That the charge Voltage is raised, when feedin is enabled, is “normal” - Otherwise the battery could not be hold at 100%. If battery and MPPTs would run at the same voltage, the multis would draw power from both, when trying to feedin.
That would be acceptable, but Victron is still moving with this voltage up and down in regione of 0.4V which results in battery charging and discharging and charging and discharging. 55.2V wants BMS, Victron adds 0.4V fine so total is 55.6V. But MPPTs voltage is moving from 55.2 to 55.6…just a 0.1V would cause problems.
By your answer “to keep battery up” i fell like you admit that battery must disable charging mosfets…because that is a “old style” behavior.
So, the MP has to act as “Man in the middle” to grab the power from the mppts, but leave the battery “idle”.
Yeah this is the issue with DC coupling design and I think it is just wrong for these types of BMSes (which there will more and more). If battery wants some voltage, you just cannot move up and down as you like, otherwise = current flows. For MPPTs to output more power, they need to move with voltage → problem. What is a problem and victron could react to is, when battery is discharging power with feeding enabled goes to grid…why? Victron see that so it can increase voltage a bit to hande it.
And to the end, you have on victron community few threads with this issue, I have access to a lot of installations with this issue, it is not about resistances.
The Voltage Up/Down is not by design. My System (2 MPPT, MPPT RS and Pylontech) is not doing that - at least not at the values your graphs show. There are a lot of users using BMS Like the JK with no Issues as well.
I think you should consider it’s an hardware issue, that may cause this. It could be caused by the installation itself (wires, terminals, breakers, fuses), or ofc. by a single defective unit, that has an Issue to keep a constant voltage. (Broken capacitors or sth.)
I would try each MPPT alone, see if the issue is present, if it just happens with multiple mppts, if it just happens with ONE certain mppt, then eventually swap the mppt-connections, see if the problem “moves” to identify whether it is “path” related or device related, etc.
It may not be caused by the mppts, could be caused by the BMS itself or by the multis. Just from voltage curves it’s not really possible to tell that. Hence, starting with the easiest to “remove” parts may be a good idea. If you are lucky, you immediately find a setting without issues, then you know “its one of these 3” in either way.
All JK users with enabled export have this problem. JBD with better settings (charge voltage setpoint) has this issue as well.
It is not about resistances and HW issues. Check video and focus on MP voltage (DC bus voltage) based on solar array power. Less sun → lower voltage → battery discharges but to grid as well. More sun → higher voltage → battery charges.
In this moment BMS wants 54.8V from Victron. It gives maximum 55.2 based on 0.4V addition. But fluctuates between these two values.
Pylontech is not doing that, we need to find out why. My guess is charging mosfets disabled, bad BMS voltage calibration and maybe combination with Victron 52.4V limit.
My next tests will be - control voltage by victron, dvcc used only for SOC.
Set JK to disable charge mosfets when on 100%.
@dognose Have you seen my video? I’m experience similar problems as well. I’m using a JKBMS (V15) (2x = 32kW) latest firmware, Multi RS Solar, 2x MPPT.
OK my first test to encounter charging discharging problem when export is enabled went as predicted. I simulated old fashioned “charge to the limit” BMSes.
BMS settings: RCV set to 3.599V per cell (totall charging voltage 57.5V plus 0.4V from Victron = 57.9V) , OVP to 3.6V to be sure I do not overcharge the cells but still hit OVP alarm and force battery to disable charging mosfets… And guess what… it works… 0 current flow. Battery is kept charged at 100% while still easily exporting to the grid.
What is a “problem” is that battery in this case battery sends OVP alarm to Victron. If user is informed that it is normal, it can be acceptable.
Maybe I will count with the 0.4V addition and set charging voltage 0.35V lower and still hit OVP. I will send updates.
Next test with DVCC on but voltage by Victron later.
