diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   When optimizing, go for Lowest 3rd harmonics .. not THD!! (http://www.diyaudio.com/forums/solid-state/134776-when-optimizing-go-lowest-3rd-harmonics-not-thd.html)

lineup 13th December 2008 06:14 AM

When optimizing, go for Lowest 3rd harmonics .. not THD!!
 
hi.

When designing using simulation and/or measurements of a real amplifier prototype,
it is usual to select currents/resistors and other components
using lowest possible THD value as a target.

Lately, I have started using another target value:
:) -> Lowest possible 3rd order harmonic distortion <- :)

I can use first THD as a hint where I have low distortion.
But at this point I do Fourier while changing component value. Often one resistor.

It is very rare that the point where is minimal THD is the same as where is minimal 3rd order dist.
Often is a rather high value of 3rd order hamonics at lowest THD.

If you go for lowest 3rd, you will get a bit higher THD.
This is most often due to the fact, that at this point often 2nd is not maximally dampend.

When you go for lowest 3rd, you get almost automatically lower 5th, 7th, too. The odd order harmonics.

Another observation is, that at the point of lowest 3rd
you will have that nice spectra of succesively falling magnitudes of the individual distortions:
2, 3, 4, 5 etc.


Any comments are welcome ;)
Me, I hate to think of all the times where I went for minimal THD,
which resulted in higher odd order harmonics distortions.
THD is a value that should be handle with caution :att'n:
It does tell you far from the whole story of how good your amplifier will do in reality.


Lineup :) regards

Mooly 13th December 2008 06:58 AM

It is the even harmonics that are pleasing to the ear, the odd harmonics are not.
That's one reason why I think the OPA604 OpAmp sounds ( to me ) so good, as the evens predominate.

roender 13th December 2008 07:00 AM

Lineup,

I don't trust simulator performing FFT spectrum analysis because we don't have good enough component models, and more than that we discard, most of the time, parasitic PCB and component leads L and C values
If you speak about THD measurement, i fully agree with you.

Cheers,
Mihai

nelsonvandal 13th December 2008 08:02 AM

I don't think I like 2:nd order harmonics. I don't think I like any distortion at all. I have no means of measuring distortion, but listening to a JLH type of amp, I think the sound is warmer than neutral but also grainy and harsh. Adding a diff stage should lower the 2:nd, I think, and makes the amp far more enjoyable with a cleaner sound.

When it comes to monolithic opamps, I like those low distortion ones like AD797, AD8599, ADA4899, LM4562 etc.

Mooly 13th December 2008 08:55 AM

Interesting thoughts.
The single ended ended input stage generates mainly even harmonic distortion rising gently at 12 db/Octave and the original JLH is certainly not "grainy" IMO. This was researched and fully documented by D Self in his exhaustive study into amplifier types.
Perhaps it comes down to personal choice and how well or not the amp and it's implementation is carried out. With the LTP the distortion it produces, that is to say, even or odd predominating, all comes down to the DC balance of the input stage.
Many folks like the warm sound of valves, with their corresponding high levels of even harmonic distortion.
I found my musical nirvana with my single ended input MOSFET amp. I have built many designs over the years, and bought a few, but nothing comes close to this. It's better than it has any right to be when you examine the circuit. But it all comes together beautifully and will comfortably see off amps costing many thousands.
Technical perfection is fine, I just don't like listening to it :)

Nico Ras 13th December 2008 09:38 AM

Hi Lineup,

THD is the measure of the power density of the sum of all harmonics generated in the audio spectrum. Whether it is 2nd, 3rd, 9th etc. The highest one will govern the THD of the system.


As Nelson rightly commented no harmonics are what we aim for. THD is a simple measure for the percentage power of harmonic content to that of the original input.

I and others have made some study of this phenomena and a general statement could be that:

A system with higher odd harmonics sound crisper due to the fact that odd harmonic added to the pure sine wave besides squaring it the rise time of the stimulus increases and hence crisper and faster sound. Because the system THD increases with increased frequency this may be more evident at the higher frequencies.

On the other hand even harmonics present a warmer sound and this is that adding even harmonics to a pure sine besides triangulates it actually decreases rise time which tends to sound slower and less vibrant.

Depending on the order of harmonic you accentuate it will make quite a difference in the sound of the tone so that a higher H5 would sound more crisp than an H3 of the same power.

What I read in Lineup's introduction is that he prefers the crisper character of an amplifier, others may prefer the warmer slower character from even harmonics even though both amplifiers have exactly the same THD specification. It has all got to do with the power spectral density of harmonic order.

There is no subjectivity in this awareness, believe me it is real and what one prefers and the reason you like one amplifier sound above the other. Neither is better only different.

Regards

Nico

Nico Ras 13th December 2008 09:57 AM

One must also consider the fact that the ear has a limited dynamic range and deteriorates with age.

I am hard pushed to accept that anyone will hear the difference in tonal characteristic of systems with THD 40 dB down and the difference between 10 000 watt and 1 watt (at the same instant) or maybe between a big drum beat and a mosquito tap dancing on a thick carpet.

I fail to see the reason for amplifier designers striving for with THD 100dB down, which is the difference between the drum beat and a dead mosquito tap dancing.

I know we all try to achieve absolute zero, but before the advent of simulators and only using measurement equipment, few had the time nor inclination to sit and tweak component values to achieve better than maybe two order of magnitude of THD.

So for these amps there may have been a significant difference in tonal character, but for the things we discuss in this forum, I doubt very much anyone could really hear this.

homemodder 13th December 2008 10:23 AM

While I feel it is better to have a favourable distortion spectrum where the H2 dominates the others, I agree with Nelson that one should be doing targeting this in view of having the lowest possible distortion. H3 is not so bad anyway, the higher odd and even are more damaging. Build Nelson Pass simple Jfet son of Boz which can be optimised to have only very low levels of H3 and youll realise that H3 is not bad at all. Amps with very high levels of H2 and very little H3 sound muffled, optimum that I find is a declining order of all the harmonics from H2 and preferable where after the third they are either absent or so low as to be insignificant. This is well known and thoroughly tested among tube lovers. Ever wondered why the symanym sounds as good as it does, look at the FFT spectrum.

Too much emphasis is being put on how amps sound, how does your source sound, and your speakers?? Junk into a amp, junk out of your amp, sometimes a less than optimum amp to compensate for bad sources is better, an amp should ideally be a wire with gain. Having some level of H2 in amps is good because cdplayers in particular sound cold and the H2 seems to disguise this well. Ever pondered why tubes have been making a comeback, think of the advent of the cdplayer. Even the coldest and lifeless sounding cdplayers can be made to sound reasonably good if coupled to a tube amp with ample H2 distortion figures. One has to match components carefully for optimum sound, include your speakers too.

Then there are offcourse those that prefer high doses of distortion, to get what I would call sound effects. While some of these amps do sound very good and pleasing to the ear, I am afraid they are larger than life and dont accurately reproduce music as played by real instruments. When compairing amps dont compare to another amp, compare to a live instrument being played in the same room. Dont get tricked by the mind, there is a difference between good sounding and real life sounding.

AndrewT 13th December 2008 10:49 AM

I like the idea being promoted by Lineup.
I wonder if there is a cheap way to actually perform the adjustment of minimising the third and the other odd harmonics?
Curl talks about eliminating the seventh. Could this be along similar lines?

Pan 13th December 2008 10:59 AM

Some thoughts reading this thread..

The idea with really low THD/IMD from any piece of audio gear is to make sure that the sum of all distortion in the complete recording and playback chain is low and prefereably inaudible if Hifi is the goal.

As for CD players sounding cold, that is not my experience and it would be interesting to know what lead to that conclusion. The CD format as such is extremly neutral and a well designed CDP/DAC will be more or less transparent.


/Peter


All times are GMT. The time now is 07:07 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2