I have currently built the following three-phase setup:
3 × MultiPlus-II 6K5 units.
Lynx Power In (battery connection)
Lynx Shunt
Lynx Distributor (MultiPlus connection)
Cerbo GX
3 × 52 V batteries connected in parallel to the busbar (with equal cable lengths)
During the build, I fitted the MultiPlus units with DC cabling (with the polarity correct), provided AC out 1 with only the neutral and live conductors (the earth conductor was still on order), and fitted the AC in with all cabling. Both AC in and AC out 1 have their own independent neutral rails (not connected to each other). The earth rail is common. The AC in wiring was terminated at a four-pole main switch, and AC out 1 at a four-pole RCBO. Both were still switched off.
At this point I started configuring the three-phase setup using the MK3 dongle on the daisy-chained bus and the VE Configure app. I assigned each of the three MultiPlus units its own phase and sent the settings to the three units. I also set the grid code to Europe and added the ESS Assistant, among other things. These settings were configured and sent separately for each unit.
So far so good: the units switched on neatly without any faults, showing the “inverter on” status and making a humming sound. I then switched all the units off one by one using their switches and turned off the DC supply.
Once the earth wire arrived, I added it to AC out 1 on all the MultiPlus units and installed the cabling to and from the consumer unit. It was now time to commission the setup. I left the AC in/out main switch and RCBO switched off, and turned on the DC supply with an incandescent light bulb connected in series as a precharge resistor. I then connected the Cerbo and the shunt, and switched on the MultiPlus units one by one using their individual switches. When I switched on the third MultiPlus, a MOSFET in that unit blew, and the fuse on the distributor also failed.
Is this simply a case of bad luck and faulty hardware, or have I missed something glaringly obvious and caused this myself?
I would really appreciate your expertise on this.
The wiring diagram matches the diagram shown by Victron in their three-phase training video. The only difference is that I have not wired AC out 2.
Down below, next to the blue LAN cable, you can see the central earthing screw. These three PE conductors should be connected to the internal PE busbar! I can’t see that here.
The same applies to all MP units.
It is quite possible that an unwanted potential was created there (in inverter mode, the earthing relay at the output is closed to bridge Nout to PEout), and that this is why the MOSFET failed. I assume that a high current briefly flowed through the N conductor during start-up, which had some effect on the L3 MP (reverse current).
I had posted an incorrect wiring diagram. All the earth wires on both the AC input and AC output were connected. The earth wire hanging loose in the photo was disconnected afterwards.
Could just be bad luck. Why didn’t you use Cerbo to switch the devices off one by one after programming them and then start them centrally?
The guy with the YouTube channel (Der Elektrotechniker) also always leaves the inverter station as a precaution after switching it on, when the countdown starts after starting the inverter. MOSFETs have blown there once as well—you really don’t want to be nearby when that happens.
Exactly—the second time switching them on. The first time, I configured them as a three-phase system, switched them off and, once everything was ready, switched them back on one by one. Then it failed after the third Multi came on and they tried to switch to inverter mode.
Could something like this also happen because of an incorrect setting or start-up procedure? For example, the ground relay setting (even though it couldn’t be changed), or switching things on separately versus switching everything on at once…
