[Amps] IMD

[Dan Mills]

On Tue, 2012-01-10 at 13:39 -0500, Bill Fuqua wrote:
> Wait,
>    Something is not right here.
> How did you get the baseband envelope?

I generated my square wave approximation in the obvious fashion (sum of
series of sines), then stuffed it through a IIR hilbert transform (and
appropriate delay) to get the I,Q pair, standard phasing method logic. 

Baseband envelope is then simply sqrt (I^2 + Q^2), being the magnitude
of the resulting phasor. 

Basically I wrote a very simple minded SSB generator in software, had it
write a half dozen data files, then loaded them into gnuplot to produce
the graphs. 

You can get something very similar with the following mess:


gnuplot> plot sin(x + pi/2.0) + 1.0/3* sin((3*x)+pi/2.0) + 1.0/5 * sin
((5*x)+pi/2.0) + 1.0/7 * sin ((7*x)+pi/2.0), sqrt((sin(x + pi/2.0) +
1.0/3* sin((3*x)+pi/2.0) + 1.0/5 * sin ((5*x)+pi/2.0) + 1.0/7 * sin
((7*x)+pi/2.0))**2 + (sin(x) + 1.0/3* sin((3*x)) + 1.0/5 * sin ((5*x)) +
1.0/7 * sin ((7*x)))**2),sin(x) + 1.0/3* sin((3*x)) + 1.0/5 * sin
((5*x)) + 1.0/7 * sin ((7*x))

I still say doing it in C was easier, mathcad, matlab, or octave would
also be entirely possible. 

Now, what have I misunderstood about SSB? 
This has a direct impact on an exciter I am developing, so knowing if
and where I am losing it would be a very good thing. 

Regards, Dan.