|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
64.114.95.25
In Reply to: Does single ended have less odd harmonics than PP? posted by violinist3 on March 16, 2007 at 14:13:58:
Ideal push-pull circuits cancel even order and increase odd order. Maximum even order cancellation is complete cancellation (infinite attentuation), maximum odd order increase is 6 dB per harmonic. P-P amplifiers using single-ended front ends of course can't (normally) cancel the front end's 2nd harmonic so things get complex quickly, especially with feedback.
Thermionically addicted.
Follow Ups:
This is flat out wrong and one of the often quoted 'old wives tales" of audio!An ideal P-P ciruit should cancel ALL even harmonics but practical amps never do.
But there is no reason why they increase odd order at all - just that with lower amounts of even order the odd's stand out more on a spectrum analyser's screen.
Of course, adding in a high amount of negative feedback will mess up the relationships and then all bets are off on any paper predictions.
That post seems to have struck a nerve. Which part is a wive's tail?"An ideal P-P circuit should cancel ALL even harmonics..."
Exactly what I wrote, saying nothing about practical amps save that they're more complex and depend on topology. In the latter case complete even harmonic cancellation, as in any signal nulling, is so sensitive to exactly matched level, phase and distortion component levels it seems to me a near-miracle tubes do it at all without heavy feedback around each stage to force gain and distortion levels between legs into alignment.
Regarding odd harmonics, I agree it was poorly phrased and misleading. Odds do add, but so does the fundamental, so the ratio of fundamental to odd remains the same. It's easy to confirm in Spice with inherently ideal components. If odds didn't sum, they would appear 6 dB lower relative to the fundamental on the output compared to either driven leg. It's a differential stage, what other choices are there than add or subtract depending on phase?
Finally, I said nothing either way about the desirability of the resultant harmonic distortion spectrum.
It's important to pay attention to the NET harmonic distortion of an amp. The 12AT7 based differential phase splitter I favor combines WELL with PP "finals" and yields a PLEASING "waterfall" HD spectrum.OTOH, the 'T7 is WRETCHED in SE circuitry, as 2nd order HD is way over the top. Horses for courses!
Eli D.
This post is made possible by the generous support of people like you and our sponsors: