I've read all parts of this thread. However, you've repeatedly asked someone else to debug your code for you. Naturally, different people will interpret your request differently.

You must admit, though, that it looks as though you don't want to do it yourself. At least twice, you were asked, "What does the datasheet tell YOU?" (My emphasis on YOU) The point being, what did you glean from the datasheet in this context?

You haven't offered to share your experience. All you've done is to provide your code and a link to the datasheet. You haven't said what the outcome of your own efforts has been. If you've adhered to the specifications of the device you are using, your code should work. If you haven't ... well ... you should figure out why not.

Have you got an oscilloscope and/or logic or protocol analyzer? Have you used them to examine the interaction between the SEEPROM and the MCU? How do you generate the "power loss" in question?

"Between the lines" of your writing is what appears to be a desire to write one code set that will satisfy the requirements of more than just one type of SEEPROM. If that's the case, you should say that. if not, why not attempt, with your code, to meet the requirements of the device you've chosen, and report your findings.

If, having done that, you want to modify your code such that it works with more than one type of SEEPROM, perhaps that can be done, but first you must get one type to function. Why not try that and let everyone know what your results are. Having once done that, perhaps someone with the necessary experience can tell you what may be going on. Please be sure to include the MCU type and oscillator frequency, along with any relevant configuration options.

RE