Re: Windows Real-time

Art,

Are you just speeding up the 8254 timer chip, and then calling the
"normal" INT 8 (IRQ 0) timer interrupt code every 500 times? I was
trying 1ms and .1ms rates under DOS, but now I think my 33Mhz '386 may
have been too slow to allow good results. Do you bump a software
counter in the new interrupt code, or directly update the parallel port
registers? or something else?

(14318180.0/12.0)/65536.0 = 18.2065 (54.9 ms normal)
(14318180.0/12.0)/1190.0 = 1002.67; ( 0.9977 ms)
(14318180.0/12.0)/119.0 = 10023.24; ( 0.0998 ms)

(14318180.0/12.0)/145.0 = 8228.84; ( 0.000122 ms) ???
(14318180.0/12.0)/146.0 = 8172.77; ( 0.000122 ms) ???

thanks,

Alan KM6VV


Art Fenerty wrote:

>
> Phil:
>
> I am using int8 with 8192 pulses/sec from the clock chip. The VXD is
> about half & half asm and c++. It might
> be that the k6 cannot allow the clock to interrupt at that rate. Thanks for
> the test though, it gives me something to
> focus on. Although I use the RTC I also use the TSC for
> subtiming inside the interrupt.
>
> Art

<snip>