• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 83 subscribers
  • Views 1769 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

F28335 flash locked out from factory or spontaneous locking?

Fred P
Expert 2235 points

We had many F28335 getting locked all of a sudden in the past but it had stopped when we updated to CCS 5.2.1 . We now wonder if it was only a coincidence and that the root cause is still present because we are having the same issue with 3 new F28335. We never succeeded at programming them from the start.

We have other boards working fine and using the same xds100v2 device to program them. Vdd3vfl at pin stays in valid range when programming (well, on board that work...). Therefore, hardware issue is not very likely but suggestions on this side are welcome.

We use xds100v2 device to program the DSP.

Here are my observations:

The first flash programmation gave that:
"Flash Programmer: Error erasing Flash memory.
Flash operation timed out waiting for the algorithm to complete. Operation cancelled.
Trouble Writing Memory Block at 0x33fff8 on Page 0 of Length 0x8
GEL: File: D:\Brio\Code\HYD001\trunk\controller\Debug\controller.out: Load failed."

Looks like the flash locked itself somewhere between production line and my bench.
I tried to unlock it through the On-Chip Flash plugin in CCS and got that nice message:
"Device unlocked. To clear the programmed password, please erase Sector A of Flash memory."

I tried to erase sector A as suggested. I got that:

"Flash Programmer: Error erasing flash memory. Flash operation timed out waiting for the algorithm to complete. Operation cancelled."

Looks like it is still locked...

Having read about the locked/unlocked state reporting bug in CCS 4.2.4, I have tried to lock it so that I could unlock it (even though we use 5.2.1).

Flash Programmer: Error locking flash memory. Cannot lock device if password on the device is all 0xFFFF. Lock operation cancelled.

I have filled all the key boxes with something else than 0xFFFF and clicked on "program password". Got that:
"Flash Programmer: Error programming password. Flash operation timed out waiting for the algorithm to complete. Operation cancelled."

I tried again erasing the flash. Failed again.

Saving program memory 0x33FF00 to 0x33FFFF to a file reveals 0xFF all over the place. I feel that is what we want in there so it does not help me much...

I have tried Unlock_CSM script along the way. Made no difference.


What else can we do?
Any hint on what is happenning?

Are we the only ones reporting spontaneous locking problems?

All troublesome IC have CA-23AJ8NW marking. Is there any known issue with that batch?


Thanks

  • 0
    Expert 2235 points

    Hello hello hello...  Is there anybody in there?  Just nod if you can read me...

  • 0
    TI__Guru* 93800 points

    Fred,

    If I understand, it seems that the device is not locked and this is not the root of the problem.   I say this for the following reasons:

    • Password location read shows all 0xFFFF indicating the password is unlocked
    • If the device is locked, then program/erase/etc would report that the device can not be programmed/erased/etc because it is locked.  Instead the algorithm is timing out.  This means CCS thinks the algo was downloaded and started, but the code never returned.    The flash programming is done by executing code on the device by the CPU.

    So I am wondering - are you able to load execute any code on these devices in RAM?     Could there be a power supply issue (voltage, current)?

    Regards,

    -Lori

  • 0 in reply to Lori Heustess
    Expert 2235 points

    We have replaced one of the three IC in problem. It now works.  If it was related to power, I would expect the problem to persist.  In addition, that board has been running for weeks before a sudden locking occured (which lead us to switch the IC to a brand new that never worked...).

    As I mentioned at the start of this post, we had a very similar problem occuring all of a sudden in the past (using CCS 5.2.0). At the time, we could still program the RAM. However, we had not tried out all of what I have listed here (did not try unlock script, did not dump memory).

    We will try to load code in RAM on the 2 others IC in problem we still have, double check the supplies and let you know how it goes.

  • 0 in reply to Fred P
    TI__Guru* 93800 points

    Fred P said:
    We have replaced one of the three IC in problem.

    Fred,

    If the removed IC is available and in good physical shape you can work with your distributor to send it to TI for Failure Analysis (FA). 

    This would answer the question if it is really locked or otherwise damaged.    From your previous descriptions it doesn't sound locked to me.

    Regards,

    Lori

  • 0 in reply to Lori Heustess
    Expert 2235 points

    We had a similar issue last summer.  We sent IC to you for analysis.  They were confirmed as locked.

    Unfortunately, we have discarded the three IC I mention in this post.  We should not have...  

    Next time I will make sure to send them out to you.