19.6608MHz instead of 20MHz

Will Reeve · Joined: 30 Oct 2003 Posts: 209 Location: Norfolk, England

Hi,
I am limited for space (physical) and a perfect size crystal is available at 19.6608MHz. It's a USB project using the USBN9603 (which is using it's own 24MHz crystal).
I was wondering how good the CCS compiler is at working out it's delay_us etc, if I set the delay(clock=19660800)? I can't see any other issues for not running at 20MHz as I wanted?

Keep well,

Will

SherpaDoug · Joined: 07 Sep 2003 Posts: 1640 Location: Cape Cod Mass USA

CCS works very well with odd clock frequencies. I use all sorts of clock frequencies to make frequency divisions easy.
_________________
The search for better is endless. Instead simply find very good and get the job done.

dyeatman · Joined: 06 Sep 2003 Posts: 1934 Location: Norman, OK

I use 18.432 in many of my designs as an exact multiple for accurate (0% error) baudrates.

Darren Rook · Joined: 06 Sep 2003 Posts: 287 Location: Milwaukee, WI

I thought the golden rule was if you sum up all the numbers and then divide equally by three, you have 0% error for baud rates.

Guest · Posted: Fri Sep 24, 2004 10:45 pm

i just wanted to point out the digits in 18.432 add up to 18, which is evenly divisible by 3 (18%3 = 0). So you just answered your own question. (I have no idea if that golden rule is true or not, just doing some arithmetic).