My test scenario/problem:
in cerbo-GX:
—Settings/System setup/ESS I switch the “grid setpoint” to -14000 which brings the inverters within a few seconds to ~ 3* 3450W output and stabilizes there with +/-50W.
—in my test i also switched on consumers with a total consumption of 12500W in the house - inverters also stopped at ~ 3* 3450W and the gap (~2000W) was filled with electricity from the grid.
Some details about my system:
-2* 48V 16kwh battery packs supporting 400A and configured for 300A max. discharge. (config in JK-BMS, serialbattery-driver, aggregateBatteries; aggregateBatteries provides 300A discharge current to the cerboGX in the parameters)
-1* Lynx shunt between both batteries and rest of the system
-3* muliplus-II 5000 in a 3-phase configuration
-DC wiring is 50mm² between busbar and MP and 95mm² between batteries and busbar, busbar is 1000A Victron distributors and power-ins.
-MPPTs are not really relevant, the problem persists also when PV provides (partially) the power to feed the inverters directly.
-AC cabling is 16mm² to each multiplus (=>4x16mm²) connected directly to connection house <=> grid.
-Fuses for the AC between MP-II and grid are 32A circuit breakers (per phase) and 50A (melting) fuses (per phase) for the whole house.
-MP-IIs are connected to AC-IN, AC-OUTs are currently not used.
-MP-IIs are controlled via Cerbo via the EM540, which is placed first, when cables come into the house from the grid (after the 50A fuses).
-Software/firmware: I am using latest/stable for all devices, so e.g. v3.60 on Cerbo.
-System is running for 3 months, Victron equipment bought new in 08/2024.
What I did:
-checked multiplus settings via VE Configure
-checked JK-BMS, serialbattery, aggregateBatteries for possible wrong limits
-checked CERBO-GX settings (v2):
— System Setup/ESS/Peak-shaving: “Multiphase regulation” => Total of all phases
— System Setup/ESS/Grid feed-in: “DC-coupled PV - feed in access” => on
— System Setup/ESS/Grid feed-in: “Limit system feed-in” => off
— System Setup/ESS/Peak-shaving: “Peak-shaving” => Always
— System Setup/ESS/Peak-shaving: “Limit system AC export current” => off
— System Setup/ESS/DVCC: “DVCC” => on
— System Setup/ESS/DVCC: “Limit charge current” => off
— System Setup/ESS/DVCC: “SCS- Shared current sense” => on
— System Setup/ESS/DVCC: “SCS- status” => Disabled (external control)
— System Setup/ESS/DVCC: “Controlling BMS” => AggregateBatteries
-checked VRM portal settings:
—Dynamic ESS is enabled, but not used due to being switched off in the main dashboard
—Still all settings should be fine.
------System: “Maximum export power” is set to 15kw
------battery: “Maximum discharge power” is set to 15kw
-checked temperature above the MP-II: It goes up to max 40° (According to Victron docu the expected limit is at 45° ~3700W, so no explanation for the limit of ~3450W)
My strong assumption is that the MP-II should be capable of producing at least 3x3700W, so >11000W even at the highest measured temperature.
Also a technical issue inside the MP-II I would consider very unlikely, because the output rises more or less linear with constant speed to the 3450W.
When I loaded 5k Multis in warmer ambient temps, it ran to its 4000W for a few minutes then derated to 3500W.
I could reproduce this at temps below 25C as well.
Exporting is less efficient, that needs to be considered as well, there are a few variables.
I had a similar issue.
Issue was different voltage measurements between MP2 internal and BMS, due to voltage drop on battery cables.
I ensured diagonal battery wiring - plus on bat 1, minus on bat 2.
I added a voltage sense line from MP2 L1 to bat 1 plus and bat 2 minus.
Now I can utilize full potential.
Hi Björn, that is actually a very helpful tip!
Do You know, which value the MP takes for comparing battery vs. MP-measured voltage? (Cerbo DVCC battery monitor or ESS battery monitor)
Did You try to turn off the MP battery monitor in the VE.config settings?
When You say voltage measurement, do You refer to these optional voltage sense cables, that are described as optional additional equipment for the Multiplus? (The MP quick installation guide shows just 2 thin wires to connect to battery + and - with 0,75mm² according to handbook)
I now connected the V-sense cables. One end is the green connectors in the main multiplus and the other end is the bus-bar, just where the batteries are connected to the bus bar. I considered this as close enough to the batteries, since cable length is always 1m and I have 90mm².
Unfortunately no change in behavior - The MPIIs stop inverting at 3450W.
When doing the modification, I also checked the length of the cables from the busbar to the MPIIs and they vary between 2 (nearest MP) and 3m.
Does anyone have experience with such constellations - might this be the root cause for the limited inverting power?
Measure the actual voltdrop on each cable at maximum output from each Mp from busbar to Mp and between busbar and batteries on both polarities using a decent DMM like Fluke.
Also In VE.Configure under Inverter → General, you can set “Limit inverter power”. This caps the maximum AC power the MultiPlus can deliver from the battery. Useful if you have a battery with limited discharge current.
90mm² battery cable per meter is approximately 0.0034 to 0.0040 Volts per amp depending if the cable is a copper conductor.The general formula for voltage drop in a DC circuit is: V_drop = (2 * L * I * R) / 1000
V_drop: Voltage drop (in volts)
L: Total length of the cable (in meters), summing the positive and negative cable lengths.
I: Current (in amperes) flowing through the cable.
R: The resistance of the cable per unit length (in Ohms/km or mV/A/m).
Another setting to check is “Grid current limit”. Victron is conservative and if you set a grid current limit of 16A, it will export at most about 15-15.5A.
In any case, the max continuous output of a Multiplus 5000 is 4000W active power.
Understandable and interesting. I was asking myself under which configuration/circumstances does the MultiPlus-II take the V_sense (if connected) into account? Anytime?
I did the exercise properly now under ~100A load from each of the battery banks, 10A only from the smallest MPPT, ~ 10kW load on the inverters and got the following Voltage drops:
Battery 1 => Busbar : +81mV / -97mV
Battery 2=> Busbar: + 84mV / -83mV
Busbar => MP1: +171mV / -71mV
Busbar => MP2: +177mV /-73mV
Busbar => MP3: +181mV / -91mV
This gives an approximate average voltage drop ( BT1=> MP3) of 81+181+91+97 mV which gives 0,45V.
This should be still fine based on the aim <2.5%.
Or did I do something wrong?
Does anyone have experience with the voltage drop difference between MP1 and MP3 of 171+71=242 and 181+91=272 (mV)? Is 0,03mV difference in Voltage drop too much?
4.5.5. Voltage sense
For compensating possible cable losses during charging, two sense wires can be connected with which the voltage directly on the
battery or on the positive and negative distribution points can be measured. Use wire with a cross-section of 0.75mm² (AWG 18).
During battery charging, the inverter/charger will compensate the voltage drop over the DC cables up to a maximum of 1 Volt (i.e.
1V over the positive connection and 1V over the negative connection). If the voltage drop threatens to become larger than 1V, the
charging current is limited in such a way that the voltage drop remains limited to 1V.
See Appendix A for the location of the connector.
So I read it in a way, that it is mainly there for CHARGING to not overcharge the batteries, which perfectly makes sense, especially, if You have no BMS, that will cut-off charging with voltage over-protection. But seems to not impact inverting ;-( (unfortunate for me)
- The MPIIs stop inverting at 3450W.