question

Hi @gyrovague

I have had similar issues with some of my sensitive equipment. I have a 8kVA Quattro and although most of the equipment doesn't notice the load-shedding some of my more expensive test equipment does... I solved the problem by placing an inline UPS on those loads. I already had the UPS's so it was at no expense to me... I assume that because the equipment is rather far away (in wire length) from the inverter, there is a larger delay in voltage change over, large enough for the system to notice it and shut down...

I have been trying to read on using AC caps for low voltage ride through... but haven't had a lot of time to read on that. I cant see why it will work but I have had a few colleagues mention it, so I will have to get my oscilloscope out and test it one day.

Yes, I'm not sure either what the purpose of having the ET-112 is... The way it was described to me was that the ET112 is necessary in order to have non-essential loads on the input side of the multiplus II, I think it essentially then feeds-in from the MP2 to those non-essential loads, and uses the ET-112 to measure the true value at the grid side. So something like:

          __________________GX
          |         ________|___________
grid -- ET-112 --+--|AC-in     AC-out-1|-- essential
                 |  ---------+----------
            non-essential    |
                             +--MPPT--PV
                             |
                            bat

(I may have some of the details wrong, doing this from memory, and I'm by no means an installer).

I think the issue is that between the ET-112 and the GX and the MP2, somehow it reacts very slowly (15 seconds or more) to large non-essential load changes (maybe due to the RS-485 to USB converter?), so ends up feeding in too much and thus leaking into the grid, which resets the prepaid meter.

What I don't understand, is how the above is supposed to be any better than doing simply this:

                            GX
                    ________|___________
grid ---------------|AC-in     AC-out-1|-- essential
                    |          AC-out-2|-- non-essential
                    ----------+---------
                              |
                              +--MPPT--PV
                              |
                             bat

As far as I can tell, the AC-out-2 is connected to the AC-out-1 (on my model of MP2) via a relay that is controlled by default based on the grid presence on AC-in (measured by an internal CT inside the MP2 I think). This relay can also optionally be controlled by assistants to allow one to have non-essentials enabled during grid failure while (say) battery is above 80% or whatever, thus also allowing non-essentials via PV+MPPT during grid failure.

The installer maintains the former is "the way the system is supposed to work" and that my 2nd drawing is "not the correct way", but from the things I've been reading in the manual and these forums, I honestly cannot see a reason why he says that, and in fact I have a feeling this configuration stems from before the MP2 with AC-out-2 came out. It seems to me this would eliminate the whole feed-in slow response problem altogether, while still allowing me to use PV+battery energy on non-essentials while grid is working, and prevent non-essentials using battery when grid fails... So essentially all the features I want/need, without any of the problems, and would've been slightly cheaper.

I'm probably missing something. Tagging @Daniël Boekel (Victron Energy Staff)

Many grid-tied installs (especially when backup power is a concern) have more appliances attached to them than the inverter can power on its own, with passthrough from grid covering the slack. This is a much cheaper way by avoiding massive rewiring.

Solar/batteries may only power part of the load or an ESS inverter limit is set, so it is no guarantee of a seamless failover.

The issue seems to be more about brownout as suggested before.

If you can drop the supply breaker and nothing has an issue, the problem is on the grid side.