A simple circuit with an optocoupler creates a “tube” sound

[FSFX]

Sounds tube amps produced all along.

I have a Wurlitzer jukebox from 1954 with a tube amp. It is far from perfect in reproducing the sound.  As well as being far from perfect in reproducing the true sounds of the 45 records, it also adds noise and hum. That is just like all of the other tube amps from that era. Nostalgic sound but I prefer the purer sound of more modern amplifiers.

[BubbaFet]

Even with a tube based circuit, the range and comparative level of harmonic distortion can change significantly depending on the value of the cathode resistor and the details of its bypassing as well as the input signal level.
This is pretty obvious as these things affect the operating conditions of the tube in the circuit.
I spent quite some time experimenting with this and analysing the distortion of different arrangements when designing and building a 'guided-missile' pencil tube boost and overdrive pedal a couple of years ago.
When it comes to 'tube sound' and harmonic distortion it really just depends on circuit configuration anyway as a Class A preamp or single ended output stage will create a whole lot different distortion to a push-pull type of output stage.
There are many other factors involved as well as have been extensively documented over many decades.
   

In my 'distorted' view  , the best explanation of 'tube sound' and its mechanism is to be found in 'Radiotron Designers Handbook 4th Edition'.

http://www.tubebooks.org/Books/RDH4.pdf 

Specifically...  Chapter 14 - Fidelity and Distortion, starting on page 643 of the pdf.

--------------------------------------------------

Meanwhile, Chat-GPT3.5 offers up this response...

me: What exactly is 'tube sound'?

ChatGPT: "Tube sound" refers to a distinctive audio quality or character associated with vacuum tube amplifiers, also known as valve amplifiers. Tube amplifiers use vacuum tubes (also called valves) to amplify audio signals. These tubes are composed of a vacuum-sealed glass envelope containing various elements and electrodes.

The term "tube sound" is often used to describe the unique sonic characteristics produced by tube amplifiers. It encompasses several qualities, including harmonic distortion, compression, and a warm, smooth tone. Here are some key aspects typically associated with tube sound:

1.) Harmonic Distortion: Vacuum tubes tend to introduce harmonic distortion when amplifying signals. This means they add harmonics, which are additional frequencies that are multiples of the original signal. This distortion can impart a pleasant, euphonic quality to the audio, often described as "warmth" or "richness."

2.) Compression: Tube amplifiers exhibit a natural compression effect. When the input signal exceeds a certain level, tubes gradually saturate, compressing the dynamic range of the audio. This compression can add a pleasing smoothness and sustain to the sound, particularly in the context of guitar amplifiers.

3.) Soft Clipping: When a tube amplifier is driven into overdrive, it exhibits a form of soft clipping, which differs from the harsh clipping produced by solid-state amplifiers. Soft clipping introduces a more gradual and smoother distortion, often preferred by musicians and audiophiles for certain genres of music.

4.) Harmonic Content: Tubes tend to emphasize even-order harmonic content, particularly second-order harmonics, which are perceived as adding warmth and richness to the sound. Solid-state amplifiers, on the other hand, generally produce odd-order harmonics, which can sound harsher to some listeners.

It's important to note that the perception of "tube sound" can vary among individuals, and different tube amplifiers can have unique sonic characteristics. Additionally, modern solid-state amplifiers and digital modeling technologies can emulate some aspects of tube sound, blurring the line between tube and solid-state audio quality.

[FSFX]

QUOTE
4.) Harmonic Content: Tubes tend to emphasize even-order harmonic content, particularly second-order harmonics, which are perceived as adding warmth and richness to the sound. Solid-state amplifiers, on the other hand, generally produce odd-order harmonics, which can sound harsher to some listeners.
END QUOTE

I don't think that statement above is actually true.
It just shows that things like Chat-GPT are just as likely to give wrong answers as right ones, typical of the 'garbage in - garbage out' that computers are known for. Don't trust AI to tell the truth. It is based on a myth that has been around since the early days of solid state amplifiers that used certain circuit configurations such as the typical push-pull output stages. It can be proven quite easily that simple solid state amplifiers can be designed to give predominantly second harmonic distortion, just like tube amplifiers.
Further more, the square law forward transconductance characteristics of JFETs make them ideal for generating second harmonic distortion.   

[ElectricDruid]

Yep, anything that regurgitates what it read on the internet is going to be at least 95% full of rubbish.

Not that I'm blaming ChatGPT - that applies to humans too!

[Rob Strand]

Further more, the square law forward transconductance characteristics of JFETs make them ideal for generating second harmonic distortion.   

And we know that doesn't automatically give you a "tube sound".

IMHO, there's two tube sounds:

Clean: which isn't that much different to any clean sound, including opamps. 

Overloaded:  this is where things really change.   The changes are more than just second harmonics.  In fact a push-pull tube output stage has tube sound and very little even harmonics**.   Some key points are how tubes clip at low VPK, especially pentodes,  and the softness of the GK diode when the gate is forward biased.  The GK diode is *very*soft and difficult to emulate with solid state.


** A caveat is a device with second order harmonics on a sine-wave will produce a particular intermodulation spectra with a complex tone.

[Ben N]

IMHO, there's two tube sounds:

Clean: which isn't that much different to any clean sound, including opamps. 

Overloaded:  this is where things really change.   The changes are more than just second harmonics. 

But to me - and I can't begin to explain this in technical terms - the really interesting region that makes tubes special is where there these two flirt & cross over. I agree that clean-clean doesn't really reveal anything special, almost by definition. Linearity just doesn't allow for much variation by definition. But where a little bit of saturation starts to happen and picking dynamics start to have this outsize effect on how the amp responds -- that's where the magic is and where tubes shine. I don't really play heavily distorted amps -- if I'm getting that heavy, it's with a fuzz -- so I don't really appreciate what's happening in a modern high-gain amp from a player's perspective, and maybe there's something special there with tubes, too, though as a listener, it all kind of sounds like a distortion pedal with good eq to me. But the edge of breakup dynamics is where tubes are still king for me.

[teemuk]

ChatGPT: "Tube sound" refers to a distinctive audio quality or character associated with vacuum tube amplifiers, also known as valve amplifiers.

Keyword: distinctive.

1.) Harmonic Distortion: Vacuum tubes tend to introduce harmonic distortion when amplifying signals.

Any non-linear circuit does.

2.) Compression: Tube amplifiers exhibit a natural compression effect.

This is natural to transistors too.

3.) Soft Clipping

Is not inherent to tubes. Some tube circuits soft clip, some don't. Solid-state circuits can also soft clip.

4.) Harmonic Content: Tubes tend to emphasize even-order harmonic content, particularly second-order harmonics

Again some do and some don't. Solid-state circuitry can also be designed to emphasize 2nd order.

It's important to note that the perception of "tube sound" can vary among individuals, and different tube amplifiers can have unique sonic characteristics.

So much for "distinct" then.

This response reads like copy & pasted amateurish tube advocate websites, or the super cluttered Wikipedia article of the same topic. I honestly expected a bit more from AI but it obviously just mishashes various cliches without actual abstract or critical thinking to weed out facts from falses.

Before it cracks up the secret of "tube sound" (and learns to draw hands) I'm not worried it develops self-conscience and destroys the mankind.

[Steben]

There is no difference between amp harmonic distortion and amp harmonic content. Ild say the guitar has content the amp adds distortion.

[FSFX]

It is interesting to read the actual original article that started this all off and see that there are quite a few disputable points.

https://habr.com/ru/articles/697508/

Here is a quote directly from the article.

QUOTE
I invited the sound engineers I knew to listen to the sound. We listened to the audio of the narrowband channel and full-fledged tracks. That's when my attention was drawn to the similarity of the sound of the processed audio with the sound of tube amplifiers. In the process of working on this scheme, I looked through various publications, including patents, among which I found patents of the American company Aphex for a device called Aural Exciter. The principle is almost the same, only instead of a multiplier, they have a cascade on an offset diode, generating higher harmonics, which are added to the original signal. Then I began to look for other ways to get such a sound.

There are many circuits designed to produce a sound similar to "tube" and they are all reduced to two different types. The first are circuits with limiters on diodes that provide "soft" clipping of the signal, and the second are circuits on field-effect transistors that are not included in the general negative feedback circuit. Both types have, in my opinion, a common drawback, which is that these schemes, due to the fact that the transition characteristic is described by a logarithmic equation, generate odd harmonics, which are known to adversely affect the sound quality.
ENDQUOTE

It is interesting that the author attributed a logarithmic equation and odd harmonics to field-effect transistors. Certainly true for diode clippers but definitely not for JFETs which have a second order (square-law) equation and generate even harmonics.

FWIW, the original article mentions a circuit that uses a four-quadrant multiplier to generate additional harmonic content. This multiplier will act as a frequency doubler, just like many octave type of circuits and produce mainly even harmonics. The original article is well worth reading to see the subject of this thread in context. The version of the article I found was in Russian so needed to be translated to read it.   

[Clint Eastwood]

It makes more sense to me, and I think this is Teemuk's point,  to speak about "tube amplifier sound". The use of high voltages, giving lots of headroom, little or no global negative feedback, and the output transformer coupling alone makes a tube amp sound different from a solid state amp.

[Steben]

Personally I never understood the devotion to "headroom"...
Low headroom means too much gain. It's all relative.

[FSFX]

Personally I never understood the devotion to "headroom"...
- - - -    It's all relative.

Me neither. Many modern audio devices run at less than 5 volts. I think 'headroom' and 'dynamic range' get confused a lot.

[Clint Eastwood]

I should not have used the word headroom, but it seems to me there is something like 'high voltage' sound. Or maybe I am just imagining it?

[Steben]

I should not have used the word headroom, but it seems to me there is something like 'high voltage' sound. Or maybe I am just imagining it?

Frankly, still not quite sure what you mean. Triodes usually clip harshly either way whether on 100 or 300V.
Power amps in class B tend to squish but this can be emulated.

Many tube guitar amp DO sound and feel great. Let that be clear 
Yet question is whether this is device based or topology based.

Do fireflies and other 1W amps sound great? They do not have sag for example yet have push pull structure.
Ss rectifier high power marshalls have not that much sag nor cathode bias shift. Though 1974 has it. Ac30s have shift but hardly a sag. Tweed deluxes have shift, sag and some grid blocking. These are called downright fuzzy often...
And sorry ... again .... I adore the t dynamics power amp emulation. Analog solid state.

[amptramp]

Just to explain the idea that a square-law transfer function leads to the generation of second harmonics, use the trig formula:

cos(2a) = 2cos²a - 1

where a is an angle.  But if we take 2*pi*f = a to be an angle, we can back into the frequency domain from an angle.  The 2a in the first equation is a double angle that corresponds to double the frequency.

Thus, FET's with their square law transfer function should generate a lot of second harmonics.

[ElectricDruid]

Just to explain the idea that a square-law transfer function leads to the generation of second harmonics, use the trig formula:

cos(2a) = 2cos²a - 1

where a is an angle.  But if we take 2*pi*f = a to be an angle, we can back into the frequency domain from an angle.  The 2a in the first equation is a double angle that corresponds to double the frequency.

Thus, FET's with their square law transfer function should generate a lot of second harmonics.

In general, transfer functions that are mirror-symmetric generate even harmonics. Ones that are rotationally symmetric generate odd harmonics. Non-linear functions that are neither produce a mixture of odd and even.

http://musicweb.ucsd.edu/~trsmyth/waveshaping/waveshaping.pdf

[BubbaFet]

ChatGPT: "Tube sound" refers to a distinctive audio quality or character associated with vacuum tube amplifiers, also known as valve amplifiers.

...

This response reads like copy & pasted amateurish tube advocate websites, or the super cluttered Wikipedia article of the same topic. I honestly expected a bit more from AI but it obviously just mishashes various cliches without actual abstract or critical thinking to weed out facts from falses.

Before it cracks up the secret of "tube sound" (and learns to draw hands) I'm not worried it develops self-conscience and destroys the mankind.

Methinks that you may have just put yourself at the top of the 'A.I. KILL LIST' for writing this 

[Rob Strand]

Just to explain the idea that a square-law transfer function leads to the generation of second harmonics, use the trig formula:

cos(2a) = 2cos²a - 1

where a is an angle.  But if we take 2*pi*f = a to be an angle, we can back into the frequency domain from an angle.  The 2a in the first equation is a double angle that corresponds to double the frequency.

Thus, FET's with their square law transfer function should generate a lot of second harmonics.

The book Microelectronic Circuits by Sedra & Smith actually derives the second harmonic level for a JFET and gives a formula.  (Well at least the earlier editions.   Stuff seems to get edited out of books these days.)

[samhay]

I was playing with optocoupler delays a few years ago.

https://www.diystompboxes.com/smfforum/index.php?topic=116243

The topology was different - the phototransistor was set up common collector, and the current was low as I was trying to drive it using distant LEDs. In any case, you could easily get distortion by setting the LED current such that the transistor wasn't biased sensibly. It can't have been too earth shattering though, as I never built one.

One thing I found puzzling was that it had voltage gain, despite being set up common collector. You can see this in LTspice sims too, so it must be a known property of the optocoupler. Any idea what's going on there?

[Eb7+9]

One thing I found puzzling was that it had voltage gain, despite being set up common collector. You can see this in LTspice sims too, so it must be a known property of the optocoupler. Any idea what's going on there?

common collector (or emitter) implies common to two circuits ... here, the circuits are decoupled

so, voltage in >> LED current >> photons/lumens >> base current >> hFE times base current times emitter resistor >> emitter voltage

potential for overall voltage gain production is there provided product of all transfer ratios amounts to something greater than unity


the bipolar transistor is working as a current amplifier here, and not a voltage follower since there is no voltage source at the base