[sdiy] adding PWM to a cmos clock?

Tom Wiltshire tom at electricdruid.net
Sun May 17 21:46:37 CEST 2009


On 17 May 2009, at 19:15, Dan Snazelle wrote:

>
> so in other words you cant feed it with a square...is that because  
> you need a continuous waveform? is it because the comparator will  
> just turn on and off exactly like the square wave and you need  
> something which will allow more transition?

Yes and yes. Scott had it exactly, and you've got it right too.
A ramp wave or a triangle wave are ideal for PWM because they're  
continuous. The fact that they're linear means you get a linear  
response between your PWM CV and the actual pulse width you get out.  
This is usually what you want. but Scott is right, the charging curve  
of a capacitor is also a continuous slope and will give you a  
reasonable PWM response. You'll find that the pulse width changes  
more quickly at lower voltages where the slope is steep, and less  
quickly higher up where it begins to flatten out.

T.





More information about the Synth-diy mailing list