S29GL256P90TFIR20 data corruption

[okernytskyy]

We are investigating fields failures and we have 2 cases when NOR flash memory data became corrupted after few month operation at customers location. Why NOR flash memory data become corrupted?

[Apurva_S]

Hi @okernytskyy 

Could you please give us more details about your application and the corruption?

• Is the device being continuously programmed/erased at the customer location?
• Are there chances that the device may have exhausted all its P/E cycles?
• Are there any chances that power failure/fluctuation may have occurred while programming/erasing the device?
• Please elaborate on the nature of the corruption, number of bytes/location etc.

Regards.

[okernytskyy]

The device was not intended to be programmed/erased at the customer location

It is very low chances  that device may have exhausted all its P/E. It was programmed with boot loaders and two kernel images by contract manufacturer, and we do not expected any other intension to re-program it (except FW update, which is not the case).

We do not have information that power failure/fluctuation occurred while programming/erasing this device. We made thousands products and saw only two failures of that type happened at customer location.

We dumped data from the suspicious NOR device and compared it with the source data. There are 5 bytes difference between rootfs1 to roorfs2.

//RED=rootfs1 (crashed)

//GREEN=rootfs2 (still good??)

//Each set has one byte difference.

 

[Apurva_S]

Hi @okernytskyy 

Apologies for the delay in responding.

Do both the devices show same type of failure (similar corrupted values and on the same locations)?