On Sun, 9 Jun 2002, Olin Lathrop wrote:

> > ----/\/\/\/\------|-----
> >      3.3K         |
> >                   = .47uF
> >                   |
> >                  GND
> >
> > If I read the stuff I found on passive LPF design right, that should give
> > me a 102Hz cutoff, correct?
>
> Yup, 102.6 Hz, +-20% for the component <g>.

Well, seems to work OK.  Wow, crash course in audio LPF design.  I could
even do it with an op amp now (and still may).

> > The PIC (I'm using a 12CE674 for development, but will switch to 12C671
> > for real) looks for a low pulse on its input, which *should* be the peak
> > of the cycle of the string's fundamental frequency.  We start Timer 0
> > running and wait for the next pulse, indicating the next cycle.  If Timer
> > 0 overflows, we increment a high-byte counter, ending up with a 16 bit
> > counter for the cycle time.  We accumulate 32 of these samples, then
> > divide by 32 for an average (crude, yes, I know - I may work on that).
>
> I'm not sure I follow exactly what you are saying, but you can certainly use
> timer 0 with pulse interrupts to measure the period.  Using a CCP module in
> capture mode will make your life a lot simpler and will improve accuracy
> because you don't have to worry about being in a timer 0 overflow interrupt
> when a pulse comes in.  I've done a bunch of frequency/period measuring
> projects, and CCP modules are real handy for that.

Sorry, I did meander about the point a bit.  I'm using Timer0 because I
just want to use the 12C part, and it's a pretty simple task.  I'm 99%
sure it's all OK aside from the analog input stuff.

By the way, I just tried the above circuit, fed by a 1uF cap and connected
to the existing 330K/56K bias resistors.  Almost there, the amplitude is a
tad low but it looks a lot cleaner.  The PIC did lock onto the E string a
few times, and saw A and D a couple of times.  At this point I'm thinking
it might be a lot easier to connect to the amp output instead of the
guitar input.  Anyawy, looks like we're almost there.  Now I just have to
figure out how close to perfect pitch is close enough for a bass.

Dale

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu