CC2530: Internal Flash High Temperature Operation - Limitations

"I have seen in the datasheet that for RAM in temperatures over 70 degrees and in certain power modes the RAM contents could be altered.": Do you have a reference for this?

The datasheet would have stated if a temperature limitation existed. 

What we typically have seen when customers have issues with flash corruption is that it is caused by bugs in the application code.Typically a flash write to the program memory causing the program pointer to go strange places etc. 

Is this reproducible? 

Hello TER,

Thank you for your response. The reference can be found on the cc2530 document as you can see below in the red box (document : swru191f.pdf)

As for you other question: I have not been able to reproduce the issue on my lab yet.

BUT, we have a customer where we have seen this issue a number of times.

We have seen that placing a device in a specific spot where there is high temperature, the flash gets corrupted after resetting. All replacements that have been given to the customer and were put to that specific spot end up having this same issue.

As for your note on checking if there is any code that could end up in a faulty write in the flash, I am also checking if this could be the case.

By the way, is there any danger writing in the flash while the device is operating in high temperature?

Thank you in advance,

Kind Regards,

Nikos

I only checked the datasheet, hence I didn't see the note. 

With regards to the note: Do you any operations during PM2/ PM3 that could be the cause of what you are seeing?

I assume that when you test this in the lab you use the same temp as in the customer installation. What else is specific for the failing spot other than temp? How is the power supply etc? 

What is the failure rate on that given spot? 

Hello TER,

Thank you again for your response.

The failure rate on this spot is 3/3. I mean 3 devices were given in total. The 1st and 2 replacements. So I would say 100% :)

We don't use any PM2 or PM3, i just quoted the text due to the temperature note (over 70 degrees). In other words the 1st question was: if the RAM can affected in specific temperature ranges and voltages can also be the flash?

Power supply is checked and there is no issue. The same power supply is given to 3 other devices some centimeters further that belong in the same Zigbee Nwk as the faulty one.

We are also trying to see if the customer sends commands that could trigger many flash writes (like changes in the binding table etc), but the weird thing is why the other devices are not affected like the one in the specific spot? (they are in the same Zigbee Nwk)

What about the other questions?

• Have you seen the flash getting corrupted in high temperature?
• Have you seen the flash getting corrupted over 125 degrees that is the limit of operation?
• What could happen if a write to the flash is done during  device operation in high temperature?

Thank you again,

Kind Regards,

Nikos

We have not seen any flash issues as long as the chip operate within the parameters given in the datasheet. 

- "Have you seen the flash getting corrupted over 125 degrees that is the limit of operation?": This is outside the spec of  the chip and has not been tested. 

"The same power supply is given to 3 other devices some centimeters further that belong in the same Zigbee Nwk as the faulty one.": I would assume that more than th faulty device operate at high temp? Since a node at a given placement fails but not the node that is places fairly close: What is different with the failing node (from a software/ usage point of view)

Hello TER

Thank you again for your response. The failing device runs the same firmware as the others. I will update you with more info on this. Thank you again for your support.

Kind regards

Nikos