12F675 problem with OSCCAL and SLEEP

[itchy_frog]

Hi Folks,
I hope someone can help with this. I'm running code on a 12F675 that uses the SLEEP command, then the device is woken from sleep by a reset on the MCLR pin. When I use the standard code to get the OSCCAL value (CALL 0x3FF), the program runs fine but on wake from sleep, the device runs for a second or so then dies/sleeps. If I read the OSCCAL value and hard-write it into the program, the whole thing runs as it should and wakes properly from SLEEP. I'm programming with the K8048 using PicProg2009, which I think might be part of the problem - certainly there is only one shot at programming then it deletes the OSCCAL value (however I now know to always record it first).

I'd also welcome advice on a more robust, cost effective programmer.
Kind regards,
Andrew

 

[nickelflippr]

The **broken link removed** can auto regenerate the lost OSCAL value for the 12f675 type devices.

 

[Pommie]

I suspect you are leaving the RAM in a state that causes a crash after the sleep. Can you add a Clear_RAM routine to the initilisation code to see if that fixes it.

Mike.

 

[itchy_frog]

Problem solved

Thanks for the input folks - problem now solved. The SLEEP command was in a routine at the bottom of the code. On wake-up the code was sent to the RESET section, missing the START code where the OSCCAL value is retrieved and put into RAM. It simply needed a GOTO following SLEEP to send the program back to START, now it functions perfectly. No more reading the chip first, hard-writing the OSCCAL into the program code and individually programming each chip!