[sdiy] my VCO not oscillating :(

[harrybissell]

not being afraid to embarass myself (lol)... I usually try and force oscillators

to work. If you think its the schmitt trigger that doesn't work, clip in a
pushbutton
and crobar it.  If you get a ramp, whose slope you can control with current...
then
it is the schmitt trigger that is not working. Then disconnect the ramp circuit
and
force a voltage into the schmitt, and see how that behaves.

It will help if you say to it... "now... Obi Wan... I am the MASTER"

H^) harry

Michael Ruberto wrote:

> >
> >Ok, here goes nothing:
> >U1 is a standard summing network that scales the inputs to 18mV output
> >for 1mV input,
>
> no problem, it's tracking the CV just fine.
>
> I think (experts?).  Check for this first at the top of
> >the tempco.  The 1583 will do the exponential conversion sourcing an
> >exponential current from the linear (18mV/Octave) input to the base.
> >Q1 etc I believe to be high frequency compensation (but this is really
> >just a guess).
>
> Q1 seems to be the current converter for the servo loop since it's in the
> feedback path of U2. The HFT adjustment sets the initial level for the error
> signal. I think. Measurments here indicate steady unchanging currents.
>
>     hmm... I'm back to the datasheet for the 4302,
>
> Yes, this also has been a thorn in my side. We had a hunt for it about 6
> months ago and some list members found it's specs but the datasheet itself
> and the pinout are still unknown. I have used this device in a VCF connected
> as DSG viewing the flat side and it works. However, in this application it
> wouldn't make a difference whether I took the signal from the D or S. At
> least I know which is G :)
>
> >FET's operate in enough strange ways I'm really not shure how it will
> >act here, wheather we should consider the current out from the mirror
> >or the resulting Vce as more important..
>
> Well it looks like an upside-down Mu-amp. It's a source follower but the
> jury is out on whether the bottom FET is contributing to the load or just
> sourcing. I was looking through the archives where they discussed this very
> circuit back in 99 but couldn't find a real conclusion....
>
> There is alot of feedback in
> >the circuit after that *scrathes head..*  Experts?
> >
>
> By this I am assuming you mean the network comprised of D1, R23, R28 and
> R29? I think that (for lack of better terms) this creates a negative "sump".
> As the sawtooth rises it eventually pokes it's head up enough to reverse
> bias D1 just enough to trip the Schmitt trigger consisting of Q4 and Q5. I
> have been looking rather closely at this portion of the circuit with the
> thought that maybe the "sump" isn't at the right level and the trip never
> happens. On the cathode of D1 I measure an almost perfect -4V and
> -3.5 on the anode. Problem is I have no idea what should be here. In this
> situation it would be so helpful if I had some idea of what voltages should
> be present here.
>
> I find that Q5 has .7V on it's base indicating that Q4 is clearly conducting
> but it doesn't seem right considering Q4 sees -3.5V on it's base. With the
> .7 on Q5's base I know it's on and that's verified due to the fact there is
> a steady 15V on the Gate of Q2. Now where I scratch my head is looking back
> at the FET source follower. With 15V on Q2's gate shouldn't Q7 be biased
> off?
> With R22 running warm it's obvious that Q7 is conducting. It's like an old
> who-done-it movie. I have several suspicious characters but without knowing
> what clues to look for I may make several false arrests.
> Now I know some peeps on this list have built this circuit. Now would be a
> good time for those peeps to chime in before I embarrass myself even further
> ;-)
>
> M. A. Ruberto