janco93

Im so sorry for bothering you all, i will now copy paste here what artifical inteligence write to me when i was asking about it. Im notnsure if it is right or wrong, it is quite long piece of text but for me it makes a sence but i dont trust it 100% that is why im asking here in forum. AI chat : (We've reached the absolute core of the issue, where at first glance, electrical engineering logic dictates this is an unsolvable "Catch-22." Your reasoning is flawless in that if the current split like this through a standard domestic Residual Current Device (RCD), it would trip immediately.
But here is the key to why this actually works on Deye inverters in "Common Neutral" mode without throwing a leakage current error (Leakage / F24).
The entire secret lies in the inverter's hardware architecture and EXACTLY WHERE and WHAT the internal sensors (RCMU) measure.
1. Where is the RCMU sensor located in the inverter?
Your concern stems from the assumption that the inverter has a leakage sensor directly on the Load N and Grid N terminals, measuring the physical current flowing through them. However, this is not the case.
The internal RCMU (Residual Current Monitoring Unit) in hybrid inverters does not act as a standard domestic RCD for the bypass circuit. Its regulatory and technical purpose (according to the IEC 62109 standard) is to protect the AC grid from direct current (DC) and high-frequency leakage that could infiltrate from the PV panels, battery, and the inverter's IGBT power bridge.
Therefore, the RCMU sensor is physically located at the output of the power inverter (after the generation stage), before the circuits split via the internal bypass relays to the Grid and Load terminals.
2. What happens during grid operation (Bypass Mode)?
Imagine you have a "Common Neutral" setup wired—meaning both Grid N and Load N are bonded at your main domestic PEN busbar. You are connected to the grid and turn on an appliance in the older (TN-C) part of the house.
The operating current flows from the grid, passes through the Grid L terminal, goes through the closed internal bypass relays in the inverter, and exits via Load L to the appliance.
It returns via the old wiring (PEN) to your main distribution board onto the massive PEN busbar.
And here's the trick: From here, 99.9% of the current flows straight out through the thick main cable back to the grid (street).
Only an absolute fraction of the current would theoretically "loop" through the parallel connection of Load N -> inverter -> Grid N, because the voltage potential difference across a single solid copper busbar is practically zero.
3. Why doesn't it throw an error?
Deye doesn't detect any "Leakage" because:
It bypasses the sensor: The operating current flowing from the grid to the appliance and back (bypass) doesn't pass through the inverter's active power bridge at all. Consequently, it doesn't pass through the internal RCMU coil that would be looking for asymmetry.
Power measurement is on the phases: The inverter measures the operating currents and household power consumption using its own Current Transformers (CT), which are placed solely on the phase conductors (L1, L2, L3). The inverter doesn't care at all which path the current takes to return to the distribution transformer; what matters to it is only what it measures on the phases relative to neutral (voltage).
4. What about battery operation (Off-grid)?
When the grid goes down, the inverter starts generating power itself (via the aforementioned power bridge).
Power flows from the bridge -> through the RCMU sensor -> out through Load L.
It passes through the appliance and returns to your PEN busbar.
Since the grid is disconnected, the current must return to the source, which is now the inverter. From the PEN busbar, it travels via the neutral wire to the Load N terminal and back into the power bridge.
The RCMU sensor inside the inverter sees that, for example, 10A went out on the phase and 10A returned on the neutral. The vector sum is zero. The inverter is satisfied, reporting no leakage.
Summary of this paradox
Nothing multiplies, and no error pops up. Your concern would be 100% justified if you ran Grid N and Load N through an external 30mA RCD in the distribution board—that would trip in a heartbeat.
However, the Deye inverter hardware inherently handles this topology (Common Neutral) without error messages because its internal leakage detection system is not designed to monitor the return path of bypass current, but exclusively to monitor its own power generation. This is exactly how hundreds of hybrid solar systems are successfully running on older grid systems (TN-C) across Europe.)

Ok so i can confirm, if i disconnect one of inverter solar input and connect to separate charger everything work great. So solved problem for now but itnis not ideal solution for me, anny suggestions to how connect it back and not flicker? 😀

watchpower for me was useless, until i found  MultiSIBcontrol.
it happend a few times with watchpower where it wouldnt display any PV readings until i set all the watchpower settings to default, and then it would only display the pv input
try this MultiSIBcontrol
http://www.multisibcontrol.net/