selfbus.org

Hi.
I've made hardware bug. In ISP-ICP converter i connected two resets together. Later I noticed that reset from target processor is controlled via GPIO from converter controller.

Hello everyone,
I would like to share my work results to bring china uC working, after spedning some time playing with.

uC state shipped from china :
- no ISP bootloader
- some security bits were set ( thanks to ICP bridge it was possible to read the bits )
- not possible to erase all pages because of active security bits.

Security bits programmed once, are not possible to erase by ISP or ICP.


I bought TOP2013 programmer which is provided with TopAll8 software. It shall have possibility to program parallelly LPC922 and "global erase" order, shall also erase security bits.
"It Shall" but there is a bug in this programmer or in TopAll software.
Programmer shall send the "global erase" order and then pool for the status and observe BUSY bit. ( 0xF0 changes with 0x70 ).
But the programmer, after 3ms of pooling, breakes the power where status is still 0xF0 and return SUCCESS.
Security bits in this case were not erased.

Ok. Hmm... So I took Nucleo board, used 14 I/O ports and I have generated parallel sequence to send "global erase".
That has worked, and status BUSY has been changed to 0x70 after around 7ms. Security bits were cleared. Small success.

Then, with TOP2013 I have programmed blinky.hex bootloader and it has also worked.
Bootloader version has changed to 02/02 from oryginal 06/06, what made my microcontrollers like they should be.

ISP works now and I am able to program uC with using Programmer from SelfBus page.

If I find developer of TopAll software, I will ask him to fix the problem. ( break the power after min 10ms or pool for status )

For now, I don't know any other software supporting TOP2013 to test with.

Regards,
Rafal

This is kind of cool.
I really like the idea since i played around with a Galep programmer many years ago to flash the LPCs via ICP.
A nice side-effect is that there is a couple of bytes more that could be used without the serial loader in place -- as it's very tight to squeeze the firmware in for some devices and sometimes things needed to be excluded.

I somehow liked these old LPCs since the code was so small, on the other hand the limits of the old silicone gave us a lot of headache at the end. When the Lib needed to be patched (long frame support was needed with the later ETS versions) I would have wished for a few bytes more...