cemments below...

Olin Lathrop wrote:

> Bob Barr wrote:
>
>>On the 16-series devices, so does avoiding the use of bit set/clear
>>operations on ports by rolling your own read-modify-write operation.
>>Reading the whole port in, modifying W as you need to, and then
>>writing the whole thing back works well in situations like this.
>
>
> That just creates your own manual version of the read-modify-write problem.
>
> The problem is that when reading the port bits, you get what the voltage is,
> not what the you are trying to set it to.  The BSF and BCF instructions do a
> read, alter the indicated bit, then write the result back to the whole port
> register.  This is exactly what you are doing the hard way.
>
> The problem can be avoided by keeping a shadow copy of the port register
> desired value.  To change a bit, you first change it in the shadow register,
> then write the whole shadow register value to the port.
>
> In practise the read/modify/right issue isn't as big a deal as the endless
> discussions about it might lead you to believe.  I have used shadow
> registers on isolated occasions, but most of the time it's easy to get
> around the issue as long as you're aware of it and understand it.  If your
> external circuit holds the PIC pin in the opposite state the PIC is driving
> it for more than a few cycles, then there is likely something else wrong
> anyway.

Well-stated on all points, Olin.

One unexpected advantage of shadow registers is that you now have a
reliable expected value, which allows you to check for external faults,
such as solder or trace shorts, or defective peripherals. Just read back
what you wrote; if it doesn't match, something is hanging up the pin.
Then pass the info onto the tech via on a diagnostic pin (manchester
mark code, RES232 serial, etc).

Makes the production line troubleshooter's job MUCH easier.

--Bob


--
              --------------
                Bob Axtell
        PIC Hardware & Firmware Dev
          http://beam.to/baxtell
              1-520-219-2363

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu