Quote:
Originally Posted by
tubetwister
Back to the non linear distortion thing I think I get this for the most part ,
"For larger values of u, the higher order coefficients such as a_2 and a_3 come into play." -wikipedia -
not to sure about this stuff so much though ,
I'm going to try to keep this discussion math-less. ;-)
So far we've been talking about nonlinear distortion of the most obvious and clear cut kind, which is to say: clipping. In clipping the input signal tries to drive the output to places it was never designed to go and it completely balks.
the Wikipedia example of clipping was:
"For many devices, a linear model is accurate only for small signal levels. For example, at 2 volts input, a typical audio amplifier might put out 20 V, meaning the linear gain is 10 V/V. For 3 V input, it might then output 30 V. However, the model implies that at 50 V input it would produce 500 V, which is not possible with most amplifiers."
The other form of nonlinear distortion is not nearly as obvious nor as clear cut.
My paraphrased change to the above describing the other more gradual form of nonlinear distortion than clipping would be:
For all real world devices, a linear model is accurate only for small signal levels. For example, at 2 volts input, a typical audio amplifier might put out 20 V, meaning that the linear gain is 10 V/V. For 3 V input, the amplifier might output 29 V, meaning that the gain has dropped from 10 V/V for the first 2 volts of input to 9 V/V for the last volt of input.
Basically, the figurative little engine in the amplifier that accurately pulled the output from 0 to 2 volts lost a little steam, and when presented with a command to put out 10 more volts, it only managed to put out 9 volts. The dynamic range of this kind of device also eventually runs into a brick wall (clipping), but first there are these smaller errors.
Here is what happened graphically:
The blue line represents an ideal linear amplifier, while the pink line represents this particular form of nonlinearity.
This graph shows why this is called third order harmonic distortion:
The upper wave is at the fundamental frequency. Note that in the given time interval there is just one positive peak. This represents the pure signal.
The lower wave is at 3X the fundamental frequency. Note that in the given time interval there are three positive peaks. This wave represents the distortion.
If you add the two waves together, note what happens at the time marked with the red line. The first negative peak of the third harmonic cancels the positive peak of the fundamental (makes the peak value of the positive sum of the two waves less positive, which is what happened in the other plot).
Also note what happens at the time marked with the blue line.
The last positive peak of the third harmonic cancels the negative peak of the fundamental (makes the peak value of the negative sum of the two waves less negative, which also is what happened in the other plot).
So we can say that the addition of the third harmonic distortion accomplished the same graphically as we saw happen in the first plot of the output of the amp plotted against the input to the amp.