Improvement: Ve.Direct, FrameHandler, Handle timestamp of multiple frames

[SW-Niko]

This PR can be one possible solution of issue #1499

Testing is ongoing on system with 1 smart shunt (2 frames) and 2 solar charger (1 frame)

Note: isDataValid() use now the batteryVoltage_V_mV instead of the serialNr_SER because batteryVoltage_V_mV is available on all VE.Direct devices and serialNr_SER only on solar charger.

[schlimmchen]

I don't understand the _actFrameNumber and _endFrameNumber voodoo.

We support the Smart Shunt and the SmartSolar MPPTs. The former sends two frames, the latter only one. So... Shouldn't we set the last update timestamp to millis()

1. when a frame without the V label was processed for Smart Shunts (and if at least one with V was processed?) and
2. when any frame was processed for SmartSolar MPPTs?

[SW-Niko]

Hello @schlimmchen.
since some days a have a simpler solution running.
This version take care regarding your concerns "Data can move between frames" and also removes the "_actFrameNumber and _endFrameNumber voodoo".

We support the Smart Shunt and the SmartSolar MPPTs. The former sends two frames, the latter only one.

Not always right. At least we know one MPPTs (250|85) also use 2 frames. See #689

Just give me some time. I want to finish my tests before I commit the updates.

[schlimmchen]

Not always right. At least we know one MPPTs (250|85) also use 2 frames.

Oh, okay. That changes things...

Just give me some time.

Of course! Why would I pressure you while I take my sweet, sweet time to prioritize your other PRs 🙊 There is so much stuff to do...

[schlimmchen]

Feel free to un-draft when you think the PR is ready.