Well I have learned a lot today about the shaft encoder, and it is
working for the most part. I use a 1ms interrupt to read the encoder
inputs. I must have 3 readings the same to set the encoder valid flag.
That's a fairly short debounce because when you spin the knob the
transitions are 5ms apart. So this works pretty reliably. Now from the
spec and my observations, I am capturing 60 transitions per revolution
and there are 30 detents per revolution. That makes for an interesting
situation :-) I was hoping to make each detent bump the freq digit by 1.
But because of the mechanics you can get 1 3 or 2 2 transitions per
detent. If I choose 1 revolution for ten increments of the frequency
digit it is 3 detents and 6 transitions. Well I'm learning :-) I'll
figure it out tomorrow :-)
if ( Flag.encoder_valid ) {
if ( new_encoder != old_encoder ) {
if ( ( ( new_encoder >> 1 ) ^ old_encoder ) & 0x01 )
delta_freq = 10;
else delta_freq =
-10;
// if ( new_encoder ^ old_encoder == 0x03 )
delta_freq = 0;
old_encoder = new_encoder;
Flag.freq_change = 1;
}
}
and
//===================================================================
char old_encoder, new_encoder;
__interrupt void isrVmtim(void) {
static char counter = 0;
char temp;
MTIMSC_TRST = 1;
temp = PTBD & 0x03;
if ( temp == new_encoder ) ++counter;
else {
new_encoder = temp;
Flag.encoder_valid = 0;
counter = 0;
}
if ( counter >= 2 ) {
Flag.encoder_valid = 1;
counter = 2;
}
}
//===================================================================
--
Regards,
John
=========================================================
email: [EMAIL PROTECTED]
photos: http://www.flickr.com/photos/k5jhf/sets/
videos: http://www.youtube.com/profile?user=k5jhf
files: http://briefcase.yahoo.com/[EMAIL PROTECTED]
web page: http://www.geocities.com/[EMAIL PROTECTED]
call sign: K5JHF
=========================================================
