Why we need to check on the harmonics and THD?
In the past years, I built amplifiers according to the schematics and kits. Also, I believed that using those high end audiophile components will yield better sound from the amplifier that I built.
Few years ago, I heard from the electronic engineer who had been working on design and calibration of electronic amplifier and equipment that it is important to minimize the distortion of the amplifier in order to get the best sound out from it. I wasn't sure about it because I didn't have the equipment for perform such measurement. Luckily, I bought a Keysight oscilloscope last year and found the Soundcard Scope program for PC that I could have the spectrum analysis function and THD measurement.
During the past 10 months, I tuned most of my amplifier for the best THD figure. Now they sound so good that impressed me so much.
In the past years, I built amplifiers according to the schematics and kits. Also, I believed that using those high end audiophile components will yield better sound from the amplifier that I built.
Few years ago, I heard from the electronic engineer who had been working on design and calibration of electronic amplifier and equipment that it is important to minimize the distortion of the amplifier in order to get the best sound out from it. I wasn't sure about it because I didn't have the equipment for perform such measurement. Luckily, I bought a Keysight oscilloscope last year and found the Soundcard Scope program for PC that I could have the spectrum analysis function and THD measurement.
During the past 10 months, I tuned most of my amplifier for the best THD figure. Now they sound so good that impressed me so much.
To each his own. Although from a perspective of a literature reader I strongly agree to precision, from listeners perspective I am a nightmare to argue with.
I like certain colorization and expansion in music I listen. First off dudes I listen are dead or promote dying. They surely would not mind me applying disarmonia to their pieces, due to success and victory having many different roots too. Second off, I am not analyzing much of a technique they use for playing or anything, I prefer to align things in "cores" of importance. This leads to uniqurness of a listener. I just generate more additional cores instead of "cloning" the "plant, child, painting" of artist. Call me silly, call me fool, I will not care for ultralow/any other pattern THD, because it promotes misperception of self-value.
I like certain colorization and expansion in music I listen. First off dudes I listen are dead or promote dying. They surely would not mind me applying disarmonia to their pieces, due to success and victory having many different roots too. Second off, I am not analyzing much of a technique they use for playing or anything, I prefer to align things in "cores" of importance. This leads to uniqurness of a listener. I just generate more additional cores instead of "cloning" the "plant, child, painting" of artist. Call me silly, call me fool, I will not care for ultralow/any other pattern THD, because it promotes misperception of self-value.
Last edited:
Post provided by "Philosopher" followed by "Overactive Imagination" by Death (Individual Thought Patterns, 1993.)
Yeah. To each his own. I fall on the ultra-low THD end of the spectrum. I used to build tube amps and had an excellent 300B design that sounded heavenly on simpler music. I wouldn't call it a precise amp by any stretch of imagination, but it sounded pleasing as long as the source material wasn't too demanding. Throw some Joe Satriani at it and the sound turned to mush.
For the past 6-7 years, I've been firmly in the ultra-low THD camp. I find that such an amp sounds excellent on all types of music.
Ultimately, sound quality comes down to your personal preference. Some prefer that the amp adds a little something-something to the sound. For example, the lush midrange provided by a 300B can be really flattering to some types of music. Some like tweaking: Use a different cartridge/amp/DAC/headphone/whatever with different genres/artists/albums/tracks/whatever. Others, me included, prefer a set-it-and-forget-it approach. I figure that if the artist, recording engineer, or producer wanted a lush midrange, they would have mixed the track that way.
As far as I've read, the scientific literature is pretty clear: People, on average, in blind randomized trials, prefer audio gear with good measurements (plural!) This is true for casual listeners, trained listeners, audiophiles, professional musicians, etc. But few of us live in a double blind, randomized scientific experiment (or maybe we do; maybe the rats are running the show 🙂) so we're influenced by factors other than the physical stimulus that reaches our eardrums in our evaluation of sound quality. That's well-documented as well.
Tom
For the past 6-7 years, I've been firmly in the ultra-low THD camp. I find that such an amp sounds excellent on all types of music.
Ultimately, sound quality comes down to your personal preference. Some prefer that the amp adds a little something-something to the sound. For example, the lush midrange provided by a 300B can be really flattering to some types of music. Some like tweaking: Use a different cartridge/amp/DAC/headphone/whatever with different genres/artists/albums/tracks/whatever. Others, me included, prefer a set-it-and-forget-it approach. I figure that if the artist, recording engineer, or producer wanted a lush midrange, they would have mixed the track that way.
As far as I've read, the scientific literature is pretty clear: People, on average, in blind randomized trials, prefer audio gear with good measurements (plural!) This is true for casual listeners, trained listeners, audiophiles, professional musicians, etc. But few of us live in a double blind, randomized scientific experiment (or maybe we do; maybe the rats are running the show 🙂) so we're influenced by factors other than the physical stimulus that reaches our eardrums in our evaluation of sound quality. That's well-documented as well.
Tom
Last edited:
It is just like a cup of tea or coffee. Everyone would have his/her own preference. It is simply my own preference and strike for the lowest THD that I found it suits my taste.
The way I can explain my experience with it is that I found it sounds much more realistic when the THD is low. The vocal of my favorite singers sounds right to my ears. Also, it yields better clarity and definition in both treble and bass.
The way I can explain my experience with it is that I found it sounds much more realistic when the THD is low. The vocal of my favorite singers sounds right to my ears. Also, it yields better clarity and definition in both treble and bass.
I guess I like to have a cup of coffee at guests a lot. 
my amplifier cannot do what I would ideally preffer - at the moment I am having high R air core chole loaded germanium amp, which cannot provide enough clarity on some fast drumming sections, yet I like it "philosophycally" a lot and I feel rewarded at other musics. (Calmfully accepts difficult situation). I recall having LME49830 based MOSFET amp I had for few years, and I must say it did a lot more of "satisfying duty" for my being. But sometimes we thrive difficult peacefulness to feel stronger. Don't mind folks being "all involved" with their harmony generators. They are just trying to feel stronger eithin difficulties. That doesn't make you weaker. I assure you, your path is accepted. Yet ours is our own to chose.
my amplifier cannot do what I would ideally preffer - at the moment I am having high R air core chole loaded germanium amp, which cannot provide enough clarity on some fast drumming sections, yet I like it "philosophycally" a lot and I feel rewarded at other musics. (Calmfully accepts difficult situation). I recall having LME49830 based MOSFET amp I had for few years, and I must say it did a lot more of "satisfying duty" for my being. But sometimes we thrive difficult peacefulness to feel stronger. Don't mind folks being "all involved" with their harmony generators. They are just trying to feel stronger eithin difficulties. That doesn't make you weaker. I assure you, your path is accepted. Yet ours is our own to chose.I did the contrary. To see how harmonic distortions effect the sound, I add by 6n1p transconductance mode even order harmonics. Up to 10% it is unaudible above it sounds brighter but nothing disturbing up to 14%. Odd order harmonics of 1-3% with opposite phase to fundamental gives cleaner sound and extra dynamic as it act opposite to the loudspeaker's distortion. To remind that a magnetic pickup cartridge as Shure V at nominal level has 2.5% 2nd and 0.5% 3rd harmonics.
IF I was interested in ultra low distortion I guess I would be turning back to SS amplifiers. I sincerely doubt that differences in distortion under 1% can be heard.
I check every year my ears , 1.5% for 1khz/500hz and 0.5%1khz/333hz. If you are in tube world, because you want to drink mineral water and not pure H₂O of SS.
The measurement way is surely not the listening way. There were studies being made where people prefered tube sound to listen to. When listening with low distortion SS equipment, they say that it sounds cleaner but they became listener fatigue.
At some point in my DIY career, when building big SS amps leads towards this goal of low to zero distortions and best measurements I noticed, that the presented audio show wasn't involving and enough emotional to me anymore. It was like playing a live band in a clinical environment and that wasn't realistic but it always brings my imagination to the conclusion that it only can sound in real life this way when it happens under those circumstances. And we all know that music shouldn't sound this way.
At some point in my DIY career, when building big SS amps leads towards this goal of low to zero distortions and best measurements I noticed, that the presented audio show wasn't involving and enough emotional to me anymore. It was like playing a live band in a clinical environment and that wasn't realistic but it always brings my imagination to the conclusion that it only can sound in real life this way when it happens under those circumstances. And we all know that music shouldn't sound this way.
Last edited:
I like the simplicity of tubes. I imagine I can see the electrons flowing through the vacuum from cathode to anode. I can actually see them after I have smoked a special cigarette!
Why not post some before and after results so we all can be impressed by the improvement you tuned into your old circuits? What did you do to make the great changes in distortion levels?
I believe it's even order cancellation. Regardless, the interaction between amplifier and loudspeaker distortion merits more investigation.
Oscilloscope is mandatory to see obvious distorsions when tubes are not biased right , but after that you can't do much ... only using negative feedback 😀
Of course for an 12AX7 the "sweet" spot is pretty much known , that is a guarantee for low distorsions anyway , harder for other tubes .
Of course for an 12AX7 the "sweet" spot is pretty much known , that is a guarantee for low distorsions anyway , harder for other tubes .
When people try putting in expensive well know tubes like the Telefunken ones and hoping for better sound from the amplifier, it may or may not be the case.
After I measured and tuned my pre and power amplifier with minimal distortion at low power level say 0.5W, 1.0W and 2.0 watts for PA and 2Vp-p for line amp, the amps sounds much much better to me. Effect on those amps with NO negative feedback surely is very significant. Surely, the NFB will help to improve on the THD but it causes other problems as well.
After low distortion tuning, the sound different with various tubes or different amps will be quite similar. Believe it or not.
The solid state amplifiers have ultra low THD mostly but their electric characteristics are quite different from tube amps. This will make them sound differently even with the same THD.
After I measured and tuned my pre and power amplifier with minimal distortion at low power level say 0.5W, 1.0W and 2.0 watts for PA and 2Vp-p for line amp, the amps sounds much much better to me. Effect on those amps with NO negative feedback surely is very significant. Surely, the NFB will help to improve on the THD but it causes other problems as well.
After low distortion tuning, the sound different with various tubes or different amps will be quite similar. Believe it or not.
The solid state amplifiers have ultra low THD mostly but their electric characteristics are quite different from tube amps. This will make them sound differently even with the same THD.
IMO, THD tells you if the design is working right, that nothing is messed up, assuming you know what the design typically does. It's useless for predicting what the unit will sound like, but essential if you want to trust your conclusions. I've seen too many people reach conclusions about a particular piece of gear that turned out to be nonsense, because the gear wasn't working right when they listened to it.
Sound is never only defined by a single mathematical model (which is a measurement parameter, a simplified model of the complex real world). For example, tube sound is 90% chemistry and 10% solid mechanics. If someone thinks he just have to copy the distortion spectrum of the component (or tube, maybe TFK), THD or whatever measurable parameter and then the result will be a copy of the original sound, its just nonsense.
But thats what they try to sell in audio. I have seen adverts for electronic apps that simulate the sound of a TAB V76 or any other famous audio mixer, limiter, studio tube amp with PC based solid state amplification. Just buy a cheap plugin for the mixing PC and it will sound like the famous fab four!
But thats what they try to sell in audio. I have seen adverts for electronic apps that simulate the sound of a TAB V76 or any other famous audio mixer, limiter, studio tube amp with PC based solid state amplification. Just buy a cheap plugin for the mixing PC and it will sound like the famous fab four!
Last edited:
I remember an article about double blind A-B testing that was published in Stereo Review (in the 1970s?).
It compared the sounds of several solid state amplifiers of that era, with a Futterman tube amplifier of that era too.
The amplitudes were very carefully matched.
The double blind tests showed that statistically . . . there was not a detectable difference in the sounds of the different amplifiers, such that any particular amplifier could be reliably detected, nor even which was being played of the two being compared, was it AA, AB, BB, BA, BAB AAA, ABA, etc. Wow!
I do not remember anything about the "listeners", the speakers that were used, the music signal sources, or the room.
But . . .
The common things about each and every one of those solid state amplifiers, and also the Futterman tube amp was:
No output transformer.
Totem Pole Output stage (stacked vertically).
Lots of Global Negative Feedback.
Does anybody remember that article, like I do?
Just curious.
Your comments, please.
It compared the sounds of several solid state amplifiers of that era, with a Futterman tube amplifier of that era too.
The amplitudes were very carefully matched.
The double blind tests showed that statistically . . . there was not a detectable difference in the sounds of the different amplifiers, such that any particular amplifier could be reliably detected, nor even which was being played of the two being compared, was it AA, AB, BB, BA, BAB AAA, ABA, etc. Wow!
I do not remember anything about the "listeners", the speakers that were used, the music signal sources, or the room.
But . . .
The common things about each and every one of those solid state amplifiers, and also the Futterman tube amp was:
No output transformer.
Totem Pole Output stage (stacked vertically).
Lots of Global Negative Feedback.
Does anybody remember that article, like I do?
Just curious.
Your comments, please.
Last edited:
I generated some test WAV files with various tones and harmonics and found I could detect the 7th harmonic of a 100Hz signal at -65dB, and the 5th harmonic of 200Hz at -63dB, being about 0.05% and 0.07% respectively. These higher harmonics aren't protected by perceptual masking which is why they are so easy to pick out.
Intermodulation products are also not always protected by perceptual masking as they can be at very different frequencies to the loud tones responsible for them, so some of them will likely be detectable at similar levels.
The often expressed preference for lower order harmonics forgets that you can't have them without intermodulation products, which will be much more problematic in certain signals.
My working rule of thumb is that 0.1% distortion is not objectionable, 0.01% is probably not detectable by the majority of humans in double blind trials, and is probably only achievable with headphones anyway. However there's little excuse for making electronics that's not nice and linear, its not hard to to achieve with the tech we have.
Intermodulation products are also not always protected by perceptual masking as they can be at very different frequencies to the loud tones responsible for them, so some of them will likely be detectable at similar levels.
The often expressed preference for lower order harmonics forgets that you can't have them without intermodulation products, which will be much more problematic in certain signals.
My working rule of thumb is that 0.1% distortion is not objectionable, 0.01% is probably not detectable by the majority of humans in double blind trials, and is probably only achievable with headphones anyway. However there's little excuse for making electronics that's not nice and linear, its not hard to to achieve with the tech we have.
For discussion, start with a push pull amplifier that has No negative feedback, and that only has 2nd and 3rd harmonic distortion, and no higher order harmonic distortion products.
Suppose that at medium signal levels, it is -46 dBc 2nd harmonic, and -34 dBc 3rd harmonic. -46 dBc is 0.5% second harmonic, and -34dBc is 2% 3rd harmonic.
According to RSS . . .
That is 2.06% Total Harmonic Distortion. All is OK, these are only 2nd and 3rd harmonics, just like instruments produce, like the French Horn and the Clarinet respectively.
Now, apply negative feedback.
The 3rd harmonic distortion comes from the negative feedback signal, and goes back through the amplifier again. That 3rd harmonic is processed by the amplifier’s non-linearity, and becomes the 9th harmonic.
2% times 2% becomes 0.04% 9th harmonic (-68dBc). It seems like we are still OK.
Now, turn up the sound level so that the amplifier without negative feedback would have put out 5% 3rd harmonic distortion.
Then we apply negative feedback, and turn up the volume to the same output level to make up for the reduction in gain due to negative feedback.
5% x 5% = 0.25% 9th harmonic. That is -52dBc 9th harmonic.
Oh, amplifiers do not produce only the 2nd and 3rd plus only the 5th; or only the 2nd and 3rd plus only the 7th harmonic.
It is harder to hear the 5th harmonic or to hear the 7th harmonic (in isolation) when the 2nd, 3rd, 4th, 6th, 8th, 9th, and other harmonics are also present.
And yes, low Intermodulation distortion is very important too.
I will leave it up to you to decide if the 9th harmonic distortion of 0.25% is something you can hear or not.
And I will leave it up to you, if it is something that you do, or do not, like the sound of.
After we decide about amplifiers, please start working on those loudspeakers.
They distort too, especially at high sound levels.
How many of you have loudspeakers that do not distort as much as your amplifiers?
Harmonics?
Intermods?
Frequency Response as flat as your amplifiers?
No FM modulation of the tones of music (Speakers do that to some greater or lesser degree; your amps are free of FM modulation, I hope).
Worry, Worry, Worry.
I prefer to enjoy listening to music, live music, and also the playback of recorded music.
When I listen, I almost always stop worrying about my playback system.
Suppose that at medium signal levels, it is -46 dBc 2nd harmonic, and -34 dBc 3rd harmonic. -46 dBc is 0.5% second harmonic, and -34dBc is 2% 3rd harmonic.
According to RSS . . .
That is 2.06% Total Harmonic Distortion. All is OK, these are only 2nd and 3rd harmonics, just like instruments produce, like the French Horn and the Clarinet respectively.
Now, apply negative feedback.
The 3rd harmonic distortion comes from the negative feedback signal, and goes back through the amplifier again. That 3rd harmonic is processed by the amplifier’s non-linearity, and becomes the 9th harmonic.
2% times 2% becomes 0.04% 9th harmonic (-68dBc). It seems like we are still OK.
Now, turn up the sound level so that the amplifier without negative feedback would have put out 5% 3rd harmonic distortion.
Then we apply negative feedback, and turn up the volume to the same output level to make up for the reduction in gain due to negative feedback.
5% x 5% = 0.25% 9th harmonic. That is -52dBc 9th harmonic.
Oh, amplifiers do not produce only the 2nd and 3rd plus only the 5th; or only the 2nd and 3rd plus only the 7th harmonic.
It is harder to hear the 5th harmonic or to hear the 7th harmonic (in isolation) when the 2nd, 3rd, 4th, 6th, 8th, 9th, and other harmonics are also present.
And yes, low Intermodulation distortion is very important too.
I will leave it up to you to decide if the 9th harmonic distortion of 0.25% is something you can hear or not.
And I will leave it up to you, if it is something that you do, or do not, like the sound of.
After we decide about amplifiers, please start working on those loudspeakers.
They distort too, especially at high sound levels.
How many of you have loudspeakers that do not distort as much as your amplifiers?
Harmonics?
Intermods?
Frequency Response as flat as your amplifiers?
No FM modulation of the tones of music (Speakers do that to some greater or lesser degree; your amps are free of FM modulation, I hope).
Worry, Worry, Worry.
I prefer to enjoy listening to music, live music, and also the playback of recorded music.
When I listen, I almost always stop worrying about my playback system.
Last edited: