Roland Andrag wrote:
>
>I have with success (after it was suggested on the Piclist) connected a
>piezo device between two port pins to get an effective 10 V swing.
>
>RB0 -----piezo speaker------RB1
>
>Start off with RB0 = 0 V, RB1 = 5 V. Now switch them to RB0 = 5 V, RB1 = 0.
>Do this about 4000 times a second for a nice noise.  A good idea is to play
>around with the frequency - most piezo speakers have a narrow frequency band
>in which they are really loud.
>
>I cannot remember what the verdict was on whether it is necessary to include
>series resistors...
>

Roland:
        For a 'purer tone' you might want to try pulsing your piezo with a
duty cycle of 2/3 (ie +ve for 2t, off for t, -ve for 2t and off for t). I
believe if you do a Fourier analysis of this waveform, you'll find there
are no 3rd and 9th harmonics, etc.(as well as no even ones due to the
symmetry) giving you a cleaner sine wave than you would have with the 50%
square wave that you're proposing.
        For experimental use, you'd likely get away with no series
resistor, but I wouldn't do it. The piezo element is one big capacitor that
you're charging/discharging so to limit the current add some resistance.
~1K is a good ballpark to try.




        Jim Nagy
        Elm Electronics
  ICs for Experimenters
http://www.elmelectronics.com/