Flash write cycles concern | time period power | cd=07 and cd=20

[n-alex-1]

Feature Description

Thank you for the projects! I became a big fan step by step.
I have two B2500 in use - they work together in the automatic mode with the CT002 quite good.

Recently I bought the Venus E 2nd Gen
I realized it doesn't controll well by its own - so I decided to control the device via HA automation by adjusting: venus_e_time_period_0_power

I do it if the required power difference is more than +-/20W (other adjustments the other B2500 handle well as well, there are no conflicts. Input value from CT002 I get via bluetooth from the B2500 units - this works smooth as well).

My concern (I'm not sure if its a valid one - please confirm):
I thought its not a best idea to create so many write cycles to this time-period-power setting because it stores it permanently. I'm a bit scared to break the flash after a while. But many people are using this parameter as well as I understood.

I found this page additionally:
https://eu.hamedata.com/ems/mqtt/index.html?version=153&type=HMG-50&mac=replaceMe
which describes the protocol and they have two options:

cd=07 --> Timed and fixed power discharge settings (Flash storage)
and
cd=20 --> Timed and fixed power discharge settings (No flash storage)

Are my concerns justified?
Current Firmware seems to support even more timer settings 7 or 9 instead of 4 so I'm not sure if the link I posted above is still a valid one. I just accidently found it by typing: hame hmg-50 in the famous search engine.
Where do you get your source of information?

Device Model (optional)

Venus E 2.0 Gen, CT002, B2500

Firmware Version (optional)

V153

[tomquist]

Yes, your concerns are justified. The Venus doesn't have an option to change the output power via MQTT without writing to the flash. The link you posted above is actually the MQTT documentation for the B2500, not for the Venus. So what you're asking for isn't technically possible without Marstek making a change to their API. It's unlikely they'll add support for that in MQTT anytime in the future. More likely, they'll add it to the new Open API (see here).

[n-alex-1]

Ok then the question the way around:
Do other people use the same parameter to create the beautiful manual "Nullpunkteinspeisung"?
Or it is not recommended at all? (Did somebody killed his device already? I guess I did more than 50.000 write cycles already in one week)
Any idea why the automatic feature is working so much more delayed compared to the logic in B2500 for example?