Future of Tubes is TODAY

[g1]

How can we know it wouldn't have sounded just as good with the tubes bypassed?

[Chuck H]

What I'm puzzling over is that it was stated to have a "Nutube preamp section and a Nutube power amp section". What part of watt production relative to power could the Nutubes possibly be performing?

Even if the amp is essentially solid state with the Nutubes in there as ad candy I think they're onto a good avenue of product that is better than average compared to the competition. Which I guess would only be the Peavey TransTube stuff.

[teemuk]

My hunch is that thing features a "ValveReactor" power amp with a NuTube dual triode instead of 12AX7. It's output section is most likely generic class-D.

Don't know about the preamp but it would be logical to do D/A convert signal, run it through a pair of NuTube-based gain stages, then A/D convert back. Dynamic interactions that alter the "transfer curve" of a gain stage are natural in analog circuits but have to be implemented in DSP, so realistic emulation of something like a pair of capacitively coupled triode gain stages is going to be computationally more taxing than mere static "waveshaping". Likewise, adding harmonics, also known as distortion, to signal means that - due to reasons explained by Nyquist theorem - the data must be upsampled before waveshaping process. Since sample data increases considerably due to upsampling even the static waveshaping process becomes computationally somewhat taxing. And then we run into latency. Analog processing has its merits, and they can always modify the analog signal path / circuit with switches controlled by digital logic. Kinda like what some ValveReactor power amps have done for about a decade already.

Peavey TransTube concept, particularly Vypyr series, is feature-wise very similar concept: Largely DSP preamp, but distortion is actually generated in pair of analog "TransTube" gain stages. Power amp is a marriage of generic voltage power amp (can be any class, doesn't matter) and analog tube emulation (T-Dynamics). These new Vox products sound very similar to Peavey Vypyr series with the exception that key parts of analog circuitry employ VFD triodes instead of solid-state circuitry that in practice behaves about identically.

[teemuk]

Seems cool...but proprietary. Also, could kinda care less about a triode...more interesting, to me, would be a pentode version.

The idea of using VFD triode for other applications than displays is not new. The concept has likely been around as long as VFD in general and all in all "prior art" is pretty well documented in medias like Internet. Korg has merely packaged the idea, it's no longer domain of DIY tinkerers. But in practice VFD triodes have very low tranconductance figures, high distortion and nasty sensitivity to microphonics.

But look no further than Nuvistors. Real vacuum tubes (triodes, pentodes, etc.) in a package that was about the size of a vintage "metal can" opamp. Optimized for low microphonics and low current consumption. Basically, stuff to compete with transistors in the space race etc. And they were actually developed as generic vacuum tubes, not as "display units". So expect much better performance than a modified VFD can offer.

[Gregg]

There are several products claiming that they "sound and perform exactly like a tube (triode)" but are SS (like the AMT one for example - it looks like a tube and fits directly into a tube socket). Even if that was the case I think that they will have a very small niche because it's the same thing like a margarine that tastes like butter. Probably you can nail the taste but you still know it's margarine.

[J M Fahey]

Agree with above posts , but would like to add:
DIYers had to experiment with actual displays ... because that's the only thing that's available out there, specially as cheap surplus.

Noritake *might* want to give some value to 2 Billion suplus never sold or used microwave oven / keyboard / clock radio / whatever displays but I suspect what they want to find new customers for is their *Technology* , experience and already installed (and fully paid for) Factory.
Did I mention fully trained workforce?

So they can simply design (in their sleep, it's such an easy task compared to their other products) a single or dual triode and call it a day.

* filament? ... check
* grids? ... "just" two? ... not 27 or whatever? ... check
* anodes? ... "just" two? ... not 27 or whatever? ... check
What? ... no fluo dust coating? ... is this a joke?

Won't even mention glass and vacuum, because they are there.
And the bonus since they developed displays so well, that these triodes will be happy with, say, +30V or so, easy to get in SS amps.

Only minor problem is that they need, say, a single 3V well filtered DC filament supply.

So in a nutshell, they can *easily* (and that's the understatement of the year) build a double triode, in fact they could easily pack a dozen of them in a single envelope if they wanted , and call it a day.

And since plates must not be punched and twisted in weird ways to make segments or symbols I bet it can be made more linear, maybe higher transconductance and less microphonic than a regular display.

Designers would be working under way more relaxed constraints here, and we'd end up having *real* triodes, by definition, specially built for Audio, of course.

[carlg]

Seems strange to me that it's necessary to go this route. After all these years you think somebody would've just gone ahead and designed a fet with curves identical to a 12ax7. I mean, Nelson Pass has been known to get custom jfets made for his amps and he's pretty small time compared to Korg. These guys have a billion dollar. I guess they'd lose the marketing power of saying "our product has tubes in it!" but personally if I could build a circuit with fets that sounds the same as one with 12ax7s? I'd do it.

[Gregg]

Seems strange to me that it's necessary to go this route. After all these years you think somebody would've just gone ahead and designed a fet with curves identical to a 12ax7.

A single FET would hardly be possible but as I mentioned earlier there's a ready to go SS 12AX7 replacement (I guess build around FETs):

AMT Electronics USA, LLC - 12AX7-WS Warm Stone Solid State 12AX7 Tube

[Chuck H]

I saw a similar product at NAMM in 2009. Maybe the same guy.?. It was in the early stages of refinement at that time. Didn't even have the faux plastic nipple on top

[carlg]

That's pretty different than what I'm thinking of. They seem to be trying to get a piece of the replacement tube market by making something that will drop into existing tube amps circuits. I'd be happier with something that's made to get the same sound in it's own circuits at normal solid state voltages. No second heater supply or bulky high voltage components. I don't know of a reason why a billion dollar company couldn't develop a custom fet with whatever transfer characteristics they desire, but if that's impossible for some reason in a discrete fet surely a small custom IC could accomplish it.

[carlg]

I mentioned Nelson Pass having these kinds of custom silicon made before, if anyone is unfamiliar with him but interested in this stuff this is a quick article about that type of stuff: FIRST WATT SIT INTRO

He's just one man with a dream. I have to think a huge company like korg could afford to go much further if they actually wanted to.

[tedmich]

A single FET would hardly be possible but as I mentioned earlier there's a ready to go SS 12AX7 replacement (I guess build around FETs):

AMT Electronics USA, LLC - 12AX7-WS Warm Stone Solid State 12AX7 Tube

$40 each but only sold in sets of 4 ($160)


Note from Mfg. "Before you buy, make sure that the anode (plate) voltage in your device at is least 30V"

12AX7WS Warmstone - 12AX7 Tubes Solid-State Equivalents, buy 12AX7WS, AMT tubes

spec sheet
http://www.amtelectronicsusa.com/AMT12AX7WS.pdf

[loudthud]

I use what this patent calls prior art. FIELD EFFECT SEMICONDUCTOR DEVICE HAVING AN UNSATURATED TRIODE VACUUM TUBE CHARACTERISTI
As far as I know, nobody makes the parts that would be covered by the patent.

The parts I have are tricky to bias, you need an opamp to force them to behave, but it works flawlessly. I couldn't have done this without my trusty Tektronix 576 curve tracer. The only parts I use that were not available in the 1980's are the output transistors.

Anybody in the Dallas area PM (not email) me and we can setup a demo. Bring your camera.

[Gregg]

I saw a similar product at NAMM in 2009. Maybe the same guy.?

AMT is a Russian company. Their engineering team consist of people who are also guitar players and they know what they're doing. Some of them were known to be forum gurus in the past and they know the stuff very well. If they claim that SS 12AX7 sounds like a tube maybe it is so but it's still a "margarine that tastes like butter" for me.

$40 each but only sold in sets of 4 ($160)

In Europe they retail for 35 Euro the piece:

AMT Warm Stone 12AX7-WS - Thomann UK

[bea]

A solid state device requiring a heater?