[sdiy] Harmonic bandwidth

[Ian Fritz]

At 02:27 PM 1/9/2008, Nicholas Gregorich wrote:
>>>This seems like a valid point, but what bothers me is that there isn't 
>>>much of a transient effect since there are only 32 harmonics. Adding 
>>>sines from 100Hz to 3.2kHz shouldn't be causing a speaker to struggle, 
>>>should it?
>>I have to admit, I don't really know what this waveform looks like.  How 
>>fast is the rise, say 10% to 90% at 100 Hz?
>
>The 10-90% rise time on a 100Hz sine wave is approximately 1.5ms.

Ah.  So for our rough comparison I think we would take twice that for the 
peak-to-peak slew. So 3.0 ms.

>The 10-90% rise time on the saw wave [100 cps] with 32 uniform phase 
>harmonics is approximately 89.5us

So the corresponding sine wave of the same amplitude would be at 
100*3.0/.09 = 3.3 kHz.  I guess that makes sense -- the "Saw" slew rate is 
about what the slew rate of the highest harmonic would be if it had the 
full amplitude of the wave.

And then with the randomized phases what would the effective slew rate 
be?  That's what we would have to compare against.

>But in this case since the highest frequency is only 3.2kHz I find it hard 
>to imagine that the speaker is slewing already.

I see what you are saying.  But maybe between 100 Hz and 3.2 kHz, you are 
partly into the crossover regimes between the individual drivers?  So there 
could be some phase dispersion.  Maybe these tests need to be done with 
super audiophile speakers.

>I have to admit I don't know much about speaker physics. :)

Me either. Totally winging it here.  :-)

Ian