EmbeddedRelated.com
Forums
Memfault State of IoT Report

LPC2136 JTAG issue

Started by Robert Lacoste November 14, 2005
Hi guys,
A strange problem on a custom board built using a Philips LPC2136 ARM7TDMI 
chip, may be someone will have an idea... :

- Software developped and tested on a Keil MCP2130 eval board (with a 
LPC2138), programmed and debugged with a Ulink JTAG probe, no problem
- Recompiled for the target chip (LPC2136), tested on the eval board (as 
only the flash memory is more limited), no problem
- JTAG probe connected to the target board, flash program & verify cycle 
executes OK, however the firmware is apparently not working at all
- Debug mode with the JTAG on the target board (step by step starting from 
reset) : very strange, the instructions displayed are the one from our 
firmware, but their behaviour is not consistent with the instruction (ex a 
jump doesn't change PC, etc). Seems like the debugger memory and the target 
memory are not in synch...
- Tried with a slower speed on the JTAG, same behaviour

I'm 99% sure there is something wrong on the JTAG itself, but I've checked 
10 times and it seems well wired. Moreover I don't understand how a program& 
verify cycle could be ok if the jtag was wrong... Or something linked to the 
LPC rom boot code ?

Any idea will be appreciated !

Friendly,
Robert 


"Robert Lacoste" <see-my-email-at@www.alciom.com> a &#4294967295;crit dans le message de 
news: 43786cdc$0$18329$8fcfb975@news.wanadoo.fr...
> Hi guys, > A strange problem on a custom board built using a Philips LPC2136 ARM7TDMI > chip, may be someone will have an idea... : > > - Software developped and tested on a Keil MCP2130 eval board (with a > LPC2138), programmed and debugged with a Ulink JTAG probe, no problem > - Recompiled for the target chip (LPC2136), tested on the eval board (as > only the flash memory is more limited), no problem > - JTAG probe connected to the target board, flash program & verify cycle > executes OK, however the firmware is apparently not working at all > - Debug mode with the JTAG on the target board (step by step starting from > reset) : very strange, the instructions displayed are the one from our > firmware, but their behaviour is not consistent with the instruction (ex a > jump doesn't change PC, etc). Seems like the debugger memory and the > target memory are not in synch... > - Tried with a slower speed on the JTAG, same behaviour > > I'm 99% sure there is something wrong on the JTAG itself, but I've checked > 10 times and it seems well wired. Moreover I don't understand how a > program& verify cycle could be ok if the jtag was wrong... Or something > linked to the LPC rom boot code ? > > Any idea will be appreciated ! > > Friendly, > Robert
PROBLEM SOLVED ! Just in case others could have the same issue : P0.14 pin should be HIGH at reset, if not the LPC launches the on-chip ISP code and doesn't launch the application, and then the debugger is lost... Yours, Robert

Memfault State of IoT Report