LPC54608 returns invalid device ID

Started by H3llow33N, September 08, 2017, 06:16:50 AM

Previous topic - Next topic



I have a board with LPC54608 and it returns 00054000 as device ID.
If I try to read back the flash memory I get Operation failed error.
My main problem is that I cannot debug it anymore over SWD and I tried to erase the flash using FlashMagic.
In terminal I can give part of ISP commands and it looks like it is working properly.

Any idea why the MCU returns invalid ID?

Andy Ayre

Seems like it might be an engineering sample rather than a production part.

What error do you get when erasing? Have you tried lowering the baudrate?

Embedded Systems Academy, Inc.
support at esacademy dot com



here we have the same problem with the device ID signature on a development board with the same processor.
Flash magic does not read the expected device ID for the device but 0x54000.
The bootloader version returned is 19.1.

Communication with the device is done through a serial port (not the USB serial port on the board).

  • Baud rate : 9600
  • Interface : None (ISP)
  • Oscillator : <left blank>
I have tested the following command, each time overruling the device ID match warning:

  • Blank check succeeds with no error.
  • Erasing the flash succeeds with no error. (Flash is than reported as blank)
  • Display flash memory reports "operation failed"
  • Programming the flash ends with a verify error at address 0x1

When the flash is erased and than programmed, the firmware seems to work normally.

When monitoring the serial connection, the verification done is the same as when displaying flash memory.
As described above, the flash read fails.

Andy Ayre

Verification and read is fixed in version 10.82.

Embedded Systems Academy, Inc.
support at esacademy dot com


Hi Andy,

thanks for the reply.

Downloaded and checked version 10.82.
Besides reporting the wrong device ID, programming and verifying is successful.

There is a verification error at address 0x1FC, the Code Read Protection location, so I think that's an issue with our makefile.
(After removing the line with this address from the .HEX file, verify does not result in a error.)

With kind regards,