diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Solid State (http://www.diyaudio.com/forums/solid-state/)
-   -   HOW CAN I CREATE 10% THD ?? (http://www.diyaudio.com/forums/solid-state/223016-how-can-i-create-10-thd.html)

aarvin2 6th November 2012 09:09 PM

HOW CAN I CREATE 10% THD ??
 
Hi guys I know this is a very strange question but for example I am assembling an amp, and I would like to have 10% of THD at the control of a switch, what could I do to get it ??


Please feel free to be very technical about the parts and methods to be used to get 10% of distortion but then when i don't want it I just switch it off and listen to my super low distortion sound again.

I would like the 10% distortion to be added to the output signal(at any volume), instead of just distorting the output signal by cranking the volume.


Thanks in advance for your help guys!! :)

richie00boy 6th November 2012 09:16 PM

You need to think more about what spectral content you want. You could have it all as second harmonic, or spread over harmonics 2-10 for example.

Maybe stating why you want it will help guide you.

Either way, it's not going to be easy to achieve with a switching type action unless you use DSP.

dsdjoy 6th November 2012 09:22 PM

I am not sure what you are up to...

10% Distortion is A LOT for an amp.

A difference between 0.015% THD and 0.030% THD at 1khz & 1Watt can be recognized already. Anything above 1% should be avoided.

Then 2nd harmonic distortion sounds different from 3rd harmonic,....

If you would like to have just really distorted sound,
Google for "Guitar Distortion Pedal" and put such device on the signal input of you amp.

Or distort some mp3 musik using PC Audio effect tools, burn it onto CD-rom and enjoy ;)

Quote:

Originally Posted by aarvin2 (Post 3231211)
Hi guys I know this is a very strange question but for example I am assembling an amp, and I would like to have 10% of THD at the control of a switch, what could I do to get it ??


Please feel free to be very technical about the parts and methods to be used to get 10% of distortion but then when i don't want it I just switch it off and listen to my super low distortion sound again.

I would like the 10% distortion to be added to the output signal(at any volume), instead of just distorting the output signal by cranking the volume.


Thanks in advance for your help guys!! :)


cotdt 6th November 2012 09:32 PM

People cannot recognize even 5% harmonic distortion added into their music via DSP in double blind tests. With pure piano music, low notes only, no high frequency content, it is 3%. I would say 10% is reasonable.

0.1% may be a lot of distortion for a DAC or an amp, but you sure as hell can't hear it according to psychoacoustics research. The papers are publically available for anyone interested.

Harmonic distortion is a meaningless measurement that a lot of engineers believe in, but is not supported by any scientific research.

sreten 6th November 2012 09:36 PM

Hi,

Its a relatively easy stage to add before the volume control, though its hard
to add so quiet passages are distorted the same amount as loud passages.

10% of THD covers a multitude of sins, from just pure 2nd harmonic which
will sound OK to say an even spread of higher harmonics (still 10% THD)
which will sound utterly grim / unlistenable, with horrible intermodulation.

A CMOS inverter might work quite well, but I know little about adding
distortion to a full range music signal, a lot more for guitar and bass.

rgds, sreten.

cotdt 6th November 2012 09:42 PM

Can this be done with analog? It is very simple to do with a DSP.

DF96 6th November 2012 09:42 PM

You need a very efficient compandor process to get a constant signal level in the middle. Feed this at the level of 10mV to a BJT. That will give you 10% 2nd, which will be roughly 10% THD - you can adjust a bit up or down. The expander part of the compandor will then give you the original signal level, but now with 10% distortion.

I can't imagine why you would want to do this.

dsdjoy 6th November 2012 09:50 PM

Quote:

Originally Posted by cotdt (Post 3231241)
People cannot recognize even 5% harmonic distortion added into their music via DSP in double blind tests. With pure piano music, low notes only, no high frequency content, it is 3%. I would say 10% is reasonable.

0.1% may be a lot of distortion for a DAC or an amp, but you sure as hell can't hear it according to psychoacoustics research. The papers are publically available for anyone interested.

Harmonic distortion is a meaningless measurement that a lot of engineers believe in, but is not supported by any scientific research.

I just had the opportunity to compare two completely identical amplifiers, the only difference being that one fixed at 0.025% THD, whereas the second, identical amp was fine-tuned to deliver 0.015% THD in one listenting session and to 0.025% THD in a second listening session. In both sessions the amps were compared:
0.025% THD vs. identical amp with 0.015% THD
0.025% THD vs. identical, fine-tuned amp with 0.025% THD

Measurements were done before listening session using 1khz sine-wave and 1 WAtt RMS output.

The difference in sound was very noticeable between the 0.015% and 0.025% version in direct A/B comparisons: The added 2nd and 3rd harmonics of the higher THD-version added some richness and full-ness to the sound.

Of course measurements with 1khz sine-wave can not be compared with playing complex signals (Musik). With Complex signals also intermodulation distortion comes into play.
And also I listen to music usually at higher output levels than 1 Watt ( Higher level = higher distortion from the power amp).

Here is a good article about distortion:
6moons.com - industry features: "Distortion &*Negative Feedback" by Nelson Pass

cotdt 6th November 2012 09:53 PM

dsdjoy, interesting results. How did you increase distortion in one of the otherwise identical amps?

dsdjoy 6th November 2012 10:35 PM

Quote:

Originally Posted by cotdt (Post 3231287)
dsdjoy, interesting results. How did you increase distortion in one of the otherwise identical amps?

Amp design is a push-pull stage with two transistors ( precisely a MU-follower with 2 SIC Jfets in output stage).
Amp is fed with balanced signal into input stage consisting of long tailed pair, using small amount of negative feedback.

These amp design and the 2 output jfets work mostly in perfect symmetry, resulting in cancellation of 2nd harmonics.

By doing minor tweaks of resistor connected to source-pin of one of the output transistors ( or by inserting a less well matched jfet) the symmetry is disturbed and 2nd harmonics is increasing, also 3rd to some extend.

You can argue that such small circuit changes resulted in sound changes, and that it was no the change in THD. I can't prove the opposite.

What led me to my conclusion is that I played with many variations of the circuit, always measuring (THD and 2nd to 5th harmonics), listening and comparing the modified versions to the original amp. In all attempts I was only able to have the same sound from both amps, if 2nd and 3rd harmonic levels of them were identical.

Also I can't argue that I can hear 0.015% THD. The given values are only reference values measured at low output level. In real listening situation with real speakers for this amp the values rise to approx. 0.15% to 0.30% THD at 10 Watts, and then also intermodulation distortion (I can't measure) probably plays a role.

Also I should note that I used Air Motion Transformer tweeters which extend to beyond 40khz and have extremely low distortion when reproducing music at mid and high frequencies. This way smaller differences in amp distortion are not masked so much by - usually - higher speaker distortion.


All times are GMT. The time now is 02:43 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