Re: 12F675 Startup Problems

Started by Eirik Karlsen December 8, 2007
Well...as I've said before serial coms should always be done with x-tal or maybe resonator osc.
Trying to get a RC osc to run within 5 - 10% in accuracy and stability just isn't possible
in practice...you might get it to work but when temperature changes, or a wire gets too
close to chip or tracks, or whatever that can pull the osc happends you'll get tagged.

kg4pid wrote:

> To make it work I had to change the the value that gets moved into
> OSCCAL. Since the serial routine that I used had been around awhile
> I can only assume that the problem is in the Pic oscilator. Don't
> know what to do to fix it though as I don't want to have to
> calibrate each pic chip.
>
> Max
>
--
*******************************************
VISIT MY HOME PAGE:

LAST UPDATED: 23/08/2003
*******************************************
Regards
Eirik Karlsen
I have run serial comms at 9600 baud and lower using Microchip internal
RC oscillators quite successfully for years. The secret is to keep your
dialog length very short, one or two bytes at a time so that clock drift
won't affect your data much.

DLC

Eirik Karlsen wrote:
> Well...as I've said before serial coms should always be done with x-tal
> or maybe resonator osc.
> Trying to get a RC osc to run within 5 - 10% in accuracy and stability
> just isn't possible
> in practice...you might get it to work but when temperature changes, or
> a wire gets too
> close to chip or tracks, or whatever that can pull the osc happends
> you'll get tagged.
>
>
>
> kg4pid wrote:
>
>> To make it work I had to change the the value that gets moved into
>> OSCCAL. Since the serial routine that I used had been around awhile
>> I can only assume that the problem is in the Pic oscilator. Don't
>> know what to do to fix it though as I don't want to have to
>> calibrate each pic chip.
>>
>> Max
>>
> --
> *******************************************
> VISIT MY HOME PAGE:
>
> LAST UPDATED: 23/08/2003
> *******************************************
> Regards
> Eirik Karlsen
>

--
-------------
Dennis Clark TTT Enterprises
www.techtoystoday.com
-------------
--- In p..., dlc wrote:
>
> I have run serial comms at 9600 baud and lower using Microchip internal
> RC oscillators quite successfully for years. The secret is to keep
your
> dialog length very short, one or two bytes at a time so that clock
drift
> won't affect your data much.
>
> DLC

Hi DLC,

I must stress here that it is not the length of the packet that
compensates for the use of an inaccurate oscillator but the start bit
of each character sent. The leading edge of the start bit is what
causes the receiver to reset its timer. This happens for every
character that is sent. Keeping the packets short has no effect on
this re-sync.

Regards
Sergio Masci

http://www.xcprod.com/XCSB
.