Python file gets corrupted during AC fluctuations

[shivams]

I have installed a few ESP32s for a test project. Some of them are powered using an AC-DC adapter (with 5V DC output). Rest are powered using direct DC from batteries. The ones powered by the AC-DC adapter have had a high failure rate. They often get corrupted: some of the Python files are not readable, while the firmware itself is intact. Re-uploading the Python files make them work again. I am hypothesizing that this problem is caused by the AC fluctuations. However, since the firmware survives and only the Python files are corrupted, then MicroPython also may have shortcomings. I'm not sure. I live in a region in India where AC fluctuations are common. Has anyone witnessed the same?

And yes, I am planning to upgrade my power supplied to something which gives a more refined DC supply. However, I wish to know if something can be done at the MicroPython end to make it more resilient.

[pythoncoder]

The only solutions to electrical problems are electrical. Many USB adaptors are fit only for charging phones rather than for supplying continuous power. But even the best adaptor will fail if the AC voltage drops too far. In that case you need to look at some form of battery backup. Some ESP32 boards have support for LiPo battery power, otherwise you'll have to engineer your own solution.

[shaoziyang]

If the system is reset when writing a file, the file system will be damaged.