MSP430G2402: Strange problem with VECTORS LOCATION in memory (again)

I will loop Dietmar into this thread so that we can re-start this conversation. Any further details you can provide from the past four months would be appreciated.

Regards,
Ryan

Good Morning Ryan!

First of all thank you for your fast reply.

In order to join some dots...

--> Both the boards I referred in the previous old posts of September and this new 2 boards had MSP430G2402 uP on them. What I notice is the fact that in the RESET VECTOR there isn't the address that I expect (the main address) but another one.

--> In the second of this 2 new one boards that have this strange problem, I added on the board a TRANSIL SMAJ 5.0 against electrostatic discharges to avoid any voltage spikes but unfortunately the problem is still there.

--> In every processor for this application I use the information memory as EEPROM and I write/store regularly data on it. Is it possible that when I write on this part of the flash memory I enable unintentionally writing in all the flash memory?

Can this interfere with the waste of the correct data in the vectors? How can I do to protect the VECTOR's region from random accidental writing?

It is very important for me and my company try to resolve this problem definitively.

Any ideas and advice would be really appreciated.

Best Regards,

Luigi

Good morning Luigi,

Are you able to share schematics and a code excerpt of your flash memory write sequence? How easy is it to replicate this issue? Can failing units be re-programmed to work properly? Can any device be susceptible to this behavior?

Regards,
Ryan

Good morning Ryan,

thak you for your fast reply.

I attach at the end of this post the schematic of the HW and also a code extract of the FLASH_WRITE procedure I wrote.

Unfortunately I'm not able to replicate the issue because it is a random problem. Moreover, the strange fact is that when I re-program the board everything works well, but I don't have any certainty that the issue would recur again (until it really happen or not).

I hope in good news from you,

Kind Regards,

Luigi

2021.Schematic.pdf

 

FLASH_WRITE_ok.h

Hello Luigi,

Please let me help to support this case.
From the history I see there was a concern from Dietmar about the supply voltage.
I don't see an exact answer to this, but I have noticed your application is operating at 3.25V and 16MHz.

Is this correct?

Hi Lukas,

I'm glad to read that you want to help me.

Yes, what you said is corect. The machine operates under 3.3 V  at 16Mhz.

Thank you very much for your answer.

Kind Regards

Luigi

Goodmorning everybody!

I'm here after an entire week to inform you that I did several tests on the board giving an electrostatic discharge in order to reproduce the issue... But no score at all!

Have you got some ideas about which is the source of my problem? Were the schematic and the code for the flash writing procedure helpful?

I really appreciate any comment and help you would give to me.

Best Regards,

Luigi

Hi Luigi,

Thanks for your update.

I"m still concerned about operating the device at the Edge of 3.3V with 16Mhz.

Do you have to change the voltage/frequency and check this again?

Best regards

Lukas

Good Morning Lukas,

thank you for your reply.

To answer to your question, not yet... I didn't try to change the voltage/frequency of the device yet.

The very strange fact, in my opinion, is that I 've been using these devices from 10 years, with this settings for voltage and frequency, and this is the first I have this particular problem on some machine I produce.

I have a question: if the problem is that the device operates at the edge of the TI specification, wouldn't it be much reasonable to expect the complete damage of the FLASH memory (specifically the cells' memory reserved to the vectors) instead of the loss of the correct data in the vector's locations?

For complete damage I've meant that the FLASH memory isn't no longer re-programmable.

I hope someone could let me understand better this problem.

Thank you for the attention,

Best regards,

Luigi 

Hi Luigi,

Can you isolate this issue to certain Lot Trace codes?
How many fails do you see?

Best regards
LUkas

Hi Lukas!

I see about 15 fails at the moment and every board has uP with this word: MSP430G2402, 34KG4, CF30A.

I really need to find what is the exact problem in order to resolve it as soon as possible, so, if you have any ideas please let's discuss it!

I hope this would help to find a solution to this issue.

Thank you very much.

Kind Regards,

Luigi