board (possible) hw problem

Go To Last Post
2 posts / 0 new
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi all

We were working on a development board attached to the pci-like slot of the stk1000 that wasn't going really well as it had a static ram chip attached on the EBI interface than seemed to mess up with the data and/or address bus lines thus corrupting the bus contents and made u-boot stop the booting process and reset the avr32. When this card was off the slot the system booted ok.

We tried to probe this pcb to see if it was short-circuiting the data/address lines and when we took the card off, the board persisted the erroneus behaviour as if something was burned.

More specifically u-boot gave use the following message:

U-Boot code: 00000000 -> 0000e150 data: 24000000 -> 24002388
SDRAM: 8 MB at address 0x10000000
Testing SDRAM...FAILED at address 0x10000000
SDRAM: read 0x7b42fe3f, expected 0x0
malloc: Using memory from 0x0ffc0000 to 0x10000000
Flash: 3842 MB at address 0xf01ff01f
DRAM Configuration:
Bank #0: 70987098 3842 MB

*** Unhandled exception 2 at PC=0x00003e4a
Bus error at address 0xffffffff
pc: 00003e4a lr: f1dcf1dc sp: 24007f34
r12: f1dcf1dc r11: f01ef008 r10: 80000000
r9: 24000420 r8: c160c160 r7: 0000007e
r6: 24000410 r5: 24007fac r4: 0000007c
r3: 24000038 r2: 24000030 r1: 00010008
r0: 00000014
Flags: qvNzc
Mode bits: hrje....G
CPU Mode: Supervisor

Stack: (0x24007f34 to 0x24008000)
7f20: 00004626 2058818a 0ffc0008
7f40: 24007fac 2800c10c 24000c5c 10000000 00000000 00000014 000002b0 2058818a
7f60: 0ffc0008 24007fac 2800c10c 31313532 303000a2 90b043d3 3fefa63f 6ef9c4a6
7f80: 66f337aa 01de0db6 2b41ed0a c7f8c89b 2576c6ef df0b2c58 f4fe8ea9 6b5355d7
7fa0: b4e315fc ab72a89b 28c6f74f 0ffc0008 00000000 000089dc 00008994 0001c200
7fc0: 00000000 00000001 00000000 007f0004 00000001 07270de2 00000000 f8d00083
7fe0: 2058818a 24000938 4884f886 2800c10c 424f01aa 0000003f 00000000 00000014
Unhandled exception

We said ok we burned it, but... after some cold resets and some sweet talking to the board it became fully functional again, except the lcd that flickers a little. Twilight zone.

It seems that the data or address lines are getting corrupted easily, who knows, static electricity? The noise capacitance maybe? I don't know. Our stk1000 board on the other hand was one of the first out on the market (more a prototype) and maybe it lacks some quality. Moreover we've been beating it every day for almost a year now so it maybe starting to kick the bucket.

Can anyone with a greater expertise give as any hint of what may went wrong?


  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have see this type of error when I pull an address- or data-line to ground or 3.3 volt.

Do you have the latest schematics for STK1000, it is on I would double and triple check your board connector.