During a global semiconductor shortage, is it time to look at the most recent innovations in thermionic valve technology?

   You are probably wondering why on Earth I am writing a modern technology article about thermionic valves. Perhaps for some readers, you may not even be familiar with the often-romanticised predecessor to the transistor, whose related technologies have become integrated into almost every aspect of our modern world. So that we are all on the same page, I think it is best to start this thought piece with a brief description of the thermionic valve and its similarities to what followed.

  Thermionic valves, also commonly known as vacuum tubes, amongst other names, were first invented by Sir John Ambrose Fleming in 1904 as an emerging technology that could be used to control the flow of electricity through electronic components, hence their nomenclature, valve. As their other name suggests, a thermionic valve is, by its construction, an evacuated glass tube. Inside this small vacuum chamber are a set of electrodes with pins emerging from the enclosure for connection to the rest of the circuitry. Though there are diode valves, the electrodes of the triode valve feel most relevant when comparing them to a transistor. Triode valves consist of a heated cathode, a grid, and an anode, each separated by a small vacuum gap. When the cathode reaches temperature, the heating often supplied by an additional element, thermionic emission causes it to give off electrons. These electrons then travel across the vacuum and are received by the anode, completing the flow of electricity across the device. Since the electrons can only flow in one direction because of the emission process, this two-electrode valve is our diode valve. By placing a third electrode between the cathode and anode – a grid – one can now mediate the flow of electrons across the device by applying a voltage to it; a negative voltage on the grid will slow the electron flow to the anode, and a positive voltage will increase said electron flow. This operation is not unlike that of a Field Effect Transistor (FET) which has a source, gate, and drain; much like with the triode valve, a voltage applied to the gate mediates the current flow between the source and the drain. Like solid-state devices, valves evolved to have complex architectures, such as multiple grid devices – tetrodes and pentodes – and to have multiple devices in a single package; dual-triode valves, i.e. two triode electrode configurations within a single enclosure, are still widely used to this day in audio applications.

  Now that we are all familiar with what I mean when I say valve, let us quickly examine why the development of the transistor in the late 1940s eventually led to the demise of valves as the preferred technology. We can already see that triode valves and FETs can have a very similar usage in an electronic circuit, so what was it about transistors that caused this decline in valve use?

  Simply put, the answer comes in the greater efficiency of transistors in terms of energy consumption, ease of manufacturing, and overall durability. Since a heating element is required for thermionic emission in valves, much of the power used to make the device work is ‘wasted’ in this heating. Valves also require much higher voltages to operate, thus the transistors that were built to do a similar job were far more energy efficient; they did not require heating or high voltages. Operationally, valves also require a warm-up period before they are at a suitable temperature to work, whereas transistors are ready to work as soon as a device is turned on. When it came to their construction and size, valves stood no chance in the supremacy of transistors. Transistors were smaller, more robust, and easier to produce. Being a glass tube with suspended electrodes in vacuum, valves are far more fragile and difficult to manufacture. Since the emission of electrodes also ‘burns away’ the cathode, valves additionally have a definite operating lifetime before requiring replacement in a device. As they found themselves being employed in an increasing number of applications, it was clear that transistors were far more suitable for a greater range of smaller and more field-intensive devices and technologies than valves in their current form could ever be.

  But this does not mean to say that valves did not have their advantages over the transistor. Despite its difficulties, vacuum will always be a readily available resource, rather than the germanium of older transistors. The move towards silicon in the mid-1950s was certainly useful in ensuring an abundance of raw materials to create new devices for the foreseeable future. Another benefit is that a broken or damaged valve is easier to replace when compared to transistors installed as part of a circuit board and soldered in place. Since valves were known to have a limited lifetime, they were typically installed with a socket; changing one is not much more difficult than changing a light bulb. Perhaps the greatest advantage of valves is their resilience to heat and, more importantly, surge voltages and electrostatic discharge. While high currents can both damage the semiconducting layers of solid-state devices and the minute metal tracks themselves, the vacuum medium of valves does not have this issue, and it would take far higher surge currents to damage the relatively macroscopic metal elements of a typical valve. This makes valve technology far more resilient, though not immune, to electro-magnetic pulses (EMPs) that can easily break solid-state electronics.

  Though now almost completely surpassed by solid-state technology, valves used to be found in most places where we would expect transistors to sit today, and valve computers using valve logic have certainly existed in the past. The first automatic digital computer, the Atanasoff-Berry computer (ABC), used a CPU of over 300 valves in order to perform arithmetic calculations; as previously mentioned, valves operate similarly to FETs, and MOSFETs are the basis of the digital logic devices that run our computers today. That said, the ABC weighed 320 kilograms and presumably gave off a lot of heat from the plethora of valves in its CPU, so we should all be thankful that transistors were a technology that could be miniaturised to the point where a modern mobile phone is more powerful than a mid-90s home PC and does not burn a hole in your pocket…usually.

  But I did say ‘almost completely surpassed’. Due to their subjectively superior behaviour when amplifying audio signals, valves remain widely used in the audiophile and musical equipment industry. This is most likely due to the historical aspects of the technology as a key component in the development of popular music, which has become a huge industry in itself. The sound of valves is the sound of rock, and so long as rock lives, valves will continue to remain a staple technology, albeit in a tiny sphere when compared to its original applications.  

   While there has been some miniaturisation of valves in the past – such as the sub-miniature vacuum tubes used in hearing aids and nuvistors used in MIG-25 fighter jets – the most notable development in the technology since the 1970s happened in 2015 when Japanese musical instrument manufacturer KORG INC collaborated with Noritake Co., a display manufacturer, to create the Nutube, a miniaturised dual-triode valve designed for audio applications but contained within a sleek, robust package suitable for direct mounting on a PCB. The device still exhibits the cathode-grid-anode structure of the traditional triode valve, but by applying technology from Noritake Co.’s vacuum fluorescent displays, the geometry has been modified, dramatically reducing the device’s size and power consumption when compared to previous valve technology.

  A quick rundown of the Nutube’s features reports that the device uses less than 2% of the power and occupies less than 30% of the volume of a conventional vacuum tube. Unlike legacy devices it has a much lower thermal output and 30,000 hours of continuous operating life, allowing the device to be attached directly onto a high-density circuit board with the confidence that the component will not need to be replaced. A direct comparison with a traditional dual-triode valve like the ECC83 (12AX7) seems a little rash as the trade-off in size and required power makes the Nutube’s operating values seem small, but with a working temperature at between -40 and +85 degrees Celsius and an anode voltage operating range of between 5 and 80 V, it is not clear that the Nutube punches that far below the weight of the 100 V minimum typical anode voltage of the legacy device. The typical anode operating currents, however, are around 40 times smaller for the Nutube, but this may simply be the efficiency of the new technology on show and not an indication of the fragility of the device to the high surge currents and voltages that give traditional valves their electrical resilience. Unlike the ECC83, the Nutube’s cathodes do not require an additional heating element, illustrating the much cooler operating temperature of this new technology.

  So far as I have been able to find, Nutubes (model number Nutube 6P1 to be precise) have only been used in commercial audio applications from KORG and their affiliates, but there is certainly some interesting variety in what circuits the Nutube has been used in: As expected, the technology is primarily being used in the preamp stage of guitar and bass amplifiers to get the coveted sound characteristic of valve amplifiers in a smaller, lighter package. Nevertheless, I have not seen any power amplifier applications for this technology; it appears even the Nutube headphone amplifier uses an op-amp for its output stage. Where things get interesting is that Nutubes have also been used to make low frequency and audio frequency oscillator circuits, the former being used to drive a tremolo effect (the relatively slow modulation of an audio signal’s volume) and the latter being used as an audio source in a synthesiser.

  If Nutubes can be used in this variety of circuit architectures, could they perhaps be used in the wider variety of applications outside of the musical sphere that valves used to occupy, where the advantages of valve technology are preferable over the transistors that surpassed them? For a technology that has been around for seven years now, I feel that Nutube has been a little underutilised and looked over, simply because only audio enthusiasts seem to care about valve technology, but I would be very intrigued to see how this new breed of valves stacks up against EMPs and whether the Nutube architecture can be expanded to create multiple-grid designs or ones suitable for power amplifier applications.

  In terms of EMP countermeasures – and I will preface that with stating that I am talking about technology that can simply withstand EMPs rather than prevent malicious EMP events – if Nutube technology could be shown to have a similar resilience to traditional valve technology, they could be useful in future defence equipment or equipment designed for high electro-magnetic noise environments. Though I would hope that current solid-state transceiver technology for military applications were fully resilient to EMP attacks – the AN/PRC-77 portable transceiver used during the Vietnam War supposedly outcompeted similar valve-based units of the era – perhaps Nutube technology could help provide more portable solutions, even if encryption and decryption is performed within a separate module; I do not believe that Nutubes can replace the advanced microchip technology afforded to us by transistors. I will admit that the fiction writer in me does wonder how far one could go with this technology, such as building an EMP resilient computer the size of a small apartment in order to perform simple calculations if, for whatever reason, the bulk of solid-state technology was deemed unusable or unobtainable in future. And if the idea of a computer the size of a small apartment seems silly, you need only think of a server room or the original valve computers to know that people will and still do dedicate huge spaces for a computer or set of computers that need to perform tasks they deem important. 

  If all this talk of EMPs seems quite doom and gloom, the reason I talk about Nutube in this capacity is unfortunately because of the current times and the news that has circulated over the course of 2022 regarding legacy valve technology and semiconductors. With the start of the Russia-Ukraine conflict, valves, most of which are made in Russia and China, found themselves on the list of banned exports. While there has been some loosening of restrictions for the global audio market, all-in-all, there are feared shortages of valves for the foreseeable future, not because there are none available, but simply because exports are curbed. Effectively, Russia is hoarding the technology, and, again, the fiction writer in me cannot see why the reason is simply to make audiophiles sad. This action and the growing China-Taiwan tensions seem to me to be a red flag for global access to both technologies, after all, Taiwan produces 92% of the World’s most advanced semiconductor chips.

  While I sincerely hope that Nutube does not become an emergency technology, it seems worth looking into this relatively recent innovation to see what it could offer certain applications that solid-state technology does not. Realistically, it is unlikely that the specifications of the Nutube 6P1 will be suitable for all ideas that may come to mind, but if there is demand, perhaps KORG and Noritake will invest in the R&D to further miniaturise the design and/or expand on the current dual-triode architecture to get us some power pentodes, for example. If someone out there can make a simple Nutube-based computer that can add 2 to 3 and get 5, that would be a great starting point for if the technology ever needs to head in that direction, but if the Nutube ends up simply being used for some really warm-sounding EMP-proof, or not, radio gear, I would still count that as a win for valves making a long-awaited comeback. 

As expected, the Nutube evaluation boards are heavily geared towards audio amplification and signal processing, and since there are a number of solid-state components as part of the circuitry, they will not be suitable for EMP testing applications in their current state. However, that does not mean there is nothing to play with off the bat. The latest Nutube evaluation board (NEB-2) sports the following features and specifications:

·        4 selectable levels of harmonic distortion: approximately [1]: 0.015%, [2]: 0.1%, [3]: 0.3%, [4]: 1% – Independent switching for left and right channels

·        Signal bypass switch for easy switching between signals “with” and “without” Nutube processing

·        Stereo output volume control

·        Bias adjust for left and right channels

·        Max Input Level +2 dBu – Measured at 6 V BIAS voltage

·        Max Output Level +12 dBu – Measured at 6 V BIAS voltage

·        Max GAIN +10 dB – Measured at 6 V BIAS

·        Frequency response 20 Hz – 40 kHz

·        SN ratio 100 dB

·        Input RCA x 2 (Input Impedance 10 kΩ)

·        Output RCA x 2 (Output Impedance Less than 50 kΩ)

·        Power Supply DC 12 V AC adapter or External power supply (+7 to +20 V, Recommend: 12 V)

  With ipXchange the new platform that it is, we have not yet been able to confirm whether we can supply this board – we need to approach KORG for a start – but perhaps with enough interest from our audience, a collaboration will become a possibility in the future. Regardless of whether the Nutube finds a suitable place within the ipXchange platform, I thought it was an interesting technology. I hope that you have found this article interesting and a good indication of the type of content that would be useful to you. Obviously, I understand this was a very niche topic to start off with, but let us hear what you have to say and what kind of topics you would like to hear about in the future!