EEPROM issue

[HighlyCompressedAir]

Using this project causes so many writes to the heat pump's EEPROM that it should break in no time. Are you aware of that?

https://github.com/Egyras/HeishaMon/issues/225

[mi-hol]

@HighlyCompressedAir how about closing this issue?

[HighlyCompressedAir]

Enabling WAR, RTC and SoftStart will cause tens of thousands of writes to a single register every year. If the EEPROM's lifetime is 100k writes which is very common, that could kill it in just a few years. I don't know why everyone just pretends like it's nothing, I don't like it. A warning on the front page is a minimum that can be done.

[geduxas]

Yes we are talking about it a lot, theoretically it's possible, how it is in reality no one knows, heishamon and this project is old enough, that should already reveal broken boards, but personally i didn't hear

[mi-hol]

Enabling WAR, RTC and SoftStart will cause tens of thousands of writes to a single register every year.

@HighlyCompressedAir can you provide evidence for this assumption?

The same usage scenario applies for large scale building HVAC automation common since at least the 1990ies. Panasonic's programmers know about the limited life of EEPROM for sure and handle writes to EEPROM as a precious resource. The original design of these controllers is over 20 years old and not known to have excessive failure rates!

Japanese companies have high standards regarding quality and Panasonic is one of the leading examples (see here and check EcoVadis assessments i.e. from 2016).

heishamon and this project is old enough, that should already reveal broken boards, but personally i didn't hear

exactly!

I'm sure US citizens would have created a class action lawsuite against Panasonic, in case of failures related to air cooling/heating products but there is none.

[geduxas]

@mi-hol it's arguably.. no one exactly know, risk always possible.. Panasonic (without heishamon and taw) have additional integrations to KNX and Modbus networks.. so yes it's prepared for bigger integration into BMS system's.. also i didn't see anywhere mentioning limit of writings..

Also it's possible that all setting are stored in RAM, and only on power failure with help of big capacitors it writes state to flash.. i am just guessing :)

[mi-hol]

risk always possible.

that is correct, but title says "issue". An issue is defined as: "a risk that has manifested itself"

@edterbak Without evidence of a real issue, this topic should be moved to "discussions"

[HighlyCompressedAir]

I wasn't aware that Panasonic supports Modbus. The fact that they do makes 100k write EEPROM highly unlikely.