--- crayola@optonline.net wrote:

> 60 updates per second is good enough for my purposes. 
> 
I believe it's a frame of 50 per second for it. Might want to check
this out to be sure.  The standard is I believe 50 with a 15ms center
point PW but I've been wrong before :)

> I understand, the speed of the servo is constant. I was planning 
> on having it simulate a slower or faster rate by giving it a
> specified
  Sic et Non, yes and no.  You CAN vary the speed of the servo as well.
It's not very easy to do however.  You will need a D/A convertor to do
it.  To vary the speed of the servo movement you have to change the
voltage powering it.  Most servo's operate with a minimum of 4V to a
maximum of 6V.  They publish the speed of servo's generally at 5 and 6V
you can compute the speed constant of the servo from this.  I don't
think varying the speed of the servo though will be entirely useful BUT
you could have a FAST and normal speed (IE 6V fast 5V normal)

> distance to move from the current point, so a rate in this case
> of the number of locations it should move to to get from point A to 
> point B. A loud sound might only have 1 location to move to (far
> away), were a softer sound might have 5 locations along the way that
> the 
> servo is instructed to move to (closer). Since the later takes a 
> longer period of time, the servo will move slower to the final 
> location versus the loud sound which will move quicker. 
> 
This works :)

> Oh I have no doubt that the servo is the weakest link in this 
> plan, but its cheap, small, and fits well in a skull. I believe 
> creating this circuit will create as realistic a mouth movement 
> as possible despite the servos limitations. 
> 
For computing volume, I suggest first rectifying the input signal using
a bridge to get your DC offset.  Then instead of using the linear
voltage use a logrythmic scale to determine your magnitude.  You can
find A-law and u-law linear to logrythmic scales on the net. I suggest
you use there interpreted values.  Are you planing on 8/10/12 bits of
resolution for your input sense? The other option is more complicated
that is putting a logrythmic amp in the signal path (and less useful).
You want to basically get your numbers in some sort of db scale (I
believe A-law matches this).  Then you can do it by 'eye', IE observe
adjust and make it look right.

Sounds like fun, now all you need are servo's to make the eyes move and
track using a pair of itnernet cams and you can have a nifty thing to
freak the neighbors out. ;)


=====
Stephen R. Phillips was here
Please be advised what was said my be absolutely wrong, and hereby this disclaimer follows.  I reserve the right to be wrong and admit it as well. :)

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
_______________________________________________
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist