Thru zero FM schematic is up!

[Rene Schmitz]

At 19:37 29.09.99 +0200, mbartkow at et.put.poznan.pl wrote:
>Thanks for very nice schematics. I wonder what is the cause of
>the distortions observed when changing the modulation depth.

Its not really distortion. What I mean with dirtyness is that there are
sudden "jumps" and "screams", perhaps because of the offsets of the OTA and
the zero crossing detector. This goes specially for the output offset of
the OTA which as I see it is dependant of the Iabc of that device. 
The trimpot labelled "output offset" really does only fix the DC-difference
between the forward and the backward mode. And I found that the triangle
shape depends on the amplifier bias current. 

Also there is a small high frequency oscillation arround the zero
crossings, which I can find at the output of the linear summing amp, and
thus at the input of the zero crossing detector. I don't know how to
improve that, since I already placed decoupling caps at the comparitor,
grounded all unused in and outputs and went the brute force hysteresis way.

>I suggest some shots from the oscilloscope would be very informative.

Yes, If I had a scanner, or webcam or at least a digital scope with an
interface...

I wonder how good the time-dependant modulation works with other TZVCOs
like the E.N. one. Thats another way of asking if Yamaha style FM is
possible in the analog realm.

>Anyway, there is one thing I would personally change in your circuit:
>I would reconnect IC2d (as a follower) directly to the output of IC2a,
>instead of cascading two inverters. I suspect that in the current 
>configuration there is a nonsymmetrical delay introduced by IC2c or
>IC2c+IC2d, depending on the polarity of the mod. linear input.

I had it like that at the beginning. Only one follower and one inverter.
There was a squarewave superimposed on the signal, from the direction
signal. I guess this is from asymmetries in the schmitt triggers
thresholds. I've added the trimpot which is driven from the direction
signal!  To preserve the symmetry for the schmitt trigger the second
invertor is used. Seems wasteful, I know. But the action of the voltage
from the trimpot would not be symmetrically with your suggested scheme.
Funny thing is the trimpot changes the voltage at the buffer! 

>BTW, what are your motivations of using BA6110 instead of CA3080 ?

Pinout! I like to matrixboard with that SIL stuff. Very handy that you can
expand your circuit both sides of the IC! I guess the circuit would work
equally well with a CA3080 or LM13600.

Bye
 René




error:asciiartisnotfully  \ uzs159 at uni-bonn.de
supportedwithyourcurrent   \ http://www.uni-bonn.de/~uzs159
displaydriversettings...    \ http://members.xoom.com/Rene_Schmitz