RedDragon

Hi Freddie. I Would recommend changing that e/l to a Type A(IR) Gewiss Earth leakage (GWD4208). They are bit expensive, but worth it. The Kingstar is no good. The type A(IR) earthleakage helps especially when there are possible pulsing DC current present and also has a greater resistance against surges and atmospheric discharges. Also the contactor move you have going on there to connect neutral to earth I am unsure about. Disconnect that for now and make sure your incoming neutral and earth from Eskom or your main earth from your earthing network is properly bonded. It can be as Frivan says, but sound to me like a DC pulse issue. Also test your earth continuity, must be below 0.5ohm, but try getting it as close to 0 as possible by implementing trenched earths. Also, take a clamp on meter and measure if any current is present on your incoming earth conductor when your inverter is supplying any load. There should be none or very little. If there is any. improve your earthing. Are your pv panels grounded to negative? make sure that is not the case.

Re did earthing to inverters and also from output with 10mm individual house wire each one going seperately to earthbar of combiner box and got rid of that current on the earth conductor. Also did an earth continuity test at incoming supply, got 0.7ohm which is not ideal. Noticed that the installation of bundle that feeds to property does not have a seperate earth, but it does have a earth trench with spike where it enters property and is only connected neutral to earth at transformer point which is about 350m away. So did a neutral to earth connection where bundle enters property and improved the earth continuity to 0.2ohm. Thats mucher better. So far, so good. Haven't experienced that same fault again since. Switched AC input off and on about 10-15 times with different types of loads on all above 2kw and switch over is happening perfectly now. So it seems that was it. Fault resolved until further notice. 🙏 So to sum it all up. It seems that parallel inverter systems are much more likely to cause floating earth issues which can cause this sort of problem, especially when your incoming supply is not properly earthed and connected neutral to earth. Thanks for all the tips and pointers @Coulomb. Next time I will buy you that coffee.

Sent this this morning. only went through now.

Checked firmware just now .U1 is 72.00 on both and u2 is 36.03.U3 is 00.00. Battery voltage per module is on 53.6 when fully charged.Charging voltage is 56V. I did measure a current on the earth. It was 0.630A AC on the main earth when inverter is on and receives AC input and when inverter is in bypass and battery mode and gives an AC output to load.

Have a seperate earth spike installed which goes to inverter DB and connects to main earth from Eskom supply with 10mm earth wire. From there it distributes to each of the inverter inputs, inverter outputs and batteries. Going to install a 40ka 2p surge arrestor before input aswell or do you think i should rather do it on output? Want to test the incoming earth directly from line and follow it back to transformer which also feeds a pump station to check if eskom earth is still in one piece. They love stealing the eskom earth wires around here.

They definetly can take the load. I have pushed them to about 8kw (4kw on each) load output for about an half an hour continiously and the heat increase on the cables between inverters and batteries go to about 27°C. Used The FLIR thermographic cam to measure. But I agree completely that thicker is better in this case. So I am going to double up with an additional pair of 25mm² cables from busbar to each inverter. Can't recall the u1 now, but i know the U2 version is 36.03. Will check tomorrow and relay that. They do share load currently evenly, did check that. Thanks for that one. Did not make it to the site today, was sucked into cooling units of wine cellar on my way there.😅.Will try that tomorrow before doing an earth continuity test. There is definetly something funky going on with the earth. Have a AVS unit before inverter input, which I disconnected because it goes crazy when Mains/AC input is on. Keeps hopping around between low and high voltage warnings when there are none present, I measure a stable 230 -233v when this happens. Also the First time this happens. We typically use AVS units in our installations to keep Eskom surges and low voltage issues at bay before AC Input. They provide Low voltage, high voltage, spike and surge protection which can be set on a delay of between 10s to 10min intervals. When a fault is present as set via high and low voltage adj setting the unit will wait out its set delay time and will only reconnect the AC input with output once the preset fault has been cleared. My idea to use the AVS units to delay the loads did not pan out either, spoke to the SA distributor of Sollatek products and he pointed out that the AVS units they make are limited to minimum 1s 'blind' time and will ignore the change over procedure of inverter. Wanted to set the delays on two seperate AVS units to switch heavy inrush currents of Aircons and refrigerators at different delay times as to limit the inrush current during the change over process, but im am actually glad it won't work because that will not solve the cause of the issue at hand , will only bypass it.

According to the Kodak OGS 5.6 manual on page 50 under the parallel section the recommended battery capacity is 200ah. So i assumed that the two fl5.12 (100ah each) batteries would not be a problem when i was planning the installation. Also the the rated max,continuous discharge current of batteries is 100a. Same for max. charging current. the peak discharging current (30s) is 150A and peak charging current (30s) is 105A. Setting 2 on inverters are automatically set to 20A for max charging current on both inverters after setting 5 is set to pylon on master inverter. Almost on my way to site. Will measure the voltage of batteries when i get there.

Hi. Yes switches are on and and does provide output to loads and charges batteries at same time. Battery cables are 25mm² from each battery to busbars and also 25mm² from busbars to each inverter (through disconnectors).Would you recommend we upsize to 35mm² from busbars to inverters? I have also been speculating that maybe fault occurs as a result of over current on the DC side which causes BMS to cut out communication. Funny thing is only that the batteries do not display any fault when this happens. All the lights on batteries are still the same as they were before the inverters shut down. So i am assuming if that is the case it is the inverters that cause the cut off and not the batteries? Do you think I can test this by connecting a Automatic voltage switch to high loads(fridges,AC units) after inverter so that the high inrush current from compressors can be reduced? Haven't connected a AVS after inverter output before, will it be able to read the voltage loss in the change over process?

If anyone has any pointers for me I would greatly appreciate. 🤕

Good evening Comrades. New to this forum and i am hoping someone might help me with the solution to my problem. Have been able to find a couple of good pointers and solutions previously just by reading all the different topics, but this time i could not find a topic that fits my exact problem. Installed x2 Kodak OGS 5.6 inverters paralleled in single phase which are connected to x2 Kodak FL5.12 LifePo4 batteries, no PV panels yet. Both inverters shuts down and loses comms after AC input is switched off. Have to reset the system after that to get comms back up. Only happens when load is above 3kw (1.5kw on both inverters). Replaced the communications cable between master inverter's BMS port (which has setting 5 on Pylon) and master battery's RS485 port and get the same result. Also slave inverter is set on User @ setting 5. The parallel cables and current sharing cables are all connected exactly according to manual. My first time connecting the OGS 5.6 in parallel. So i am wondering if maybe I missed something, but i am very sure I did not. Double checked everything. Also the firmware of both inverters are the same and I comfirmed that with one of SEGEN's Tech guys who are could not really help guide me to a solution. I did notice a floating voltage on the neutral output of inverters. Measures 76V AC between earth and neutral, but am unsure if that can be the cause of the issue i am having. Going to try and get rid of that tommorow and am also planning to test each inverter individually and see if it happens with only one of them or at all. That way I might be able to determine if it is an individual inverter fault or if the fault lies with the parallel communication. Have no monitoring devices connected currently and are also struggling to connect watch power as our Internet and cellular signals are messed up at the moment, due to loadshitting.