# What are tube amp harmonics?

Larry F
April 27th, 2007, 08:56 PM
I'm very unknowledgeable about tube amps, but understand harmonics as an acoustical concept. I know that for a frequency f, the harmonics are 1f, 2f, 3f,...In discussions about the characterstics of tube amps, I hear a lot about some aspect of design or some component producing harmonics. Does this mean that if I play a guitar note with a fundamental frequency f, that the harmonics 1f, 2f, 3f,... will be either amplified or generated? And will these frequencies change when my fundamental frequency changes? How about if I play a sine wave through the amp? (Recalling that sine waves are devoid of harmonics.)

Or are what people call harmonics simply higher frequencies?

Or are these harmonics of a fixed frequency? So the amp would be acting as a resonance system.

Signed, puzzled

Bogo
April 27th, 2007, 09:03 PM
Harmonic distortion is probably what you should research.
http://www.eqmag.com/story.asp?sectioncode=36&storycode=116

dibber124
April 27th, 2007, 09:29 PM
When an amplifier distorts it produces harmonics at the intervals you mentioned. Most tube amplifiers have a higher percentage of even-order harmonics (especially 2ND harmonic) than a solid state amp. Even order harmonics are more euphonic (ear-pleasing) than odd-order harmonics. The more even -order harmonics present in the audio spectra, the more pleasing the sound.
In addition, the output transformer in tube amps tends to roll of the transient peaks and smooth the sound somewhat. The combination is responsible for the well-loved "tube sound".

Larry F
April 27th, 2007, 10:03 PM
Thanks. This inspires more questions:

1. If I play a sine wave into the amp, will the output contain harmonics?
2. Are the harmonics generated?
3. Are the same harmonic numbers present for all fundamental frequencies of my input?
4. Even-order harmonics comprise the harmonic series, 1f, 2f, 3f, 4f,... of the frequency 2f. That is, the even-numbered harmonics form a complex wave that is an octave higher than the original signal. Why don't we hear this higher note?
5. The article linked says that even-order harmonics are more pleasing to the ear than odd-order. But in my experience, odd-order harmonics sound very nice. They are similar to a square wave or triangular wave (depending on amplitude and phase) and are reminiscent of a clarinet, a very warm, hollow, throaty instrument.

kp8
April 27th, 2007, 10:28 PM
1. If I play a sine wave into the amp, will the output contain harmonics?

No.
An amp with reasonable headroom and no processing with give you
a pretty good sine at the output. I gig with a pair of oscillators as part of my rig. (don't ask) That sine will be louder or softer depending on the freq response of the amp. Guitar amps are not very hi-fi and flat. Nevertheless, if you scope your input and output you will see the output is very close to the input (sine).

2. Are the harmonics generated?

If you have distortion then the squash you would see in the time domain would result in alteration in the spectra and some additional components. But in a clean amp, you would see very little at the output that was not part of the imput. It is an amp not a synthesizer. :wink:

3. Are the same harmonic numbers present for all fundamental frequencies of my input?

Any frequency biases in your amp are at fixed frequencies, no? As for the additional richness you might get from the lack of headroom and the waves being squashed, i supposed that would depend on the amplitude. More amplitude = more distortion of the waveform = more complexity (good or bad) in the spectra. I am not sure if i am understanding your question right.

4. Even-order harmonics comprise the harmonic series, 1f, 2f, 3f, 4f,... of the frequency 2f. That is, the even-numbered harmonics form a complex wave that is an octave higher than the original signal. Why don't we hear this higher note?

The period common to all those waveforme is still @ 1f. So that is what our ear tracks.

5. The article linked says that even-order harmonics are more pleasing to the ear than odd-order. But in my experience, odd-order harmonics sound very nice. They are similar to a square wave or triangular wave (depending on amplitude and phase) and are reminiscent of a clarinet, a very warm, hollow, throaty instrument.

You and i both know that most folks spout off at the mouth about "harmonics" and "overtones":

1. have no clue what they are talking about
2. hope that you don't either
3. hope that you will be fooled or impressed by thier use of jargon.

jcayer
April 27th, 2007, 10:31 PM
1. If I play a sine wave into the amp, will the output contain harmonics?

If you have a perfect sine wave then there will be no harmonics. The best way to measure this is with a spectrum analyser. Larry, if you work in a university, maybe you can ask someone in the electronics / physics dept. They can show you this.

2. Are the harmonics generated?

As soon as you have a signal into the amp that is not a perfect sine wave, you'll 'generate' harmonics

Someone correct me if I'm wrong...:oops:

Bogo
April 27th, 2007, 10:31 PM
1-3: yes but the harmonics will vary in magnitude
4: You are assuming very high magnitude harmonics which is not the case. Picture the resulting sinewave as distorted with small deviation from the input. Hearing the higher note would not be the intended output of a normal amplifier.
5: I don't know, I think you may be comparing the sounds of different shaped fundamentals and not their harmonics.

April 27th, 2007, 10:57 PM
Well, to be non-technical here, I think of it as the ability of an amp to reproduce harmonics as well as the fundimental. I believe that some amps need reverb because the amp doesn't reproduce the harmonics very well. The reverb gives some depth that takes the place of the harmonics. I base this on the first times I played through a good tweed amp and an AC-30 with Blues. The harmonics were rich and clear and gave a tone that didn't need reverb.
YMMV

April 27th, 2007, 11:15 PM
from my understanding, the harmonics (2f, 3f, 4f, etc.) don't replicate the original signal. instead, they are encountered when the tube saturates and starts to cut off the waveform. if you think of the tube as a door, any signal shorter than the door gets in ok. turn up the volume and make the wave taller and it lops off the top and or bottom of the signal. in a perfect world you'd get a squarish wave out with a flat top/bottom. a tube driven beyond it's design limits, however, will start to add harmonic content during the swing of the signal where it is saturated. so the harmonics that are talked about are the jiggle on the top and bottom of a distorting waveform. this jiggle can get quite pronounced in a heavily overdriven tube, but the main signal will still predominate. this is where classic high gain amps tend to sound better overdriven than fenders - the high gain amps are designed to introduce the distortion a little bit at a time and send it into subsequent gain stages, so that the harmonic content gets subsequent amplification and distortion. fenders, with their few gain stages, tend to give it up all at once, without the 'distorted distortion' effect.

EMan
April 28th, 2007, 11:46 AM
First, as mentioned above - A perfect sine wave has no harmonic content. A perfect square waveform generates an infinite number of odd order harmonics. A sawtooth waveform generates an infinite number of even order harmonics.

Tube amps do not 'play' in any one particular fashion, but a combination of all three, and at various magnitudes. Fundamentally, what we hear is a very complex sound, and impossible to describe in frequency domain mathematical terms.

One of the better explainations of tubes and distortion I've read is here - http://www.milbert.com/tstxt.htm

Larry F
April 28th, 2007, 01:27 PM
First, as mentioned above - A perfect sine wave has no harmonic content. A perfect square waveform generates an infinite number of odd order harmonics. A sawtooth waveform generates an infinite number of even order harmonics.

To clarify these basic waves:

A sawtooth wave consists of all harmonics at amplitudes A/1, A/2, A/3, ... where A denotes the amplitude of the waveform. The even harmonics of a sawtooth wave produce a waveform whose frequency is twice that of the original sawtooth. This waveform, a sawtooth, sounds an octave higher and has half the amplitude.

A square wave consists of odd harmonics at amplitudes A/1, A/3, A/5, ...

Picking up on another post. If an input signal is clipped (as described by the door analogy), wouldn't that create a square wave? And if so, then it would generate odd harmonics. But I'm also told that tubes create even harmonics. My question then, is there a different process by which that occurs other than clipping? I'm having a hard time seeing that distortion creates even harmonics. I don't disagree that even harmonics are present (but my ears don't tell me that, so I'll take the word of people who have done spectral analyses). I've got a spectral analyzer upstairs, so if I have time, I'll try to look at some stuff. Might take a while creating appropriate sound samples, though.

April 28th, 2007, 01:45 PM
prior to fully clipping the wave, a tube that is starting to saturate will skew the waveform, giving you a sawish shape. here are some screenshots of a saturated sine wave (note the assymetry in the first image) a skewed, fully satruated wave and a clipped wave:

http://mysite.verizon.net/vzeowear/id8.html

Larry F
April 28th, 2007, 02:53 PM
prior to fully clipping the wave, a tube that is starting to saturate will skew the waveform, giving you a sawish shape. here are some screenshots of a saturated sine wave (note the assymetry in the first image) a skewed, fully satruated wave and a clipped wave:

http://mysite.verizon.net/vzeowear/id8.html

Fantastic explanation and great wave pics. Got it, thanks a million.

kp8
April 28th, 2007, 03:09 PM
prior to fully clipping the wave, a tube that is starting to saturate will skew the waveform, giving you a sawish shape. here are some screenshots of a saturated sine wave (note the assymetry in the first image) a skewed, fully satruated wave and a clipped wave:

http://mysite.verizon.net/vzeowear/id8.html

rocking~

EMan
April 28th, 2007, 03:21 PM
Add to all that, the fact that the signal coming off the tube plate(s) may not be symmetrically clipped. I doubt that there's more than a few Champs (for example) today that have perfect symmetrical clipping.

Wally
April 28th, 2007, 04:01 PM
First, as mentioned above - A perfect sine wave has no harmonic content. A perfect square waveform generates an infinite number of odd order harmonics. A sawtooth waveform generates an infinite number of even order harmonics.

Tube amps do not 'play' in any one particular fashion, but a combination of all three, and at various magnitudes. Fundamentally, what we hear is a very complex sound, and impossible to describe in frequency domain mathematical terms.

One of the better explainations of tubes and distortion I've read is here - http://www.milbert.com/tstxt.htm
From the above link....
"The transistor characteristics which our subjects noted were the buzzing or white-noise sound and the lack of "punch." The buzz is of course directly related to the edge produced by overloading on transients. The guess that this is white noise is due to the fact that many of the edge harmonics like the seventh and ninth are not musically related to the fundamental. The ear hears these dissonant tones as a kind of noise accompanying every attack. The lack of punch is due to the strong third harmonic which is inaudibly "blanketing" the sound. This is correctable by using a large enough pad to prevent

Vacuum-tube amplifiers differ from transistor and operational amplifiers because they can be operated in the overload region without adding objectionable distortion. The combination of the slow rising edge and the open harmonic structure of the overload characteristics form an almost ideal sound-recording compressor. Within the 15-20-dB "safe" overload range, the electrical output of the tube amplifier increases by only 2-4 dB, acting like a limiter. However, since the edge is increasing within this range, the subjective loudness remains uncompressed to the ear. This effect causes tube-amplified signals to have a high apparent level which is not indicated on a volume indicator (VU meter). Tubes sound louder and have a better signal-to-noise ratio because of this extra subjective head room that transistor amplifiers do not have. Tubes get punch from their naturally brassy overload characteristics. Since the loud signals can be recorded at higher levels, the softer signals are also louder, so they are not lost in tape hiss and they effectively give the tube sound greater clarity. The feeling of more bass response is directly related to the strong second and third harmonic components which reinforce the "natural"' bass with "synthetic" bass [5]. In the context of a limited dynamic range system like the phonograph, recordings made with vacuum tube preamplifiers will have more apparent level and a greater signal to system noise ratio than recordings made with transistors or operational amplifiers."

Thanks, Eman, for the link. I have never heard or read the difference explained so thoroughly. The observations of the experiment coincide with my ears. I have heard the difference explained simply this way....solid state creates harmonics which are non-musical and fatiguing to the human ear...tubes create harmonics which are musical and pleasing to the human. My ears tell me that this is so. I play tubes. anySS amp that attempts to create musical sonics will have to deal with the harmonic structure. Imho and to my ear, the Lab Series amps did this to some extent with the Multifilter section.

jjmantele
April 28th, 2007, 07:38 PM
One of the “harmonic” references I often see is that PP type outputs will “cancel the X harmonics but not the Y harmonics” and the SE outputs wont cancel either type of harmonics. I forget which is X and which is Y. It’s supposed to be one of the big sonic benefits of SE over PP. (And that SE is auto-Class A but that’s another Pandora’s box.)

I’ve never seen an explanation why the PP cancels the X harmonics.