we’re seeing an intermittent start failure with a three-phase compressor load on a 3× MultiPlus-II system and would appreciate input from anyone with experience of large motor loads.
System
3× MultiPlus-II 48/6k5/100-50, three-phase setup on VE.Bus
Cerbo GX], DVCC enabled
3× Victron Lithium NG 51.2 V / 100 Ah, Lynx Smart BMS NG, Lynx Distributor
Inverter-only operation (no AC input connected during the tests)
Load
Three-phase compressor motor, approx. 9.3 kW, rated approx. 20 A at 400 V, direct-on-line start, reciprocating compressor starting under load
Running current approx. 18 A per phase (approx. 2.75 kW per phase)
Measured with a power analyzer on grid supply: inrush peak approx. 137 A per phase
What happens
Starts are intermittent: sometimes successful on the first attempt, sometimes on the second, sometimes not at all.
Successful start: a loud growl from the inverters for about 1.5 s, then the motor runs normally.
Failed start: loud growling/humming from the inverters for about 5 s, then overload on all three phases and shutdown.
During a failed start (VictronConnect live values, see video): approx. 60 A per phase on all three phases, output voltage dropping to 143–214 V, 5.2–6.8 kW per phase. After approx. 5 s: “Overload” alarms, output 0 V.
Currents are roughly symmetrical across L1–L3, so it doesn’t look like a missing phase.
It happens independent of SoC (seen at approx. 35 %, 56 % and 95 %) and independent of whether the motor was running a few minutes before.
The noise clearly comes from the inverters, not from the motor. Its spectrum is dominated by 100, 200 and 300 Hz.
Video (17 s, live values plus sound - attention: very loud noise):
Questions
Is ~60 A per phase for ~5 s the expected current limit and overload behaviour of the Multiplus 6k5 in this situation? Are there VEConfigure settings that influence how the inverter behaves during a motor start?
Is this kind of loud growling under overload considered normal, or does repeated exposure risk damage to the inverters?
Has anyone started a comparable DOL compressor (approx. 9 kW, starting under load) on a three-phase Victron system? What worked: soft starter, VFD, larger units (48/8000 or 48/10000), or more units in parallel per phase?
VFDs are normally sized with upstream transformers taken into account as well. Due to harmonics, VFD loads must not exceed 40% of a transformer’s capacity.
Applied to the MP2: you need at least three 8000s. They might manage even without VFDs, though, if the battery voltage doesn’t sag. As a 16s pack, it needs to be able to deliver at least 800…1000A momentarily. Probably even more, but I’d experiment with some hefty capacitors.
We’ve since measured the motor start with a power analyzer and checked six months of VRM data:
3× MultiPlus-II 48/5000: output current limited to about 37 A per phase, no start succeeded.
3× MultiPlus-II 48/6k5: limited to about 66–70 A per phase, most starts succeed, some stall for about 5 s until overload.
On grid, the motor draws 125–146 A per phase.
DC current during a stalled start reached about 500–540 A for a few seconds.
We’re now considering 3× 8000 or 3× 10000 and would like to understand:
What maximum output current (RMS, per unit) can the 8000 and 10000 deliver during a motor start, and for how long? As far as we know they are built differently from the 6k5, so we don’t want to extrapolate.
What short-term overcurrent can the Lynx Smart BMS NG (500 A) and a single Lithium NG 51.2 V / 100 Ah handle, for example for 1 s and 5 s?
I didn’t think there was much difference between the 6k5 and the 8000 on the AC side tbh. They switched from naming the VA capacity on the 8000 to the Watts capacity on the 6k5
The 8000 has a better DC charger but that’s not gonna help you in this case.
Thanks. Looking at the datasheets, I agree the continuous ratings are close (6k5: 6000 W, 8000: 6400 W). The peak ratings differ more though: 10.9 kW (6k5) vs. 15 kW (8000) vs. 18 kW (10000).
What makes me unsure is our own comparison: the 5000 is rated 9 kW peak and the 6k5 10.9 kW, only about 20 % more, but we measured 37 A vs. 66–70 A during the motor start, about 80 % more. So the peak rating doesn’t seem to translate directly into current-limit behaviour.
Does anyone know whether the 8000/10000 current limit during a motor start scales with the peak rating, or has measured it?
Just for some reference, a 3 phase system with 3 x 15000 MPII, 8 x Dyness PowerBrick Plus.
Average loading is 15-18kW and still perfectly able to start a 5.5kW screw compressor with star-delta sequence and occasionally, transformer based 4kW electric forklift charger.
On DC side I see regular spikes of 5-600A.
This is for a cloudy day. If there is sun, the Fronius ECO 27kW, on AC Out, makes Victron equipment job much more easy.
Inverters, aside from double internal fans, are exceptionally quiet, no growling or strange buzzing.
For your system, please consider more, large capacity batteries.