5/12/2026 at 4:04:50 PM
I think this is an odd article. It mixes together a variety of technologies that have little in common (gas discharge tubes, CRTs). It doesn't really say anything about the operation of vacuum tubes, their advantages, or disadvantages. And doesn't even really support its own thesis. The reign of vacuum tubes lasted for less than the reign of the transistor and is in no way unusual in the world of electronics.There are quite a few interesting stories to tell here. Probably the most interesting one is that transistors still underperform vacuum tubes in many respects that would matter to purists, but that don't matter in real life because we learned to compensate for it. Well, except for niche audiophile audiences who don't believe in negative feedback or digital signal processing and want a very linear amplifying component... that they then connect to op-amps, DACs, and ADCs on both sides because that's the only practical way to do it, but there's a performative tube somewhere in between.
Another cool story: there were some "integrated circuit" vacuum tubes!
by chromacity
5/12/2026 at 4:25:02 PM
There's really not benefits to vacuum tubes pretty much anywhere. The only place I can think of where they are superior (which may not be true anymore) is high power transmission in, for example, radio and radar towers.In all other applications transistors will be superior. Especially because any problem from a transistor can be fixed by adding more transistors until the problem is gone or imperceptible.
The audiophile purists are using pseudo-intelectualism to justify a superiority complex. They frequently fail double blind tests whenever push comes to shove. The most famous example of this was them being incapable of telling the difference between a coat-hanger and a premium cable.
by cogman10
5/12/2026 at 8:32:34 PM
> There's really not benefits to vacuum tubes pretty much anywhere. The only place I can think of where they are superior (which may not be true anymore) is high power transmission in, for example, radio and radar towers.Tubes are still fairly common in 62 dBm HF ham amplifiers, but solid-state amplifiers are available now, so it's only a matter of time there.
> In all other applications transistors will be superior. Especially because any problem from a transistor can be fixed by adding more transistors until the problem is gone or imperceptible.
This is often, but not always, true. E.g. parallel MOSFETs operating in triode mode are subject to thermal-runaway.
> The audiophile purists are using pseudo-intelectualism to justify a superiority complex. They frequently fail double blind tests whenever push comes to shove. The most famous example of this was them being incapable of telling the difference between a coat-hanger and a premium cable.
Tube amps often will be audibly different in a double-blind because many of them have high harmonic distortion (as compared to a transistor amp). Most people think this is a Bad Thing, but audiophiles call it a "warm sound."
by aidenn0
5/12/2026 at 6:45:45 PM
The LHC particle accelerator uses klystron tubes for RF amplification: https://indico.cern.ch/event/1559978/contributions/6664488/a...by nolist_policy
5/12/2026 at 5:16:05 PM
Eh. As far as instrument amps go, it's not about perfect fidelity. It's about color and distortion. You'd never say "perfectly white lights are the best for your living room." No one actually wants a perfect white light, people want some more yellow in there because it looks better. The goal isn't equal spectrum coverage or whatever. People like the non-linearity of tubes and that's ok.by tempaccount5050
5/12/2026 at 7:20:00 PM
The thing is, much like "perfectly white lights" you can mimic the non-linear behavior of tubes with transistor circuitry. On the extreme end you can integrate a DSP into the line to add a "vacuum tube" filter onto the sound in pre-amp.Thinking you can't do that is like thinking all LED bulbs must be 5000k and only incandescent can give that warm glow (Which, funnily that color was chosen to mimic gas lights before incandescents).
by cogman10
5/13/2026 at 5:03:10 AM
But when you do that, you are still holding the tube circuit as the reference model. And what you built is more complicated; it has components that have no counterpart in the tube circuit. (In the digital case, you have hundreds of millions of tiny transistors, which are switching full on or off.)by kazinator
5/12/2026 at 11:53:29 PM
No one said you can't do it. There's just no reason to do it. Tube amps are basically a standard and it's what the amp repair guys know how to work on and it's easiest to work on.by tempaccount5050
5/13/2026 at 7:39:29 AM
There are real reasons to do that, even if you like tube amps:- Tube amps need regular tube swaps. For a small guitar Combo a set of tubes (3× preamp + 2× power amp) can cost between 100 to 150 Euros. If you're a poor musician that is a factor
- Tube amps use heavy power and output transformers. If you ever had to carry an Ampeg SVT (36 kg for 300W) you will cherish the idea of a 0.25 kg 500W class D amp. If you're a touring musician that is a factor
- Tube amps are limited in terms of sound. While this could be seen both as a feature and a bug, if you want to recall the perfect settings for each song as a touring musician, tube amps are not the right choice unless you have a horde of roadies and technicians
This is why amp sims have become more popular in the recent years with people who play music for a living.
That being said, if you're in a simple band and you want the sound of a tube amp, a tube amp gives it to you pretty reliably without making you drown in options or giving you the feeling your gear is outdated every other year, since it is outdated since decades anyways. A tube amp also needs no firmware updates and requires no subscription and has less parts that can break.
by atoav
5/12/2026 at 5:57:34 PM
Photomultiplier tubes have a solid state counterpart [1] but there's still a lot of use for the vacuum tube version.by dguest
5/12/2026 at 7:13:33 PM
And Hamamatsu (and some others) still produce and sell photomultiplier tubes. The microchannel plate PMTs are pretty nifty things [1]. You can get single-digit picosecond time resolution out of them. [1]: https://www.hamamatsu.com/us/en/product/optical-sensors/pmt/...by HansHamster
5/12/2026 at 4:52:57 PM
I disagree. There is nothing in the digitally sampled and modelled world comparable to plugging your guitar into a hot over-driven tube amp and showing the feedback who's the boss. Pure analog transistors don't give that luscious even harmonic distortion and usually just clip like a meth-addled dog stylist in a poodle-grooming station.by bregma
5/12/2026 at 5:09:30 PM
As a guitarist with over 30 years of playing, and owner of many tube and non-tube amps, I disagree. Even experienced guitarist cannot reliably distinguish between transistor and tube circuits in a blind test. Having said that, if only the knowledge of playing a tube amp gives someone a better experience, even if its not empirically distinguishable, thats a perfectly valid reason to prefer it.by n_kr
5/12/2026 at 10:17:10 PM
Maybe recently, with impulse responses and Fractal Axes-like machines, you can get any sound from silicon. But in the 90's the difference between silicon guitarrists (Dimebag Darrell or Chuck Schuldiner) and tubes was clearly noticeable.Also, some 90's lower end amps with a valve in the prev sounded way better than similar priced amps that opted for pure solid state, at least for metal music and high distortion. For the clean channel the difference between them was minor.
by otherme123
5/12/2026 at 5:48:53 PM
Minor correction:An experienced guitarist cannot distinguish between "captured" amps, or amps which at their core simulate vacuum tubes at the software level. I definitely can't tell the two apart. However, I believe it is easy to distinguish a pure vacuum tube-based circuit from a JFET/MOSFET-based one.
There do exist vacuum tube replacements like the AMT 12AX7WS [1] or Jet City's RetroVales [2], but I would argue that the fact that they try to emulate tubes via transistors is a strong indicator that the natural circuits for both sound distinct enough for guitarists.
[1] https://amtelectronics.com/new/amt-12ax7ws/
[2] https://web.archive.org/web/20190803060713/http://www.robert...
by utopcell
5/12/2026 at 7:29:11 PM
Not who you are responding to, but I'd say the point I was making wasn't that the sound wasn't different (though, the differences are almost certainly not large enough that most people can tell the difference). But rather that if that exact sound profile is desirable, it's easy to reproduce with transistors alone.The two products you list are proof of that.
by cogman10
5/12/2026 at 9:09:14 PM
I see your point, though strictly speaking, the two products I mentioned mimic 12AX7 tubes, which are preamp tubes. I'm not aware of E2E designs that also mimic power tubes.by utopcell
5/13/2026 at 6:03:09 AM
But the transistor circuit designers have gone out of their way to put in "tube like" behaviors which result in extra circuits and components that don't appear in the tube amp, and which would not appear in a "high fidelity" transistor amplifier.One very common trick that has appeared in transistor guitar amps from the 1970's and onward is current feedback. The return terminal of the speaker doesn't go straight to ground, but feeds a tiny current-sensing resistor, like 0.1 ohms. Signal from the current-sensing resistor is combined into the negative feedback. This gives the amp a nonzero output impedance against the speaker, which changes the frequency response: more power is driven at the frequencies where the speaker impedes more, like its resonant frequency (about 70 to 100 Hz for a 12" guitar speaker) and high frequencies (due to voice coil inductance: rises from about 2 kHz up or something like that).
This doesn't reproduce everything that happens with tubes, but it goes a long way.
I built a circuit like that into an off-the-shelf amplifier, with a switch. I tell you, whenever I switch that off, it's not long before it goes back on again. Without the current feedback, it's sounds blatty/tubby and lifeless. It's not just the frequency response, because even if we dial in a similar EQ curve before the amp to eliminate the difference, it's not "it".
The TubeWorks people have an interesting design in the MosValve 500 amp. Rather than using current feedback, that amp places the MOSFET output stage outside of the negative feedback loop. Negative feedback is drawn from the VAS (voltage amplification stage) before the output stage. That means that the speakers will see the impedance of the MOSFETs. Plus the supposedly "tube like" overdrive characteristics of the power devices will come into play when that thing is cranked. Because they are outside of the NFB loop, it will just be soft onset clipping. Here is the important thing: unusually, the VAS and the output stage are on separate power rails, and those of the VAS are a significantly higher voltage (+/- 93V versus +/- 71V). So it is hard to make the VAS clip; it requires a much higher signal than what it takes to make the output stage go into progressive clipping. When an output stage is included the feedback loop, like in almost every amplifier out there, the amp is perfectly linear up to the limit, and then clips really hard.
The schematic: https://img.photobucket.com/albums/v202/xr3drumx13/MV500Sche...
Tube amplifiers "do their thing" without extra gimmicks like this, though.
by kazinator
5/13/2026 at 7:10:46 AM
What you say is true if we operate gear within specs. In guitar amplifiers when tubes are overdriven tubes, they still matter. Only in the past decade (circa) digital guitar amp simulations started to become sufficiently advanced to deliver convincing organic sounding results on that front, with DSP-based hardware that is suitable for the stage.But if all you want is the sound of an overdriven tube amplifier, getting yourself an overdriven tube amplifier is still the least complex option technically, in terms of parts count, repairability, etc. It is a primitive technology, but it is also well known and works. It also has no menus, minimal settings and doesn't need firmware updates or registered accounts.
Personally I use both Class D and transistor based audio amplifiers for my musician life, since those are much lighter to carry for the same power and work for the type of sound I want.
If we talk Hifi unless they also want to overdrive the sound (they usually don't), tube amps are useless. The distortion figures of good transistor amps are much better, class D is making massive leaps in the recent years and has reached indistinguishable quality levels for a while now. These people should focus on room acoustics.
by atoav
5/12/2026 at 9:17:17 PM
The benefit is that they are incredibly simple to build.by anikom15
5/13/2026 at 6:41:04 AM
I bought new vacuum tubes just last week. For guitar amplifiers vacuum tubes are still pretty common. If you want a simple, horribly inefficient circuit that sounds good with an electric guitar vacuum tubes are great, mostly because they overdrive a bit more gracefully than transistors.For my daily band life my amp of choice is a Roland JC-120 for guitar (transistor amp) and a Eich T-500 (class D) for Bass.
Can't wait for class D to catch up to tubes, I really don't dig having to carry 30kg amps around.
by atoav
5/12/2026 at 6:18:53 PM
a non-trivial portion of the guitar-amp world is still very much set on tubes, even as amp simulators get closer to "the real thing".a cool recent development i've been following is the Octal by Verellen Devices (created by an awesome musician who also built some highly coveted boutique all-tube amps): https://www.verellendevices.com - it's basically designed to replicate the push/pull of power tubes in a solid-state package to push extremely loud guitar cab speakers. they seem to impart their own sound signature but still sound really really good, especially compared to a lot of solid-state guitar amps.
by GuinansEyebrows