Here’s a strange thing…
The system consists only of a MP2 3K GX and an 48V/100Ah battery.
The battery is slowly discharging and when it hits 30% SOC the system charge it back to 100% and the cycle repeats.
Below a graph of a cycle.
Discharging from 100% to 30% takes 100 hours.
From 100% to 30% you could say, in average, discharging 70Ah from the total of 100Ah.
From the current reported from the MP2, which is -1.3A, on 100 hours, you are discharging 130Ah…
Strange, isn’t it?..
Quite big error when measuring current…
Am I missing something?
I have seen a couple of different manufacturers also trigger false 100% soc.
It happens when the charge source drops off suddenly even if the target voltage is not met.
Have you opened one to see the ‘current measurement’ circuit? There isn’t one. I know you spend time on the forums (and in the insides of inverters) i am surprised this is not a piece of information you know?
This one is not mine and I haven’t opened it… Yet. 
Name it whatever you want, it still is a “measuring current” process… 
Probably measuring voltage drop over a properly calibrated low resistance. Just like the shunts.
As per the battery and its SOC, you can see that it’s charged to 3.51V per cell (56.2V), it’s sitting there about 2 hours to be properly balanced at a deltaV of less than 20mV and then, during discharge, all voltages and SOCs are properly synced.
In other words, the Shinwa BMS used is properly calibrated. So 30% SOC voltage is spot on.
The load is pure resistive and constant.
I will make time to reach the objective and properly measure the current with a proper DC clamp.
Comparing with Multi RS, this MP2 3K is all over the place when measuring current.
Look below when charging how looks the range of regulation. And it’s not in the most worse area…
It’s set for 14A charging and it is “regulated”, if you can name it that way, between 12A and 16A.
The multi rs has a shunt type set up in it. The MP2 from what i can see does not. It is more if a calculation.
Thats a nice predictable graph there.
2 Likes
Technically the +/-15% don’t matter, but optically its a shame…
It must have for sure something for measuring current… It can’t be otherwise.
Otherwise, one could change its opinion on the quality of these products.
I don’t know, as I’ve said, I didn’t opened one, but at least, maybe, on the drain pins of the mosfets could be some equalization resistors and some sampling on these very low resistors for the current…
Will know for sure after extracting the schematic.
Certainly Victron people wont come forward to tell… 
Yes hall effect on the ac side. Which if i am not mistaken around 4A at mains voltage (if you are charging it from there with no load would be on the threshold of where it reads ok. I could probably check the documents for them as this also affects the input current lower limit as well in some systems)
I forgot to tell, because I believe it’s irrelevant for the situation, that when discharging/inverting and measuring that erroneously -1.3A instead of about -0.7A, the AC-IN is/was disconnected…
You can’t measure DC discharge current based on AC current.
Well… you can, but It’s a very bad idea.
Although I wouldn’t be surprised, considering this: Venus error in computing DC power?