I dont know if the Venus E works right now, its new for me.
It is a new one and comes in August. But is want to know, do I really need something like this: Elfin-EW11?
or can I only use a UTP cable from the Marstek to homey? I’m not really technical, that is why I ask it 
I think I have already the answer “sorry, it is in Dutch”:
Ja, er is een véél simpelere manier waarmee je die Elfin-EW11 adapter helemaal niet nodig hebt!
De Marstek Venus E Max is een modernere generatie batterij (V3) en deze heeft standaard een fysieke LAN-poort (Ethernet RJ45) ingebouwd. Bovendien heeft de Marstek-app een officiële Local API-functie gekregen.
That is good news. But I hope that it is working on your homey app too.
Hi Kaoh,
I’m greatfull for your tremendous effort to get all this up and running based on try and error and puzzeling based on reverse engineering. THks for the 1.4 version I have installed fresh today.I have now all 3 units connected and so far they keep communiting,( fingers crossed) When setting up the connection the voltage level pops up immediately the other values are later. I’m doubting if it is not better to use the the RS485 modbus connection. I keep you posted if this is stable situation. I’m trying already multiple days and this is the first succesfull attempt. Thks!!! I have last sw version V148
Someone has experience / built concept to have peak shaving mode mode instead of self consumption and comensating to 0W. In Belgium we have to pay transport cost on the larges quarter hour peak of the month. In winter time it would be great to ask the battery to sheave the peaks above e.g. 4 kW. Most of us are self consumption and the units are compensating ro 0W. It should be great to have Homey able to set 4000W instead of 0W
You could build that yourself with advanced flows.
Nice idea, I can incorporate that easily in my Smart EMS app. I like to keep the hardware controller app clean.
I think I now know where I went wrong in the previous versions. I had a communication wrapper to prevent overloading the commands (reading / writing at the same time). The ew11 understood that, but the bit more simple utp thing did not and lost track creating a queue of pending commands.
I won’t do that again 
Depends on if the registers are the same.
See the new test version, it has a dedicated Venus D driver now.
CHeck if that works better for you.
For the rest, small steps in improved loggin in the new test version.
People with problems before please do test this.
Awesome, thanks! I updatet to 1.5.0, deleted the old device in Homey and added it again.
After restarting the app, the first capability showed 129, which seconds later became 15.36 kWh.
129 kWh is the value of “energy” in the Marstek Local API app, not the available energy.
15.36 kWh is the full capacity of the battery and currently not representative of the actual capacity. The battery isn’t fully charged, although version 1.5.0 thinks it is (SoC 100%).
Those are the only two pieces of information retrieved via modbus.
fe206cae-fd9d-43bd-af47-c497115c9dc5
dont readd your devices it should upgradew in place, thanks for testing
Venus D: Update
Moving around with some registers to read, so what this one does.
Still only those 2 capabilities, but now they are correct. At the big drop the new update was installed and the minutes after that it seemed like a bunch of data came in, but not realtime. The decline is too steep at that rate of discharge (0.7 kWh in 4 minutes, but the discharge power was 1500W). And then no more updates except at 6:40, while the discharge power was steady.
New screenshot of the past hour, with the yellow graph showing the actual measured output power and the red graph showing the SoC from the local api.
FYI: the mapped registers of the old working version some weeks ago did work fine, however I cannot confirm or check that some registers were changes with FW 149.
f4610f37-b0a5-4309-a140-e01b96828317
Another Venus D update on test channel. lets see how this one works.
It looks like that even if most registers are the same, SOME hang the modbus UTP host when I try to read them. So we are moving them in test by test to see which ones I should skip for the D.
Thanks! Unfortunately no new capabilities, but the old 2 show the wrong information. Available energy is not correct anymore and the SoC is also wrong.
I need a diagnostic report, can you send it?
Sorry, here it is: 399ef6a1-f695-4fc0-b8c1-2a329129aae4
Another diagnostic report for you.
Running 1.5.2
E 3.0
35035a80-6170-4e81-971d-7df9960e58ad
I am still running on test V1,4,1! Venus E (V1.48)
is it better or wise to go back to the final?
