No Console output after NOR flashing.

Hello,

i got a Mini2440 Board with 256MB NAND where the bootloader was lost in
NOR.

NOR Chip is S29AL016J70I02

I flashed supervivi with H-JTAG 2.1 and NOR Type S29AL016JB.

Flashing was successfully done. The USB Port is available in Linux and
Windows and tested as ok by DNW. But there is no console output, even when
i connect a TTL/RS232 converter for Mini2440 at CON1.

The LEDs are as usual when booting NOR: LED1 blinking and LED3 always on.

I tested several hjf Configurations with the same result. The actual
configuration is:

FLASH SECTION:
SPANSION
S29AL016JB
MEMORY SECTION:
16-BIT X 1-CHIP
0x00000000
0x40000000
XTAL SECTION:
NULL
TCK SECTION:
0
0
SCRIPT SECTION:
SoftReset+ + +
Setmem+32-Bit+0x53000000+0x00000000 
Setmem+32-Bit+0x4A000008+0xFFFFFFFF
Setmem+32-Bit+0x4A00001C+0x000007FF
Setmem+32-Bit+0x53000000+0x00000000 
Setmem+32-Bit+0x56000050+0x000055AA
Setmem+32-Bit+0x4C000014+0x00000007
Setmem+32-Bit+0x4C000000+0x00FFFFFF
Setmem+32-Bit+0x4C000004+0x00061012 
Setmem+32-Bit+0x4C000008+0x00040042
Setmem+32-Bit+0x48000000+0x22111120 
Setmem+32-Bit+0x48000004+0x00002F50 
Setmem+32-Bit+0x48000008+0x00000700 
Setmem+32-Bit+0x4800000C+0x00000700 
Setmem+32-Bit+0x48000010+0x00000700 
Setmem+32-Bit+0x48000014+0x00000700 
Setmem+32-Bit+0x48000018+0x0007FFFC 
Setmem+32-Bit+0x4800001C+0x00018005 
Setmem+32-Bit+0x48000020+0x00018005 
Setmem+32-Bit+0x48000024+0x008E0459 
Setmem+32-Bit+0x48000028+0x00000032 
Setmem+32-Bit+0x4800002C+0x00000030 
Setmem+32-Bit+0x48000030+0x00000030 
PGMOPTION SECTION:
ADDON SECTION:
NULL


TCK Section -1, -1 and/or removing Softreset didn't succeed.

Is the still a problem with supervivi possible or is it an hardware
problem?

Any ideas?

Thanks in advance,

Titus

How are you transferring the terminal output to CON1 from the DB9
connector? In /etc/inittab?

Have you looked at the boot messages using the DB9 connector, mini2440
RS232 cable and a PC running minicom or similar?

Hi Dave,

according to the schematics CON1 is the TTL output of COM1 (COM1 leads to
DB9 connector, but it is dead). I used CON1 and the converter breakout to
proof if the MAX3232 of COM1 is broken.

On another board, a Micro2440, the console output is ok via CON1 and the
breakout.

I tested console on Linux with Minicom and on Windows with DNW. The USB
Port of the Mini is available under Linux and Windows. DNW report it as OK.
So i conclude that the boatload is working partially or the bootloader is
working fully but the MAX3232 or some other hardware has a failure that is
killing the signal of TTL which the MAX3232 is sourced from.

So i asked if someone has experienced a partially broken bootloader because
i read in forum that other people had problems with flashing this NOR type.

The TTL is sourced directly by the CPU. Maybe the UART of CPU is broken.
But how to test it?

Yes, it is more helpful to look at the schematics!  I misunderstood:

> The USB Port is available in Linux and Windows and tested as ok by DNW.

specifically the "and tested ok by DNW" part.

Is this where the experts start using JTAG to try and see what is going on
inside the part?

Did this board use to work?  Did you change anything and now it doesn't? 
And is this first time you have had to JTAG, whatever bootloader file you
have into it?

Does the micro2440 have the same flash?

The USB port is provided by the bootloader. If you erase the bootloader
there is no USB device found by Linux/Windows. After flashing bootloader
the minis USB device was found by Linux/Windows and bind by the driver and
if you start DNW under Windows it reports "USB:ok" and if you call "USB -
Status" from menu there is the correct output. So i conclude that the
bootloader is working at least partially.

I do not know at this time if the board is working with NAND boot. I got it
from another person who did not succeed in flashing bootloader and it is
unknown if there is a system in NAND. Booting NAND there is no console
output too, but a beep signal as usual. I have no LCD or VGA adapter this
time. I ordered a new Mini2440 board from China, but it takes time to
deliver. The next days i got a VGA adapter working and will see if there is
a system in NAND. If not, i am going to flash a dump from the new Mini2440
NAND to test if there is console output from NAND boot. If not, i try to
configure an u-boot with output to TTL2 to test if all UART's are broken.
But i fear that the CPU UART's are broken and maybe other IO too.

I do not know if this board ever worked. The previous owner says it worked
till he has erased NOR by mistake.

My Micro2440 has the same NOR, but a greater NAND flash. So i could not
test a dump from it.

According to my workamte:

You can compile the boot loader to manually spit out bytes to the UART.
Writing to the TDRD register. You also have to be absolutely sure that both
ends of the UART connection are configured the same (e.g. 9600 8N1).

JTAG gives you access to the CPU so you could do this manually via JTAG
too.

****

Good luck!

Thank you for your answer Dave!

Yes, a good idea to compile a bootloader (the old vivi?) for output at
another UART or injecting bytes for output over the UART's via JTAG. I have
no experiences with doing that but will try it later.

At the moment i am too much involved in another software-only project for
earning money for my life to get deeper in that matter. So i will try it
next month.