The following message was inadvertently sent privately to me; I'm
forwarding it to the PICLIST.

-Andy

------- Forwarded Message Follows -------
Date:          Fri, 31 Jan 1997 14:19:45 -0600 (CST)
To:            fastfwd@ix.netcom.com
From:          Todd Peterson <elab@netins.net>
Subject:       Re: Re[2]: 12C508 too fast?

At 12:01 PM 1/31/97 -0800, you wrote:

>Todd:
>
>Does the CCS compiler have explicit support for the 12C508?  If not,
>and if it's just generating 16C5x code, it may be overwriting the
>OSCCAL value as part of its built-in PORTA initialization (on the
>12C50x, register 5 holds the OSCCAL value; on the 16C5x, that
>register holds the PORTA value).

Hopefully someone from CCS is on this list; I just got the compiler a
few days ago and it does generate 12C508 code; I (of course) checked
its assembly output and the first and last values of memory are
correct (load constant, put it in memory.

However, Andy, you do raise an interesting point; when I got the
compiler it wouldn't work with the 12C508 and I had to modify the
12C508.H file to switch it to port B, as per CCS's reccomendation.  I
bet they missed something in the switchover and something is being
written to port A (where OSCCAL is stored instead of port A).  They
are supposed to have a new version for updating on the web, but their
site is down.  Anyone with CCS having luck setting the OSCCAL value?

In the mean time, I think I'll check their code and see if they mess
with the non-existant port A.  CCS is a good compiler and has a lot of
features, but every project I have done I have had to download a
bug-fix.

-Todd