-1
\$\begingroup\$

I am working with an external Serial Flash memory (Winbond 25Q80). I could read-erase-write the flash memory using SerialFlash library in arduino environment. But it seems data of flash memory gets corrupted after some hours. I am suspecting either of these below two points:

  1. I supplied 5 volt instead of 3.3 volt (as prescribed in the documentation) as VCC.
  2. I was continuously reading in span of 15 seconds in the same file.

Can someone please confirm any of these above points can corrupt flash memory or not?

\$\endgroup\$
1
  • \$\begingroup\$ Giving integrated circuits too high voltage (supply or I/O levels) very often results in them breaking. I'd assume it is now broken and replace it. \$\endgroup\$
    – Lundin
    Commented Mar 20 at 7:10

1 Answer 1

0
\$\begingroup\$

Yes, either of those two could be the reason.

5V will exceed the absolute maximum limits on a 3.6V rated device, so it may be damaged.

Reading could corrupt the flash, it is just unlikely, unless the flash chip had a manufacturing error. Rare but it could be defective. It is rated for 100000 erase/program cycles and 20 years of data retention, without mention of how much it is rated for reading.

Any other thing could also corrupt the data while you read or write to the chip. For example, if you are using the 3.3V chip with a 5V MCU, the 3.3V may not be enough for the 5V MCU to read data properly from IO pin. Also you can't connect a 5V MCU directly to 3.3V memory chip so you must have a level shifter on the bus. And level shifters, when used or designed improperly, can corrupt the data too.

\$\endgroup\$
5
  • \$\begingroup\$ Thanks a lot for your response. How probable the data corruption to happen, if the flash memory is supplied 5 V instead of 3.3 V for a prolonged period of time? \$\endgroup\$ Commented Mar 20 at 6:37
  • \$\begingroup\$ @ShekharSaha I would not call this a data corruption problem - the 5 V almost certainly damaged the device, which now is unreliable at best. Any period of time is likely to damage it, granted longer periods make more damage. \$\endgroup\$ Commented Mar 20 at 7:06
  • \$\begingroup\$ After erasing the flash, it starts to work again though! This is why I am kind of confused \$\endgroup\$ Commented Mar 20 at 7:23
  • \$\begingroup\$ @ShekharSaha The chip is rated to get degraded and damaged at 5V so it might be degraded and damaged, even if it did not blow up or burn up immediately. You cannot trust the chip any more. Unless the data corruption is a problem of you using the flash so that it corrupts the data. \$\endgroup\$
    – Justme
    Commented Mar 20 at 7:29
  • \$\begingroup\$ @ShekharSaha we’re dealing with complex devices so failure could occur in many ways due to various mechanisms. If the datasheet tells us not to exceed a given value and we exceed it - expect problems. You might also have unrelated problems that all combine to confuse the issue. Even if you did have the probability of overvoltage causing the problem, how would this help you? It’s simply bad practice to exceed the manufacturer’s specs - fix the obvious problems first. \$\endgroup\$
    – Kartman
    Commented Mar 20 at 12:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.