MCF52235 doesn't start

[Kremer]

 Hi all

 We developed a board wich uses the MCF52235CAL60 and we are using codewarrior 6.4

 I'm testing the board with the Coldfire_Lite.mcp and i have a BDM hardware from PEMicro.

 I succefully load the firmware into the chip and run it with BDM attached to it. Until this point, nothing goes wrong. But when i remove the power, detach the BDM and turn it on again, nothing happens. I mean, coldfire doesn't start to run.

 I followed the Axiom schematic (only schematic i could find) to develop this board. On the RSTI pin i designed a 4K7 ohms pull up resistor like the schematic shows.

 What can i possibly doing wrong? Any suggestions are much appreciated.

 Thank you

 Kremer

[J2MEJediMaster]

If the code downloads and runs on the board courtesy of the debugger, it sounds like you're almost there. Only thing I can think of is that the debugger typically disables the MCU's software watchdog timer. You might have some lengthy operation going on that exceeds the limits of the watchdog timer, and so the watchdog interrupt fires, or the MCU resets. You should be able to check if this is the problem by adding initialization code that disables the watchdog timer. HTH.

---Tom

[Kremer]

Hi Tom

 Thank you for your help, but the watchdog is already disabled on Coldfire_Lite.mcp, file MCF5223_sysinit.c

 There is a function called mcf_5223_wtm_init() wich already clears CWCR.

 Can RCON pin be the cause? Wich is the recommended signal level for it?

 Thank you

[Kremer]

 Sorry if i wasn't clear on my first post. My problem is i can only run the code when BDM (from PEMicro, not the one wich comes on the demo board) is attached to my own development board. If i just program the chip, turn the power of, detach the BDM and turn the power on again, coldfire doesn't run the code.

 Any help is much much welcome.

 Thank you

 Kremer

[J2MEJediMaster]

Sorry, I'm more of a software guy and can't help you on those pin signals.

Have you tried downloading a simple CodeWarrior demo program onto the board and see if it runs successfully in stand-alone mode? If it does, something's wrong with your program's reset vector in the vector table. If the demo program fails to run, then possibly there's an unknown interrupt (which has no handler) firing.

---Tom

[Kremer]

 This couldn´t be a software issue since i´ve tried the same firmware on demo kit and it works, besides, any small hello world code didn´t run standalone either.

 Actually the issue was on 3.3V voltage regulator, wich isn´t as good as i expected so i had to add some extra filter circuitry to eliminate any possíble ripple that it could generate and the problem was gone. Now this excelent machine can run free as a hiena in the desert.