<!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">
The CEM and SSM based VCO's I've done (see links below) all had linear
FM. I've double checked the datasheets for the CEM3340, CEM3374, and
SSM2033, all chips I've built circuits around, and they says that the
FM inputs are Linear FM inputs. Maybe some commercial synths didn't
build their FM capability using these inputs.<br>
I did a short survey of VCO designs (mainly by looking at the
schematics of DIY sites I had links to) to see whether they provided
linear and/or exponential FM inputs:<br>
<br>
My og2 CEM3374 based, liner FM
(<a class="moz-txt-link-freetext" href="http://home.comcast.net/~sbernardi/elec/og2/og2_cem3374.html">http://home.comcast.net/~sbernardi/elec/og2/og2_cem3374.html</a>)<br>
My SSM2033 based, linear FM
(<a class="moz-txt-link-freetext" href="http://home.comcast.net/~sbernardi/elec/og2/og2_ssm2033_vco.html">http://home.comcast.net/~sbernardi/elec/og2/og2_ssm2033_vco.html</a>)<br>
My Lockbox synth VCO, based on Ian Fritz' sawtooth VCO, linear FM
(<a class="moz-txt-link-freetext" href="http://home.comcast.net/~sbernardi/elec/lockbox/lockbox_home.html#VCO">http://home.comcast.net/~sbernardi/elec/lockbox/lockbox_home.html#VCO</a>)<br>
<br>
ASM-1
(<a class="moz-txt-link-freetext" href="http://home.swipnet.se/cfmd/synths/friends/stopp/asm1vco.new.pdf">http://home.swipnet.se/cfmd/synths/friends/stopp/asm1vco.new.pdf</a>)<br>
Bergfotron (<a class="moz-txt-link-freetext" href="http://hem.bredband.net/bersyn/VCO/vco_complex.htm">http://hem.bredband.net/bersyn/VCO/vco_complex.htm</a>)<br>
Electronics for Music (EFM) vco4d and vco4e
(<a class="moz-txt-link-freetext" href="http://www.ele4music.com/kits.html">http://www.ele4music.com/kits.html</a>)<br>
Ken Stone ASM-1 based (<a class="moz-txt-link-freetext" href="http://www.cgs.synth.net/">http://www.cgs.synth.net/</a>) <br>
Jim Patchell's temp compensated VCO
(<a class="moz-txt-link-freetext" href="http://www.oldcrows.net/~patchell/synthmodules/vco.html">http://www.oldcrows.net/~patchell/synthmodules/vco.html</a>)<br>
Roman Sowa's quad VCO has both linear and exponential FM inputs
(<a class="moz-txt-link-freetext" href="http://www.sowa.synth.net/modular.html">http://www.sowa.synth.net/modular.html</a>)<br>
RayLand VCO has a linear CV input (ac couple and it would be a typical
linear FM input) (<a class="moz-txt-link-freetext" href="http://atlas.csd.net/~rjwsoft/August2003VCO.html">http://atlas.csd.net/~rjwsoft/August2003VCO.html</a>)<br>
Ian Fritz' sawtooth VCO has both linear and exponential FM inputs
(<a class="moz-txt-link-freetext" href="http://home.earthlink.net/~ijfritz/sy_cir2.htm">http://home.earthlink.net/~ijfritz/sy_cir2.htm</a>)<br>
Rene Schmitz 555 VCO, both linear and exp FM
(<a class="moz-txt-link-freetext" href="http://www.uni-bonn.de/~uzs159/vco2.html">http://www.uni-bonn.de/~uzs159/vco2.html</a>) ; VCO3 linear FM
(<a class="moz-txt-link-freetext" href="http://www.uni-bonn.de/~uzs159/vco3.html">http://www.uni-bonn.de/~uzs159/vco3.html</a>), VCO4 linear FM
(<a class="moz-txt-link-freetext" href="http://www.uni-bonn.de/~uzs159/vco4.html">http://www.uni-bonn.de/~uzs159/vco4.html</a>)<br>
Juergen Haible JH-3 dual VCO both linear and exp FM
(<a class="moz-txt-link-freetext" href="http://www.oldcrows.net/~jhaible/tonline_stuff/hj2vco.gif">http://www.oldcrows.net/~jhaible/tonline_stuff/hj2vco.gif</a>)<br>
<br>
<br>
I am not arguing the merits of linear vs. exponential FM, just that in
circuits I've seen, a linear FM input seems to be more common than an
exp FM input. Part of that may be just that you can use any 1 v/octave
exponential input to do exponential FM with, and so a dedicated
exponential FM input was not created.<br>
<br>
harrybissell wrote:<br>
<blockquote type="cite" cite="mid3FC1940D.842DE1F0@prodigy.net">
<pre wrap="">Hi Scott (et al)
I was gonna offer to put on my flamesuit, since linear / expo FM has
been
such a ..er... hot topic ??? :^P on s-diy.
Of course almost every expo VCO has the ability for linear FM as well...
I'm thinking
mostly of commecial synths (maybe with SSM or CEM chips ?) as being
mostly
expo FM. Since you get linear for free... it depends on whether the
designer brought
out the points to allow linear FM or not. As a rule... the expo FM is
there (its just the
V/oct attenuated, as you said).
OK I'll have to hook up my linear input on my old Aries VCO. I never
used it, since I
only had one oscillator at the time and used that jack for an octave
divider output instead.
Now that I have lots of oscillators I could try it :^P
H^) harry
Scott Bernardi wrote:
</pre>
<blockquote type="cite">
<pre wrap="">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.
harrybissell wrote:
</pre>
<blockquote type="cite">
<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>
<pre wrap="">--
Scott Bernardi
<a class="moz-txt-link-abbreviated" href="mailto:sbernardi@comcast.net">sbernardi@comcast.net</a>
</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>