<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
<title></title>
</head>
<body text="#000000" bgcolor="#ffffff">
Really? Almost every VCO circuit I have seen has a linear FM input. I
guess you just would use a regular 1v/octave input to do exponential
FM, just attenuate the signal. <br>
<br>
harrybissell wrote:<br>
<blockquote type="cite" cite="mid3FC1778C.71C96D05@prodigy.net">
<pre wrap="">hmmm... I'd have said that expo FM is a lot more common than linear
FM...
(especially in commercial synths)... but wht not have both in this case
?
There are advantages and disadvantages to both linear and expo fm :^P
The nonlinearity of a 'loaded' pot is often not noticible unless you
have some
panel graphics that make it obvious... Making the summing resistors
bigger
should be OK.
H^) harry
ryan wrote:
</pre>
<blockquote type="cite">
<pre wrap="">nice tutorial. thanks!I'm sure glad I asked here before getting the
PCB made. I thought that Exponential FM would have been more typical,
thats why I didn't add Linear FM. Yesterday was actually the first
time I tried putting something into the FM jacks on my synth. well,
Before yesterday, I had only 1 VCO. the Idea for this VCO is to be a
simple, fit on a small PCB, and take up little panel space. I have
Oakley Sound VCOs for my main oscillators, but I'd like to get this
one to stay in tune with those. for the pots. I'm using 100K pots on a
100K load. arg. I already spent a bunch of $ on those fancy spectrol
100k pots. I think it'd be easier to just double the values of all the
other resistors. I think that would atleast reduce the warping enough
and the 10K trimmer for V/Oct would be more centered around what the
value should actually be. I looked at that java applet, It looks
pretty nasty the way I have the circuit now. about the coarse
adjustment, I suppose +/-5 octaves would be alot more useful. meant to
change that. thanks !ryan
I've got a tutorial on pots at
<a class="moz-txt-link-freetext" href="http://home.comcast.net/~sbernardi/elec/og2/partsub_pots.html">http://home.comcast.net/~sbernardi/elec/og2/partsub_pots.html</a>.
In general to avoid loading effects, you want your voltage
divider pots to be 5 to 10 times less than the resistance
that loads them. So for a 100K input resistor, use a 10K up
to 50K pot. To see the effect of loading, I have a link to
Chris List's java applet that plots loading effects.
You might want to decrease R2 a bit to give yourself more
Coarse freq adj range. As you have it, you'll only get 6
octaves. Using a 300K will give you a 10 octave tuning
range.
The FM input you have is feeding the exponential input -
linear FM is more typical. To get linear FM, move the end of
R28 to pin 6 of U7B. Also, you'll want to change the value
of R28 to something above 1M.
I would also have a second 1v/octave input - duplicate the
R27 input. I also find having a front panel "LFO switch"
useful. This would sum a large negative voltage into your
input summer, which would switch the frequency way down.
-15v through 300K into the summing amp with switch you down
5 octaves.
The 100pF compensates for the extra phase shift running
IC1A in the feedback loop of the opamp, and it is to prevent
oscillation. I also use 100pF because everybody else does.
</pre>
</blockquote>
<pre wrap=""><!---->
</pre>
</blockquote>
<br>
<pre class="moz-signature" cols="72">--
Scott Bernardi
<a class="moz-txt-link-abbreviated" href="mailto:sbernardi@comcast.net">sbernardi@comcast.net</a></pre>
</body>
</html>