Korg's promises. So I think the filaments won't glow much. Also they will not have the high electron velocity of our beloved 12AX7. Of course it "has to glow", so Korg probably retains the phosphor glow-stuff of a VFD. That glow has nothing to do with thermionic amplification, but it is more honest than putting LEDs under glass.

[PRR]

I've manipulated Korg's image so we can see the working part.

The overall package is a hollow glass slab with pins. If you bust-open some old car-radio or VCR you will find a similar thing behind the display bezel.

I have pink-outlined the working part.



The horizontal white hair is the filament. Compared to traditional tube assembly, it is very simplified. No bends, spot-weld tabs, etc. In part because in a display system, filament power is a small part of total radio/VCR power, so thrift is not needed.

Under the filament is a hexagonal screen which is the control grid. They probably stamp this out by the acre, instead of hand-turning on a winding lathe as in traditional amplifier tubes.

The grey and blue area below that is the plate. The blue area is phosphor (glow stuff) and I suspect the grey area is bare metal.

All micro-stampings and spot-welds. Much smaller than a car-body, but not so small the workers need real magnification.

As an amplifier----

The grid-plate right under the filament hair has fair conductance, but there is a lot of grid-plate which only "sees" the filament at an angle and a long distance.

Some of the grid area and all of the grid support stamping has no control of the electrons, yet has capacitance to everywhere. This is dead-weight which reduces the gain-bandwidth. (But since "optimized" medium-Mu triodes have GBW far above the audio band, and a music flavor amplifier often wants to limit excess treble, this may be entirely fine.)

I just noted they suggest battery operation. This being a VERY teeny triode of efficient construction, that is practical. But who uses batteries today? And this is far-far from a power amp, you still need a power-plug to make a big sound.

The VFD technology, specific fabrication details, and Noritake's process, are VERY mature. The changes to bring-out K-G-P pins are trivial. So device reliability should be much better than today's perverted market for bottom-bid Chinese audio tubes, and very near the very best Golden Age vacuum tubes. We all have VFDs soldered into our lives and that would not be true if they were not incredibly reliable.

This is a win-win for Korg and Noritake. Real triodes at over-run prices.

If you know vacuum-tube amps, you know it isn't just the tubes. The VFD is not a conventional tube. Amps using different conventional tube-types tend to sound different. Heck, amps that use the same tube types and steal circuits from each other will sound different. There is a lot of tweaking and serendipity in a good guitar amp. Just throwing a triode in doesn't ensure musical happiness.

Korg of course is the kind of people who will probably do the tweaking. They've stayed too near the top of a competitive heap too many decades to be skimping the tweaking. This will be a musically useful product with "a magic feature" at a fair but not-cheap price and good distribution.

Of course if it sells a truck-load, B-ringer will find some VFD machinery and copy-cat it. Or even collect VFD rejects to dress-up a cheap DSP pedal.

[PRR]

A Guide to Fundamental VFD Operation
http://www.noritake-elec.com/vfd_operation.html

This suggests that modern VFDs are operated in the positive-grid area. This gives more plate current (glow), but is usually avoided in small-signal amplification because positive grid means "large" grid current instead of very low grid current. The grid current is not large enough to bother digital semiconductors but would be a major impact on another vacuum triode driving this grid.

All of which can be sorted by clever design of device and circuit to suit audio amplification instead of glowing displays.

Fig 10 suggests the electron flow is not thermally limited.

[StephenGiles]

Crikey Paul you never cease to amaze me with your knowledge! A bigger question is - how do you get an iphone to read a flash drive???

[bool]

... mucho deep stuff ...
Or even collect VFD rejects to dress-up a cheap DSP pedal.

First things first. Imho.

[vigilante397]

A bigger question is - how do you get an iphone to read a flash drive???

Get an Android

[amptramp]

I have an old battery tube portable AM-FM radio made by Philips that uses a DM71 tuning eye tube as the paraphase inverter for the push-pull output stage.  It is quite possible to use tuning eye tubes as amplifiers with a µ of about 15.

Tuning eye tubes are usually limited in life by darkening of the phosphor and the zinc oxide phosphor only needs 60 volts for an acceptable output, so they are restricted to these lower voltages.  It is quite possible to eliminate the phosphor or change it to P4 so the plate voltage can be raised to 250 volts and more output generated.  And the single wire cathode can be folded back on itself several times to look like the cathode on the original Audion triode to get more current.  I see this as viable - it just depends on whether anyone has the will to continue on with the development.

[GGBB]

Crikey Paul you never cease to amaze me with your knowledge!

Ditto. Thanks, Paul - extremely helpful and informative.

[frequencycentral]

Crikey Paul you never cease to amaze me with your knowledge!

This! Very informative and insightful PRR.

[Bucksears]

Regardless of whether or not THIS is the next big thing, I do like the fact that this sort of thing is being researched.

[PRR]

> How did I miss this?

You should read a forum hosted by some guy named "Aron". Many good things there.

> impede the production of tubes in North America, or anywhere outside former Iron Curtain countries was the cost of compliance with environmental regulations, given the materials that vaccuum tubes normally employ

I'm not sure this is true.

Tube demand and  production sagged VERY badly in the early 1960s, when "environment" was just being talked-about, and fingers were pointing at broad-scale pollution such as the LA Smog which rotted your spare-tire in a couple years. Many tube makers got out of the business knowing that the end was nigh, they should move their capital to semiconductors.

Yes, later some of those former tube-plants became problem-sites. FWIW, there was a mid-1960s stink about the semiconductor plants pouring toxins on the ground.

And it may be related that the last receiving-tube plant in the US, GE's KenRad, which worked military types into the 1990s, was in Kentucky which is not a high labor-cost area and tolerated massive coal-mile pollution for a century.

Of course tube production had a very long tail. Many of the US and EUR machines had already been moved to East Europe for lower labor cost and closer access to the probable long-tail customers, as well as for diplomatic reasons, and perhaps because dumping was not a problem.

> I'm curious about whether this technology is able to sidestep both the production costs and the environmental risk.

First, I would really bet these are made in China now. China's coal-smog spews heavy metals much better than any tube plant could. Tube (VFD) production is high-value, the pollution is very local, China won't care for decades.

Second... these are very teensy and low-performance. The materials are mostly not very exotic. The rare-earth oxide coating is very small even by beach-radio standards. And being used straight-line, it is probably a continuous dip instead of the spray-gun sometimes used for large folded heaters. The grid and plate are simple punchings. The bottle is mostly glass, not even a high-temp glass. Getter metals are expensive and must be worked semi-air-tight, probably not much waste there. Frankly this may be nicer to live near than a junk-auto business. (My other neighbor starts a car-fire about every month, torching out gas-tanks before selling the scrap iron.)

Costs.... the plant is in-place, the mortgage is paid-down, local enforcers have been paid, and demand is declining. Factory cost of VFDs (or a product with minimal changes from VFD) is as low as it will ever get.

Finally: KORG is not going to sell these to DIY guys, that's not what KORG does. I bet they pay Noritake $2 a bulb, put it in a "NuTube Pedal" listing for $99. Plenty of mark-up for KORG and the rock-shop which sells them. If sales are large, B-ringer will find bulbs for $1 and sell the pedal for $49, still with plenty mark-up.

I suspect the tube is so specialized (lame for general tube work) and that Korg will tweak the pedal-product so very-fine, that it will not make sense to buy a $99 pedal to extract a bulb when better tubes are readily available for $19. And that all else inside will be SMD so not moddable with my old eyes and fat fingers (the more clever folks sure will find mods to do).

[aron]

> Finally: KORG is not going to sell these to DIY guys, that's not what KORG does

Hahaha. I asked my contacts for a tube. Haven't heard back yet. :-)

[wavley]

> Finally: KORG is not going to sell these to DIY guys, that's not what KORG does

Hahaha. I asked my contacts for a tube. Haven't heard back yet. :-)

I don't know, opening products up to DIY has been one of the things they've been doing lately... free crowdsourced development.  You never know.

http://www.korg.com/us/products/dj/monotribe/page_5.php