Tubes on a Chip

The tubes you’ll find in guitar amps and high-end stereos were first designed in the 30s and 40s, and when you get to really, really advanced tube technology you’d be looking at extremely small tubes made in the 70s for military applications. For 40 years, there really haven’t been many advances in tube technology. Now, at last, there’s something new.

The Nutube 6P1, as this curious invention is called, is a full triode or half of a 12ax7 you’ll find in just about every tube amp ever. Unlike the 12ax7, it consumes 2% of the power required of a normal tube, is 30% of the size of the normal tube, and lasts for 30,000 hours.

This new tube-chip thing was brought to life by Korg, makers of fine musical equipment and Noritake Co., manufacturers of vacuum fluorescent displays. There’s no word on what these tubes will be used in and there’s no data sheet. There will be further announcements this year, so don your speculation spectacles and head to the comments.

72 thoughts on “Tubes on a Chip

      1. If I remember correctly there already was an electribe with a tube in it. However, I wonder if it was in proper configuration or in a starved plate design. Hope this new chube (CHiptUBE) will be a proper one!

  1. Well that’s clever, I guess it’s an IT instead of IC, an integrated tube. All jokes aside, I can’t see this selling too well, half the reason people own things with tubes is for looks. I mean, some audiophiles will say they’re better, but the majority buy them for looks.

    It would be neat to see some modern revisions of nixie tubes, if I’m not mistaken one hasn’t been made in the last 30-40 years. And those things are friggin’ nifty.

    1. I dunno, it looks like it is glass right? Most new tube amp projects I see they put either blue or RGB LEDs underneath to make it all pretty. I think that would work with these too.

      1. They’re saying it’s low power, so if they’re making them like vacuum-fluorescent displays, the filaments will probably not be hot enough to see glowing. Of course the people who design them into products will probably put orange LEDs behind them or something.

        I think it’s hilarious when tube folk put blue LEDs under their vacuum tubes, since with real vacuum tubes, a blue glow means a gassy, failing tube!

    2. It seems like musicians still like to use tube amps a lot, so I’m guessing this is aiming at replacing their existing slightly unreliable tube amps with a more reliable longer life tubish amp.

      I was at a gig recently that ended a couple of songs early due to tube failure.

      1. Don’t call it a joke yet.

        The neat property of tube amps is that they can handle more power at higher frequencies than transistors, so they might indeed be used as RF amplifiers to upgrade your internets.

  2. This is very cool for amplified instrument players everywhere. Most people don’t know it but tube amps are much warmer and rounder sounding than IC amps. The problem is cost and wearing out of tubes, this should help with both cons.

      1. I don’t think the sarcasm was warranted, BrotherGecko’s comment probably has anything to do with audiophiles. Guitar/instrument amps are designed to change the input sound, not faithfully reproduce it. It’s like adding spice to food. Tubes change it differently than transistors. Believing tube amps sound better than solid state amps is purely a personal preference, but like most people prefer chocolate to licorice, most musicians prefer the sound tubes amps over solid state amps.

    1. Tubes have different non-linear operating modes than transistors do. They are also in general much higher distortion devices than transistors. This makes it easy to build an amplifier out of them that pushes even order harmonics as opposed to the typical odd order you get out of fet based amps. I don’t know how to measure the warmness or roundness of this though. Could you tell me how many mV/sqrt(hz) is equivalent to a degree kelvin?

      1. Hmm, I am no fan of tubes but it’s true that tubes have lower distortion than transistors.

        That is, if you include noise as a part of distortion. Technically noise is a part of distortion because distortion is any component of the output that is NOT a component of the input.

        Transistors are exceptionally noisy when compared to tubes. Noise in tubes comes from secondary emission and good design will minimize this. Good design can’t remove noise from transistor junctions.

        On the other hand, if you consider distortion alone without the noise component then transistors have the advantage that good design will use other transistors to compensate for the non-noise distortion. In this way you can say that transistor circuits have a lower non-noise distortion than tubes.

        It’s the non-noise distortion in a tube that many use colorful terms to describe as a desired quality. So in essence the non-noise distortion in tubes is described as a pleasant addition where non-noise distortion in transistor circuits is either cancelled out or very unpleasant.

        Each to there own I guess.

        1. The number of “sides” on the tube-vs-transistor sound-quality debate seems to be exactly *two*. *Maybe* two and a half with the whole “to each their own”-style response.
          I propose: If the original mastery of the recording was intended to be played on a tube-amp, then its sound is *truer to the original* if it is indeed played on a tube-amp, distortion-included. Likewise, if it’s mastered for transistors, then transistors would be truer, and if it’s mastered on a Class-D amp, then.. you get the drift.
          Listening to an original album from the tube-era via transistor-amps can’t *possibly* have been the intent of the masterer, and I think a decent producer would know to tune their recordings for the equipment their audience has access to… Likewise, if mastered for an LP, then certainly a decent producer would take into account the effects of needle-wear…
          Dunno, just a thought… if “truer to original” is of any importance.

          1. Why can’t we all just get along? Electrons are electrons and we can move them in many strange and mysterious ways. Since I am almost as old as the transistor myself I have heard many sides to this argument over the years and have been a professional recording engineer since the mid 70’s so I have a bit of experience in the matter. Many high end ($2-4k) microphone preamps being produced today have both a tube and a transistor front end and allow selection of one or the other or a blend of both for the sonic “flavour” the engineer desires. As far as “mastering” goes, god help us when engineers start mastering for mp3 playback! Although your comment is well meaning it is somewhat misinformed. By the way I have mastered a few “records” myself, albeit not for vinyl and I do know something about the process.

          2. Actually, Miles, I think I’ve seen songs “mastered” (you’re right, I might be misinformed, is that the right term?) for MP3… I wouldn’t even know what to search for, but as I recall they’ve done some interesting things with loading *images* and text into them *for the sake of sound-generation* (as opposed to images/text in the MP3 tag which is just image/text data). I’m almost certain I remember this being an MP3-specific venture, not just images in the regular FFT of the regular audio, but again, I could be misinformed. Also, I’m sure I’ve heard of music designed to take advantage of MP3 artifacts… ;)

          3. A recording can be a “record” of a performance, say a string quartet, in which case you want the final product to sound as much like 2 violins a viola and a cello as possible when played back. A recording can incorporate elements that become the performance, such as taking advantage of MP3 artefacts (a dubious prospect) I have been involved in both and they are just different, neither one better or worse than the other. I once did a recording which was released on flexidisc and we didn’t master it at all, just sent the tape off. The artist wanted people to scratch up the disc and released it in a folded piece of clear vinyl carpet runner for the sleeve! Needless to say it had no commercial potential, but it was interesting. Mastering is highly over rated in the post vinyl age, it was originally introduced into the recording process to take into account the fact that magnetic tape had a larger dynamic range than vinyl and prevented overmodulating the grooves so the needle wouldn’t fly out on the loud bits. Now it mostly consists of compressing the piss out of everything similar to what they do to commercials on TV to make them seem louder.

  3. I dunno how much power these will save. half of a 12AX7 heater runs at 6.3V/150ma = around a watt. That’s a tiny power saving. Now, if they have decent performance at lower B+ voltage that would save cost and complication in the power supply. But thermionic small signal triodes typically require voltages of 100 – 300V to operate.
    If you want some tube ‘sound’ (the even harmonic distortion) you can run a JFET way down in the nonlinear part of its transfer function. It will even clip similarly.

    1. From the website: “Miniaturization dramatically reduces power consumption, allowing the Nutube to operate using less than 2% of power required by conventional vacuum tubes and making it easy to power the unit on batteries.”

      My guess is it will require a much lower voltage to operate.

    2. Using those numbers, the new tube would use around 20mW.
      The CK512ax from 1956 uses 0.625V x 0.02A = 12.5mW just a tad bit over 1/2 of the new one. This tube also run at 40V. I have seen it in a couple of recent Japanese audio DIY kit with other modern parts running from 6V to 9V batteries.

      1. I can design a transistor amp to do that just as well. While it’s true that a vacuum tube amp can have a transfer curve that rolls off gradually (using a tube with a remote-cutoff style grid) and can be much simpler (by parts count) than an equivalent solid-state circuit, the solid-state circuit will always cost less than that one vacuum tube alone.

        And Jacques, yes, it’s pretty funny how much better “rich harmonics” sounds than “severe distortion”, even though both terms can be accurately applied to the same amplifier.

  4. Plate power consumption is relatively low. Taking as an example a typical 12AX7 circuit with a 300V supply, 150K anode resistor, 150V anode voltage, the power consumed in the plate circuit is 300V X 0.001 = 300 miliwatts. The two triodes and heaters in a 12AX7 run this way would consume 2.5W total. Unless this is intended for battery powered applications I doubt the power saving is the goal.

  5. I used to work at a 5KW AM transmitter station and the output tube was taller than me. I am not sure of it’s exact height as it did have a base with plumbing in it for the water cooling.

  6. AHA! I’ve actually been thinking about this for over a year. My inspiration was the technology from fallout, I was basically thinking “If the transistor was not standardized, we’d have already seen miniaturized vacuum tube technology on glass chips”. Today I remembered this and was like “Man, I wonder if that actually is possible” and then I see this article. So indeed, I am happy to see this idea ACTUALLY a reality!

    Might be too late to make much of a difference now with the advent of silicon over 60 years running, but hey! Awesome!

    1. Tubes are used in space applications, a lot, actually.

      They just typically aren’t the tubes most of you are familiar with, but rather klystrons, TWTs, CFAs and so on – i.e. microwave amplifier tubes.

      They are way more robust in space than solid-state, operate much more efficiently at microwave frequencies than solid-state and tend to fail gracefully or “die slowly” so you can get an idea of how long you can squeeze a little more life out if them, whereas transistors just fail hard.

      Solid state has to be extremely tiny to work at microwave frequencies, and the smaller you make the devices, the harder it is to get the heat out of them. Yet the higher the frequency and power you need, the more heat… The technology is gaining ground, but it’s still pretty far off.

      Tubes don’t suffer the same sort of problems. Slow wave structures in microwave tubes are rather large compared to the wavelength of the frequencies they operate at and so they can stand the heat.

      Multiple transistors operating in phase is one way to get the power, but they still make tons of heat and require a lot of power input.

      You don’t see many tubes on store shelves or in skinny jeans pockets or in PC’s, but if you check out behind the scenes of the backhauls, weather radar stations, airports, satellite systems, hospitals, radiation clinics etc. etc. where the real work that makes the world go round, you’ll still find plenty of “tubes” doing their thing…

      1. How about a UNIVAC that would fit under your bed? According to wkipedia UNIVAC I used 5,200 vacuum tubes, weighed 29,000 pounds (13 metric tons), consumed 125 kW, and could perform about 1,905 operations per second running on a 2.25 MHz clock.

  7. What many people fail to recognize is that the ‘tube sound’ has much more to do with overdriving the power stage than the pre-amp stage. There are a gazillion tube quackery products made by just slapping a 12ax7a type tube in it (remember the Aopen AXB-533?) and they are pretty much just crap (need a Portlandia ‘slap a tube in it’ scene). And there’s a reason some of the best sounding tube amps are only 30 to 50 watts if not less, so you can run them wide open in a studio without blowing a hole in the wall.

    While I think this is a neat idea, I don’t see it offering anything of value. If power efficiency is an issue, as others have pointed out the preamp tube effect can be pretty well simulated with a good fet based silcon design.

    1. I agree with you on the guitar amp power tube (and don’t forget xformer saturation!) component of the tube sound but as a guitar player who is also a recording engineer tube microphone preamps and compressors are still highly regarded by many in the field. Again, the “tube” sound is often a combination of the tube distortion combined with the subtle effects of “iron” in the circuit since most vintage tube units had output xformers and often interstage xformers as well. All that said, I WANT ONE OF THESE!!

      1. Excellent points – I agree with that 100%

        The problem with most of the ‘tubeified’ gadgets is they usually run them at low voltages and avoid the transformer “iron” because those parts are typically where the production costs are.

  8. Noritake is known for their history in Vacuum Fluorescent Displays and this really is essentially a VFD structurally, using the same fine monofilament cathode, grid, and anode, so my suspicion would be that these tubes would be designed to operate with the same basic plate voltage (12-24VDC) and be limited to very small currents due to the size of the elements. I did notice in the mockup that Korg posted (both sides) that they seem to have placed a small patch of phosphor on the plate, so I suppose that would fluoresce when the current peaks. Overall, completely awesome.

  9. Pretty cool and could be fun for a number of music projects. It will be fun to see how this develops.
    Agree with the others that say we are looking at Korg’s next gen feature for it next family or modern re-issue. I had missed announcement so thanks to the submitter on this one :)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s