How Tube Amplifiers Work
robrobinette.com
robrobinette.com
[1] yes, yes...I know...always use an ad blocker.
What may not be obvious is that modern tube amp designs are an evolutionary branch from 1930's technology, with only a little coming across from the transistor->digital tech tree. The amps of the 40s and 50s were pretty closely based on reference designs that came from RCA and other tube manufacturers.
Modern passive components (resistors, diodes and caps) are made to a far higher tolerance and are better understood, but tubes and transformers are a mixed bag. The older designs were somewhat overbuilt and can be more reliable or have tonal characteristics that are not available in modern parts.
Or do you mean why people who do period-authentic tube amps don't use SMPS? That's because tube-based SMPS is very complex, often as complex as amplifier itself, and needs unusual parts [1].
[0] https://www.sweetwater.com/store/detail/MicroTerror--orange-...
[1] https://www.righto.com/2018/09/glowing-mercury-thyratrons-in...
There are quite a few effects like this. In a modern design this would be eliminated, but sometimes “bad” is good :)
A guitarist who plays electrified isn't just playing the guitar; the entire signal chain becomes the instrument. Everything from the room to the fingers of the player alter the sound and how a person plays their instrument and for some even the temperature of the room makes a concrete, quantifiable difference.
Music appreciation is largely cultural as well. The history of music is full of people hearing sounds, becoming accustomed to them and reproducing them with a novel variation. This is exemplified by many recent genres like hip hop, rap, jazz, rock, folk music and so on. There were and are entire genres of music and specific artists that revolve around certain tools. For example the Sunn brand of amplifier, especially the Model T which is venerated by some subgenres of metal or Jimmy Hendrix and his Fuzz Face pedal (and his wah and octaver and amp, and .....)
Naturally, musicians seek to pay homage to and recreate the atmosphere and feel of a specific song, instrument, artist, genre or time period. Until fairly recently, modeling and digital tools had a lot of trouble replicating the sound and interaction of these vintage, analog circuits and even today the most straightforward way to achieve a specific style is often to simply buy or clone the old-school original instruments and equipment.
While digital modelling has come a long way, arguably surpassing most of the original equipment, the rarity, variation and uniqueness leads players to continually seek out the Real Deal in order to achieve an authentic style or sound.
An example of this entire idea is the DRUMETRICS collective, whose entire purpose is to write and record new and modern performances with original vintage instruments and recording equipment. Heres a link to one: "Pale Horse" https://www.youtube.com/watch?v=vZqoFf859Xw
SMPS have been somewhat problematic with audio circuits when they were new, especially when it comes to noise and "musicality". A overdimensioned toroid transformer with a rectifier is inefficient, but an extremely simple design which allows people without too much electrical engineering knowledge to get a decent result without expensive measurement equipment.
This is a bit similar to early PCB-use in guitar amplifiers. Back then some manufacturers did a shit job with their PCBs and since then guitarists think hand-wired is always superior to PCBs.
Guitarists are traditionalists and thus the amount of innovation in that space moves slower than elsewhere.
The key diagram is the one that shows the signal path through the amplifier. Input feeds grid, plate feeds next grid, final output is from plate. Everything else is supporting circuitry.
Note that between each stage there's a capacitor in the signal path. That's to block DC. If you want an amp that amplifies DC, each stage has to run at a higher voltage than the previous stage. The plate must be above the grid in voltage. This was a huge headache in tube computers, both analog and digital.
Transistor circuits don't have the increasing voltage problem. Outputs and inputs are in the same voltage range. That's because transistors are current gain devices, not voltage gain devices.
You can also stick a voltage divider (and probably some diode clamping) in there to pull the signal off of the plate down to a grid compatible voltage for the next stage if you're just doing digital computing. That was the most common setup I've seen in tube based computing. They tended to play pretty nice with the resistors needed for the plate current anyway so it wasn't that much extra RC constant delay.
I'm trying to keep my tube computer I'm building down to ~3KW, and that's probably the biggest actual constraint on design complexity.
This is also very misleading in that all this supporting circuitry AND the stuff not even shown, such as wires routing with respect to each other and with respect to the inside or outside of a metal case ALSO contribute. All this stuff contributes to basic functionality ("noise", "hum", etc) and to finer performance (frequency response, dynamic, distortion, crosstalk, etc).
It's easy to confuse the map for the territory, the schematic for the physics of the thing. And common electronics schematics abstract away much that does matter. Engineers and builders with some experience will pay attention to this without bothering to include it in the schematic.
Pay attention when following a magazine article for example: most of the time it will point out the why of several decisions. Why they placed this and that away from each other. Why these wires are routed this way...
The delivery style gets to some people (i.e. “I’m not ___ I just play guitar…”) but i find it absolutely fine.
I moved on from tube amps about 15 years ago and now really enjoy a variety of different solid state amplification stages with varying EQ and ‘dirt’ options at various places. Turns out a lot like were Jim’s Video goes.
And gone by the time I was old enough to be interested in electronics.
Nonetheless, my curiosity about them remained and I did eventually seek out books to understand how they worked. I have since built perhaps a dozen hi-fi stereo and mono-block tube amplifiers—some from kits, some from scratch. I've built a handful of guitar amps as well (even sold some as kits for a bit). Point to point, tagboard, PCBs…
Anyone that likes to tinker in electronics I recommend they try their hand at at least one tube project (probably an amp of some kind).
Only if they are aware of the voltages and current often associated with tube setups. One bad move can be painful, or fatal in some cases.
I used to work on guitar amplifiers, doing modifications on tube amps. Messing around with the internals demanded my focus, a level of attention most "tinkerers" aren't likely ready for. Not trying to gatekeep here, just suggesting it may not be something for "anyone that likes to tinker".
But by all means, you never touch it when plugged in, you use chopsticks if you need to poke/debug a live circuit (you keep one hand in your pocket, etc.).
for people who have not had much EE education, what is important about triodes and transistors is that they amplify. you can put a signal in (a signal like from a microphone responding to your voice), and put some power in (like from a battery) and these amplifiers can make an output "copy" of the signal which is more powerful/"louder" than the original.
from this basic function, everything that we think of as "electronic" flows. we would still have electric things like light bulbs, heaters, spark plugs, electromagnets, but basically just electric steam punk frankenstein machines, and nothing subtle. Amplifiers are termed "active" electronics; without them, we'd simply have passive electricity.
I didn't read this article because I already know how these things work, and the article looks extremely confusing, and I've already read my fill of explanations that don't explain anything and (not saying this is one of those) I don't want to even risk that again. it is very difficult to find explanations for how transistors work that make any sense at all.
That cannot possibly be true. Not knowing what exactly is going on with the charge carriers at the subatomic and quantum levels is not the same as not knowing how the amplifier works: like if we fiddle with the voltage at the base, we can influence the collector current, and all the rest.
What is true is that some early transistor designs of audio amps treated transistors like tubes: they featured a phase inverter transistor that fed two non-complementary push-pull stages whose output was combined by a center-tapped output transformer.
The excuse that well-matched complementary PNP transistors were not readily available at that time rings hollow, because it's possible to create an push-pull output stage with just NPN transistors. This is called "quasi complementary" (lots of search results for this).
Output transformers, if they have multiple taps in the secondary winding, do allow for different impedances. If the end users expect to be able to plug a 16 ohm speaker into a 16 ohm output jack and a 4 ohm into 4 ohm, then they will understand that kind of amp better.
The Owner's manual extols the advantages of using transformers for speakers and describes how to use the 70V output in conjunction with external transformers.
Quote:
For complex multiple-speaker arrangements that require many speakers and long runs of connecting wire, we recommend you use a line transformer (not supplied), available at your local RadioShack store.
[...]
There are several advantages to using transformers.
• You can connect speakers with different impedances without causing differences in output between the speakers.
• You can add or remove a speaker without having to recalculate the entire system’s impedance.
• You can reduce signal loss when you use speaker wire over 50 feet long.
LOL!
https://www.atlasied.com/speech-privacy-speakers?srsltid=Afm...
The cost is fidelity. Full-range audio transformers aren't cheap, so these systems usually make some compromises because your announcements or smooth jazz over the pasta aisle don't need to be true hi-fi.
Its cool technology. Most of the speakers have variable power taps, so you can run a bunch of them in parallel on a single line and control the actual volume as-needed based on where the speaker is deployed by varying the transformer tap on each speaker.
There's no need to LOL. It's useful tech.
By using a (~nominal maximum) 70v intermediate voltage, low-impedance PA systems become high-impedance. Compared to low-impedance systems at any power level, current through the wire feeding that system is reduced. Increasing impedance reduces the size of the copper wire required, and makes things easier for practical amplifiers to drive.
This is Really Useful in systems that may have dozens or hundreds of speakers distributed over a wide area.
Just to pick an example: Let's say we have 24 8-Ohm speakers to drive in a place like a grocery store. The system never needs to get proper-loud, so we'll budget 1 Watt for each speaker (24 Watts total).
If they're wired in parallel[1], then that's an impedance of 0.33 Ohms. Most power amplifiers can't drive that kind of impedance. It's also 8.5A of current, which isn't too daunting but is significant.
But if we add transformers and run things at 70v? Things get a lot easier.
Let's say we pick a 50 Watt amplifier just so we get some headroom instead of maybe running it at 100%, and that this amplifier's output power is rated at 8 Ohms.
That amplifier produces a maximum output of about 20v. Suppose we step that up to ~70V with a 1:4 transformer (which actually gives us a maximum of 80V, but again: headroom), and drive our grocery store full of 24 1-Watt 70v speakers with it.
With the same 24 Watts, our current on the speaker line drops from 8.5A to ~0.34A -- it goes from significant, to very nearly unimportant.
Our amplifier is happy: Rather than 0.33 Ohms, it sees an impedance of 12.75 Ohms.
That's an amplifier that runs cool and quiet, probably for decades. The wire can be small (18AWG seems common-enough in the grocery store overhead PA systems I've worked on; much smaller than that, and things become harder to deal with inside of ceilings). The speakers, which are all simply wired in parallel, can be replaced, removed, or added -- and also be individually adjusted in output level by changing transformer taps -- as time moves on without thinking much about the greater system architecture.
That's a lot of words, but in practice: 70v is easy. It works. We use this technique all over the globe: Some countries use 100v, and both 140v and 25v also exist as standards, but using transformers in PA-world is ridiculously common.
The usual method doesn't involve much thought at all: Add up the total power of the 70v speakers in Watts, pick an appropriate 70v amp that can drive at least that number of Watts, and send it. It's dead-simple to get right.
70v is not common in stage or musical instrument use, or in home hifi, but it doesn't have to be. It solves problems that don't exist in those spaces.
(And yeah, our grocery store needs a ~25:1 transformer on each 8 Ohm speaker. That's fine. 70v speakers that are intended to mount overhead and that include transformers with multiple secondary taps are very, very common and are still made every day in places like Chicago and Dayton[2] -- and again, you hear them all the time when you're out in public.)
[1]: We could use series-parallel arrangements to get around this, but those wiring configurations turn both arduous and easy to screw up, and loudspeakers have non-linear impedances so mixing-and-matching them in series as time moves on becomes problematic. Series-parallel isn't broadly practical.
[2]: Chicago: https://quamspeakers.com/product-group/ceiling-loudspeakers Dayton: https://fourjay.com/background-speakers/
Some colleagues of mine are looking at the same thing for cars.
since everything that happens inside a transistor is exactly what is going on in a quantum sense, you've described "not knowing how it works". You cannot understand a bipolar transistor without quantum effects, it's the thing that creates the transistor effect.
the theory of amplifiers you go on to talk about was well developed at that time because it's the same theory for vacuum tubes.
What makes the amplifier work and what makes the transistor work are separate concepts.
That's why understanding translates from tube circuits to transistors. A transistor circuit maybe an emitter follower, which has a counterpart in tube circuits known as the cathode follower. The cathode resistor creates local negative feedback similarly to an emitter resistor. Early op amps where tube circuits. They have the same differential input stage and the same basic theory of operation. You program their game the same way with resistors. The familiar Sallen-Key filter topology was first described with the help of tube circuits for reference, back in 1955. To undestand it, we don't even need the details like how amplifiers work at the component level except when we get into design parameters in which certain issues matter, like frequency-bandwidth product, or input offset current or whatever.
For clean sound, use compatible radio preamp tubes and bias the power tubes conservatively.
For distorted sound, use the lowest overhead preamp tubes you can find, and bias the power tubes as hot as you dare without them breaking within the hour. You can always change them after a gig, or between sets. :-)
E.g. to a metalhead, any tone that doesn't "chug" is useless, including something useable to a jazz fusion player.
My first real amp was a JCM800 2203 (technically a JMP "Mk 2 master model", which is just a cascaded JMP/Plexi, which Marshall then later re-released as JCM800 when their export deal expired...but I digress), and when I got into modding this website was my first real encounter with easy explained guides of the circuits.
But tubes aren't current amplifiers, they're voltage amplifiers, like FETs.
You can look at the "characteristics curves" of tubes (plate curves and transconductance curves), which tell the story of current against plate-to-cathode voltages for fixed grid voltages.
"Improved vacuum tube models for SPICE simulations" https://normankoren.com/Audio/Tubemodspice_article.html
The intractability of the Triode is part of the reason why the Pentode exists. And, you will note, the Pentode curves in certain modes looks a lot like your bog standard MOSFET.
This also discusses how the "constants" ... well, aren't. https://www.john-a-harper.com/tubes201/
With that said, a N type JFET is not a bad start. The main rules of thumb work: The grid draws negligible current. The tube will pass enough current from plate to cathode, to maintain a roughly constant cathode voltage above the grid.
Gives overview equations for MOSFET device simulations which are probably sufficient for most purposes in Section 3.5, and COMPLETE mathematical descriptions of the SPICE MOSFET implementation in Appendix A.3. Not for the weak.
I promise you it does not contain AC when unplugged :)
Just his list of 5E3 mods (Fender Deluxe) is awesome:
A reversal, which occurs in the vaccum chamber compresses electrodes, tagging battery terminal from the +/- amplifier schema AC electricity is transformed.
* Please don't suggest I install an Ad blocker.
I mean just in general it makes the web less awful. Webpages are so much easier on the eyes without all the crap they try to stuff in there.
And it can prevent malware, especially for those less tech-inclined.
And it means you use less data/bandwidth, since the blocker prevents the request from ever being made in the first place.
If you want to support a site, just buy a subscription or donate to them or something.