Cwarrior: RTI timing off

Started by Stewart Prince November 17, 2003
We're using codewarrior to simulate code written for the D60A. We set
up the real-time interrupt to run at 8.196 ms using the following logic:

Assuming a crystal frequency of 8MHz, the Xclock should be 1/2 this rate
with the slow clock set to 1(default). Writing 0x11 to the
RTR2-RTR1-RTR0 bits should be a divide by 2^15 or

(2e6)/(2^15)2 hz or a period of .008 milliseconds.

However, when I sumulate this using the Codewarrior debugger, with an
8MHZ crystal (CPU = 4MHz), the RTI period is 16.192 milliseconds. How
can this be??

Stewart Prince
Professor, MEch Eng
CSUN



Hello Stuart.

I will check the debugger and give you some new asap about this issue.

Just to make sure we use the same setup, this is mine:

by default (after reset) SLOW=0 => SLWCLK=EXTALi
as XCLK=SLWCLK/2, XCLK=EXTALi/2
if EXTALi = 8 MHz, XCLK = 4 MHz

4E6 Hz / 2^15 = 122 Hz => period of 8.196 ms

I guess 2e6 here below in your equation is a typo.

N.B: the SLOW register (address $3E) is not simulated with "MC68HC912D60A
SAMPLE" generic simulation.

Regards,
Gilles At 08:55 PM 11/17/2003, you wrote:
>We're using codewarrior to simulate code written for the D60A. We set
>up the real-time interrupt to run at 8.196 ms using the following logic:
>
>Assuming a crystal frequency of 8MHz, the Xclock should be 1/2 this rate
>with the slow clock set to 1(default). Writing 0x11 to the
>RTR2-RTR1-RTR0 bits should be a divide by 2^15 or
>
>(2e6)/(2^15)2 hz or a period of .008 milliseconds.
>
> However, when I sumulate this using the Codewarrior debugger, with an
>8MHZ crystal (CPU = 4MHz), the RTI period is 16.192 milliseconds. How
>can this be??
>
>Stewart Prince
>Professor, MEch Eng
>CSUN >
>-------------------- >
>">http://docs.yahoo.com/info/terms/