Problems getting MCF5373L to run...

[w_wegner]

Hi all,

I am having serious problems getting "my" MCF5373L to run.

My first (software) steps were with a Cobra5329 board as a development platform. There I used TBLCF BDM adaptor just for checking I could contact the processor because I did not want to touch the bootloader, but this worked.

Now we have a custom board with MCF5373L, DDR-SDRAM and a 16-Bit Flash, FlexBus is running on 1.8V. We have 6 Boards made, and only two of them even had a running crystal oscillator. One of them stopped operation shortly after, and even on the last one with working oscillator, I am not able to contact the processor using my TBLCF BDM adaptor.
With the "bdm-chk" utility from the m68k-bdm tools, I only get to the message "CSR break not set, target failed to break, CSR = 0x00000000" - this is more than without powering up the Coldfire, because in this case, the utility fails even earlier with "Reading CSR failed: Operation not permitted". This behaviour is the same with all boards, oscillator running or not.

All operating voltages are present and clean, reset is released. What puzzles me most is that on all the boards that are not running, the crystal oscillator pins are floating at around 3V, while being down below 1V when running. Changing the processor on the board that stopped working does not change anything.

The crystal is close to the balls, 18pF series resonant, I also tried changing capacitor values - it seems not to be a problem in this area. I double-checked all (supply, system control, data and BDM) pins according to the MCF5373RM Rev. 1. RCON is low, and D[98654321] == 00000100 is driven by a 74ALVC245 during reset (static behaviour also checked).

I am really out of ideas, so if anybody has hints what I could check and might be doing wrong, please let me know!

Best regards,
Wolfgang

[w_wegner]

Hi again,

I have some more test results:
The first problem is that something is wrong with my reset configuration; during reset and when de-asserting RCON during reset, the oscillator runs fine. I did not yet figure out what might be the cause here. As already mentioned, I drive the configuration to D[98654321] using a 74ALVC245 whose OE is connected to RESET, and the RESET rising time (around 15 ns) seems to be OK to me.

The second problem is that even with default RCON setting, I can not contact the processor using the BDM cable. Now I get the message "Read register failed: Operation not permitted" upon first register access, which actually is the same as when no coldfire is connected at all...

I will post again if this gets solved.

Regards,
Wolfgang

[w_wegner]

I solved the problems now:

1. in the hurry of the last schematic changes, I tied the configuration latch OE to RESET instead of RSTOUT - after fixing this, RCON configuration seems to work, I only have to figure out why I still have 180/60 MHz operation instead of 240/80...

2. the bdm-chk utility could not contact the processor because the processor starts "executing" the contents of the empty flash and gets stuck. When inserting a reset from the BDM port at the beginning, I can access memory and (it seems) all registers. This was not a problem with the evaluation board because there the flash is already pre-loaded with dBug.

Regards,
Wolfgang