[sdiy] 4069 VCO

[Scott Gravenhorst]

=?iso-8859-1?Q?Ren=E9?= Schmitz <uzs159 at uni-bonn.de> wrote:
>At 18:52 28.10.01, you wrote:

[snip]

>Btw, my oscillator has that spike too, the unconventional 
>reset scheme does that.

This doesn't bother me at all, I will accept *any* 
waveform as long as the voltage to frequency relationship 
is linear.  (I'm not done trying with this BTW).

>Oh, you should really add the PWM stuff, because 
>this circuit has a nice floppy square that somehow 
>reminds me of the 303 square.

I certainly intend to put the whole circuit together.
In fact, one of the reasons I was interested was that
very feature!  Very clever, all with one IC.  I 
built only the VCO part to limit confusion. 

[snip the paragraph about my V/I converter which turned
out to be a piece of crap ... read on]

>I have a suspicion, have you been monitoring at the 
>circuits output?! 

Only with my oscope.  I assumed the input impedance
would not be a problem.  I was setting the oscillator
to have exactly 2 full cycles on the screen, read the 
CV with a DVM and then change the CV to get 1 cycle 
exactly on the screen and read again.  If the response
is linear, I would expect to see the CV exactly half
the previous value.

I am scoping the output of the integrator amp, because
that is where you tapped the saw output.

>If so, then the cap at the output might 
>be responsible for the no longer straight ramp (at lower 
>frequencies anyway. I figure that using the 10k the 
>frequency was rather high, so you might not have 
>encountered the highpass action of the output cap.) 

I will keep this in ming when I add the PWM part and 
the real output stuff.

>The mistracking could be due to the current that the 
>base robs, the control currents are rather tiny, and a 
>ever so small base current for the darlington could 
>disturb the linearity. You could try a n-JFET instead of 
>the darlington.

Jeeziz.  I looked at what I had breadboarded and I can't
imagine how it ever worked at all, let alone half-butt
linear!  I had the (positive) CV going into the inverting
input of the opamp.  This CLEARLY reverse biases the NPN
darlington.  So to prevent my brain for shorting out, I
decided to accept that it's wrong and not worry about why
it was 'working'.

When I put the CV into the noninvert input, it seemed to 
work, but the circuit now had a really bad frequency jitter.
The solution (and I really don't know why other than some
gain limiting and input impedance lowering ??)  was to place
a 10K resistor on each opamp input to ground.

NOW: At the low CV end of the scale, I seem to get numbers
that say "linear" for at least 3, maybe as many as 5 octaves.
If I can get 5 solid linear octaves, I'll be happy since it 
will work with my FatMan.  I also restored the schmitt trigger
feedback resistor to 22K, which seems to be fine now.  I
do notice that the ramp is still not straight.  The first 
quarter of it curves up and then it becomes linear for the
rest of the cycle.  Again, I don't care as long as the V/F
relationship is linear.

Later today, I will do listening tests to see how closely I
can get it to track my FatMan.  (After the rest of my coffee
and some breakfast.)


The schematic of my testing is at:

http://home1.GTE.NET/res0658s/electronics/4069_linear_vco_experimental.gif

If you see something else I did that's going to limit
the performance, I'd be most anxious to hear about it.

>Just wildly guessing of course, otherwise I'll be as baffled as 
>you, and need to try this my self.
>
>Cheers,
> René
>
>-- 
>uzs159 at uni-bonn.de
>http://www.uni-bonn.de/~uzs159

Well, guessing or not, your questions made me go back and
think about this, as well as LOOK at what I did.  For this,
at the very least, I must thank René profusely.


=========================================================  
- Government: The other religion.
- The media's credibility should always be questioned.
- Lambs who lie down with lions are lunch.

-- Scott Gravenhorst | LegoManiac / Lego Trains / RIS 1.5
-- Linux Rex         | RedWebMail by RedStarWare
-- FatMan: home1.GTE.NET/res0658s/TekLab_FatMan_WebSite/index.html