Help: LPC932 hang after ISP programming

[Marcus Tan]

I managed to program the LPC932 successfully via the ISP method using FlashMagic (TxD and RxD) and serial port.  I did not use the RST line. Code was running well.

After I added some code and download via ISP, the RxD and TxD line of the LPC932 are always low and it does not recognise any break detect.  Programming via FlashMagic through the serial port is not successful (due to the always low RxD line) since it cannot establish any baud rate.

What should I do now to get the LPC932 to ISP mode again?  Setting the RST of the LPC932 to 3 low pulses?  Has anyone got a ready program to do this ?

My LPC932 is already soldered to my PCB and removing it will be the last resort.  Please HELP!!

[Andy Ayre]

If your program is forcing the RxD low all the time then your only option is the reset pulses. Flash Magic can generate the pulses but you need to build the same RS232 circuit that is found on the Keil MCB900 board. You could build this on a seperate piece of hardware then connect it to the power supply and RST on your board perhaps? For the circuit diagram you need look in the application note on the Flash Magic web page regarding controlling the device using handshaking signals.

[Eduardo de Mier]

I have a similar problem.
When I program the LPC932 in my application once it is not possible to program it again. The cause is a pin that my application sets to output/high. An external pullup resistor supplies so much current, that the DTS/RTS controlled switching off of the power does not work because the LPC supplies its own power through the high output and the external pullup. Removing the pullup momentarily solves the problem.
I suggest that your SW may include a reset pulse before  (!) power is switched off so that the IO-pins are switched back to inputs. As far as i can see on my storage scope there is a reset pulse but it starts at the same time as the power-off-pulse.
My SW is the newest version.
Regards,
Eduardo

[Andy Ayre]

I'm not quite sure I understand... after power up all the pins should switch to inputs. Are you saying they don't?

[Eduardo de Mier]

I am trying to say that there is no valid reset pulse before you switch off power. In my case and maybe some other people's too power will not go down enough if there is still some pin switched to output and current flows by some pullup through the pin into VCC.
A reset pulse before (!) switching power off would reset all pins to inputs and solve the problem.
Eduardo

[erik malund]

A reset pulse before (!) switching power off would reset all pins to inputs and solve the problem.
Pray explain
1) what effect does that pulse have to do with the state of the pins with power off
2) when you turn power back on a reset must be generated.

I see no effect, whatsoever, from the pulse you suggest.

Erik