[sdiy] VCO and expo converter question
Scott Bernardi
sbernardi at comcast.net
Mon Nov 24 14:50:26 CET 2003
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.
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:
My og2 CEM3374 based, liner FM
(http://home.comcast.net/~sbernardi/elec/og2/og2_cem3374.html)
My SSM2033 based, linear FM
(http://home.comcast.net/~sbernardi/elec/og2/og2_ssm2033_vco.html)
My Lockbox synth VCO, based on Ian Fritz' sawtooth VCO, linear FM
(http://home.comcast.net/~sbernardi/elec/lockbox/lockbox_home.html#VCO)
ASM-1 (http://home.swipnet.se/cfmd/synths/friends/stopp/asm1vco.new.pdf)
Bergfotron (http://hem.bredband.net/bersyn/VCO/vco_complex.htm)
Electronics for Music (EFM) vco4d and vco4e
(http://www.ele4music.com/kits.html)
Ken Stone ASM-1 based (http://www.cgs.synth.net/)
Jim Patchell's temp compensated VCO
(http://www.oldcrows.net/~patchell/synthmodules/vco.html)
Roman Sowa's quad VCO has both linear and exponential FM inputs
(http://www.sowa.synth.net/modular.html)
RayLand VCO has a linear CV input (ac couple and it would be a typical
linear FM input) (http://atlas.csd.net/~rjwsoft/August2003VCO.html)
Ian Fritz' sawtooth VCO has both linear and exponential FM inputs
(http://home.earthlink.net/~ijfritz/sy_cir2.htm)
Rene Schmitz 555 VCO, both linear and exp FM
(http://www.uni-bonn.de/~uzs159/vco2.html) ; VCO3 linear FM
(http://www.uni-bonn.de/~uzs159/vco3.html), VCO4 linear FM
(http://www.uni-bonn.de/~uzs159/vco4.html)
Juergen Haible JH-3 dual VCO both linear and exp FM
(http://www.oldcrows.net/~jhaible/tonline_stuff/hj2vco.gif)
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.
harrybissell wrote:
>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:
>
>
>
>>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:
>>
>>
>>
>>>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:
>>>
>>>
>>>
>>>
>>>>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
>>>> http://home.comcast.net/~sbernardi/elec/og2/partsub_pots.html.
>>>>
>>>> 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.
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>--
>>Scott Bernardi
>>sbernardi at comcast.net
>>
>>
>>
>
>
>
>
--
Scott Bernardi
sbernardi at comcast.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20031124/5cb61ae9/attachment.htm>
More information about the Synth-diy
mailing list