[sdiy] analog through zero FM dummy

[René Schmitz]

Hi Martin!

At 08:36 13.02.01 +0100, Martin Czech wrote:
[...]
>Oh, really? The recent thread was about zero carrier, but sometimes
>a non zero carrier is also nice, e.g. you start with a sine wave
>that gets more and more side partials.

Ok, but at the beginning and the end of your discussion you were making a
statement about thru zero FM. The zero carrier is just a special case, but
an important one IMO, because as you say, these additive terms get in the
way. Its the specialty of these circuits that these terms dissappear. That
was what I wanted to emphasize.

[...]

>Ok, I aggree if wc=0 is given. If wc !=0 i disagree, because then (the
>circuit must be changed a little bit) and we come to my additive
>solution, I think.

Not quite, just where I wrote i0*M*sin(wm*t) you'd have to subst that for 
(i0*M*sin(wm*t)+i1)
Of course this i1 then also gets scaled along with the exponential term. 

[...]
>Usual analog TZFM circuits seem not to include wc != 0, which is perhaps
>the more interesting case.

I'd say add some constant input to the modulation input and you're there.
Or is this just another form of FM?! There are so many possibilities 
to change the frequency... The "classical" descriptions are always assuming a 
fixed carrier (i.e. for broadcasting) and only a relatively small
modulation of that.
Nowhere near "deep" FM or even thru zero!

>:::Some remarks here: one must track the modulation frequency along with the
>:::carrier frequency, only then you get constant peak freq deviation, i.e.
>:::constant timbre. (To modulate the timbre dynamically you can use the expo
>:::input of the TZFM-VCO, varying the PFD.)
>
>How does this come together with your first remark saying wc == 0?

Right, that is rubbish. The expo control input really affects the peak
frequency deviation, not the carrier frequency. And that statement is about
constant modulation index which is defined as PFD/fmod. So to get const
modulation index the PFD must be tracked to the modulation frequency. 

>:::And, there is one other thing one should notice, when applying waveshaping
>:::to the generated tzfm-triangle (to get sinusodial output) one gets a
>:::different result than with a generic tzfm-sinusodial oscillator.
>
>? I don't think so. As long as the tri wave represents the phase in a correct
>manner (and I believe it does) it will "point" to the right entry in
>our "sine table" (shaper).

Hm, although this sounds reasonable, I can't believe that the extra
sidebands that the triangle oscillator generates compared to a sine
oscillator don't generate additional harmonics when they're applied to a
shaping function.

>:::I still wonder if a real sinewave oscillator can be made running back and
>:::forth?
>
>It should be possible. The usual quadrature oscillator with instant amplitude
>stabilisation (reference signal cos(x)^2+sin(x)^2 == const) should
>be ok. Only the direction of integration must be switched (observing
>stability) , like in the Haiblesian version of Tietze/Schenk tri wave
>oscillator. Juergen choose the way to generate tri waves with 
>waveshaping to sine thereafter.

Yes, it seems that a multiplier with a current output would be a handy
thing here. 
Or something that would be a four-quadrant OTA. I think the quadrature +
stabilization is a quite good idea, though I'm not sure if I'll pursue that
soon.

Bye,
 René

-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159