Mixed Generations

[MadMoll]

Good morning everybody, hope somebody has the same situation and can help me out.
I have two Marstek Venus E, but Gen 2 and one Gen 3, Gen 3 seems much faster and that may cause the issue.

Anybody has a working config with that?

It goes back and forth even if I made ready ridiculous settings like poll 2 - 5 seconds or throttle interval very slow smooth alpha 0.1, power multiplier big deadband and and and.

The only thing that makes it working is my FHEM, just sending it only every 2.5 second and doing something like this, not good but much better. (Deadband should do the same but my seems to working better, I do not know why)

Anyone has a running config or more ideas? It is still not good, because it is slow, and I guess I really tried any AstraMeter setting I found. :)

Best,

if (abs($p) <= 2000) {
    my $faktor;
    if    (abs($p) > 1000) { $faktor = 1; }
   elsif (abs($p) > 20)  { $faktor = 0.7; }
    elsif (abs($p) > 10)  { $faktor = 0; }
    elsif (abs($p) > 5)   { $faktor = 0; }
    else                  { $faktor = 0; }
    
    $p = $p * $faktor;
}

[MadMoll]

Morning, if somebody has a similar problem, looks I got, it.
Reason was I write it into a file, even if it does not change and that makes it bad.
Now I check before writing and it looks much better.
Guess this is doing also the trick (DEDUPE_TIME_WINDOW = X), but I only tested the suggested 0.9 which was much too low in my case, because Marstek really needs it time to start.

[tomquist]

The upcoming version will include many improvements that should automatically prevent most oscillations in multi-battery scenarios. Would be interested to see if it improves things in your setup. Find instructions on how to install the version here.