This is a bit technical and only meant for those who have interest in accessing the Victron network.

Status:
I have finished the basic test to read from all connected devices.
I could read from can bus using uId 0, found the first solar controller.
I found the register with the Multi status, off/on/charger only/  inverter only field 33 , Solar charger is field 774 to switch off

I have rewritten my test code into a Modbus server to easily integrate in my central application. It is an expandable OO Java app and I created a public Github project to share as promised above.

Here is the link:
https://github.com/teutanic/Victron-Modbus-Server

I have not yet written a doc on that but I document the code pretty well for myself , since my memory is not the best anymore and I am doing too many things at once.

One of the reasons for the confusion was that the Digital Multi Control can be set to charger only, but not to inverter only unless one reduces the limiter to 0, which is not really what you want. You want the AC in, but no charger. I can do this I guess with field 33.
Begs the question why the Digital Multi Control cannot do this ?

I found a couple of issues: but not yet thoroughly verified.

Setting the current limiter on the Digital Multi Control to 0 is not reflected in the field 22 Active current limit, however all value above the min of 6.5 are. Why not ?

Another issue is the data for the inverter output, voltage, current and power do not add up and not only by a bit. Something is wrong there, but as I said I have to thoroughly verify this. I have had the inconsistencies with my  app as well as with a standard app. Not that urgent but will do this eventually.

I also have a weird problem with the DC supply to one of the Multis. Sometimes the multi reports low battery and I  measure actually the low voltage. It must be a bad contact on the supply cable, however they are double 70sqmm with massive bolts on your end as well as on the other end. I will take that connection apart again, clean it and put some conducting paste on it, just in case. I am reasonably sure it is a contact problem, one I never had with such massive cables. I have load tested the whole setup to it’s max capacity (except the solar) and everything looks ok.
The problem with the wrong DC out info is corrected after firmware update of the Multis. However I would still like to know how exactly this figure comes about.

I also finished a small documentation for the test software. It is also on Github with the sources.

https://github.com/teutanic/Victron-Modbus-Server/blob/master/Modbus%20TCP%20Server%20For%20Victron%20Network%20With%20CCGX%20Gateway.docx