Re: A89: Re: 89 timer


[Prev][Next][Index][Thread]

Re: A89: Re: 89 timer




> Hmm... The ON key actually triggers an interrupt, as we all seem 
> to agree on.  This means that the processor doesn't need to 
> actively check for it.  After getting a HALT command, the processor 
> goes into hibernation until it gets an interrupt. (the display has 
> nothing to do with that process, but for long term HALTs, the 
> screen is shut down by the preceding code... another totally 
> unrelated matter, sorry...)  The approximately 200Hz interrupt that 
> is normally running gets masked out (or maybe just plain turned 
> off, I can't remember... its been a while) when the calculator is 
> "turned off" so that the only thing to wake the processor back up is 
> the ON key interrupt.  There is no 200 times a second scan for the 
> ON key or anything.  The processor is effectively off, drawing very 
> little current until that ON key is pressed.  A realtime clock, on the 
> other hand, would require that the ~200 Hz interrupt be left on or 
> unmasked, in which case the processor _would_ be powered up 
> and down 200 times a second.  Of course, that interrupt is so 
> dependent on battery voltage that a clock isn't really practical 
> anyway, so its not a problem. :-)

Well, that's what I thought too, but according to A-86, who has flamed
me twice for thinking this, there is no DI anywhere in ROM.