Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Seplos BMS

Options
191012141537

Comments

  • Registered Users Posts: 1,874 ✭✭✭garo


    3.475 is probably too conservative for protection. I have mine set to 3.65. Yes to what reklamos says about 96%. It won't charge again until all the cells' voltage is below monomer overvoltage recovery AND capacity is less than 96%.


    I can't be 100% certain but it did seem to me that my Growatt wasn't getting energy from the battery after more than 3 minutes at 100%.



  • Registered Users Posts: 34 Fred7631


    Thanks. Still can't work out how the under-voltage alarm/protection is going to work correctly. Solis is only powered by 48V DC rather than AC.

    When the BMS hits under-voltage / % alarm it drops the amps to around 2A which is (I presume) the amount the inverter is drawing.

    This then continues to drop until under-voltage / % protection is triggered and after 1-minute this powers off the BMS which in turn powers off the Solis.

    Which of course means the Solis isn't awake to send a charge during the night when it needs it. I imagine I'm missing something here. Does anyone have the same issue?



  • Registered Users Posts: 793 ✭✭✭reklamos


    You should configure the SOC discharge on the inverter before it hits undervoltage or capacity protection on BMS. Once inverter reaches discharge SOC it stops pulling power from battery. You can see on BMS current goes to 0 and sits there till it starts charging again.



  • Registered Users Posts: 34 Fred7631


    Funny you say that as trying that now but Solis discharge is set to 10% and 8% force charge and it's just gone through both of them without charging :( It did lower discharge limit to 2A @ 10.5%.

    Also, the inverter will still be pulling power to stay on whether the BMS says 0.0. or not?



  • Registered Users Posts: 793 ✭✭✭reklamos


    Some users had issues where FOC would not kick in. Instead of FOC I use backup mode with grid charge, this has not failed on me yet. Basically you set SOC(20%) above your backup(15%). The invert would stop discharge at 20% if power cut to occur it would use discharge from 20% to 15% before cutting. If the battery is not charged for long period of time(days) it will naturaly slowly discharge but since backup mode is configured once the battery reached 17-18% inverter charges it back to 20%.

    If BMS is saying 0 current than it means 0. I have seen on some inverters where inveeter would say 0amps but Seplos would still be showing -0.6-1 amps but these were not Solis. In my case once the discharge SOC was reached I have not lost a single % overnight.



  • Advertisement
  • Registered Users Posts: 377 ✭✭munsterfan2


    Hi,

    Still having issues with my setup, yesterday ran perfectly, with battery charging / discharging between about 80% and 40% but this morning when battery was at 40% and morning load came on initially draw was -11.52A @ 52.25v with cell voltages ranging from 3.174 to 3.238 but with cooker, microwave etc, draw increased to 58a and the cell voltages start to drop quickly, especially C9 which hit 2.9 giving a remaining capacity alarm with a total voltage of 48.3 and SOC dropped to 10% (20ah) which triggered the Seplos to do a charge.

    With charge rate at 62a the voltage of the cells jumped again giving a total of 51.87, ranging from 3.269 but with Cell 9 very quickly jumping to 3.5 and triggering a "Monomer High Voltage Alarm". This drops charging to under 10a which results in C9 dropping to 3.427 within a few seconds and charging jumps again to 60a and within 10s it again jumps to 3.5 and this repeats.

    Am I correct in thinking with a 200ah battery 60a is .3c so there should be no issue at this charge / discharge rate. Battery functions fine at the 20a / 30a charge/discharge rates.

    Any suggestions, I have re-terminated the wires, polished the connectors & battery contacts with steel wool, moved the cell but issues move with it. Is it worth charging overnight at 20a, redoing the top balancing and resetting the Seplos ?

    Thanks



  • Registered Users Posts: 793 ✭✭✭reklamos


    you could try to do top balancing again but it could be that you have a bad cell



  • Registered Users Posts: 2,309 ✭✭✭SD_DRACULA


    Hmm if it moves with the cell maybe a bad cell?

    What if you halve the amount of cells you have, leaving out the possibly bad cell too?

    If it works ok then you know the cell is bad.



  • Registered Users Posts: 3,389 ✭✭✭randombar


    I'm currently in talks with Solis about updating my firmware, apparently it can't be done OTA.

    I'm going to push to get it done anyway, does someone know of a battery that only works with the latest Solis Firmware? I've mentioned Seplos but they have come back saying they don't support that at any firmware.



  • Registered Users Posts: 377 ✭✭munsterfan2


    AoBo if not listed in your current firmware will get you an update ( did for me )



  • Advertisement
  • Registered Users Posts: 34 Fred7631


    Does using backup mode with grid charge allow you to set charge and discharge times?



  • Moderators, Home & Garden Moderators Posts: 6,042 Mod ✭✭✭✭graememk


    Found this on thingiverse https://www.thingiverse.com/thing:5265259/files

    Its a cover that clips over the main bit of the circuit board, Standoffs, and a screen case (and i think the screen case can mount directly on the clip on bit)

    About a 10 hr print for everything.



  • Moderators, Home & Garden Moderators Posts: 6,042 Mod ✭✭✭✭graememk


    First print went ok.ish, it's in PETG, only the screen case lid had first layer issues in one spot, maybe have to dial in the first layer settings a bit more.

    Takes 9.5 hrs to print everything and I don't have any fittings to assemble it either ( I think it's m3), and some IPA to keep the bed clean.

    Won't be able to do anything more on this for a few weeks though. But after that (and when I work out how much it costs me) If someone wants one tag me in the thread.



  • Registered Users Posts: 34 Fred7631


    Has anyone come across an issue on the Solis where the Overdishg SOC: % gets stuck on an old value even though it shows as changed? I set mine to 15% a couple of days ago but now I can't get it to change. Discharging slows right down at 15.5% as if it was still set like that. Tried powering off AC+DC + complete shut down of BMS and multiple restarts and changing the discharge figures up and down and saving them but no joy. Any thing else I can try?



  • Registered Users Posts: 793 ✭✭✭reklamos


    Yuo can still use scheduled charging with Backup mode.



  • Registered Users Posts: 1,874 ✭✭✭garo


    Make sure your connections are tight. Worst case scenario is that you do have a bad cell. The same cell shouldn't hit UV and OV protection. That makes it unlikely to be a top balance issue and more likely to be either a poor connection or a bad cell.



  • Registered Users Posts: 377 ✭✭munsterfan2



    Have some time tomorrow so going to move the cell, and check all the connections. Was reading on diysolar about graphing cell voltage vs current to see if any cells stand out. Will try that. What voltage difference is acceptable when discharging. BMS sit's at about 18mv when idle but as per image hits 60mv @ 14amp but I have seen it climb over 100mv at higher currents. with C8 always the lowest





  • Registered Users Posts: 793 ✭✭✭reklamos


    Was able to build an awful solution to pull seplos stats remotely to Grafana. It still requires a lot of work but it is start.




  • Registered Users Posts: 377 ✭✭munsterfan2


    Hi,

    Looking at Reklamos very pretty graphs I went and plotted a few today. I seem to have a 2 batches of 8 cells that are responding together: The top group of 8 cells consists of Cells, 14, 15, 3, 9, 12, 04, 10 and 11, all within 5mv at the start of the graph. The lower group of 8 cells react much more sharply to higher current ( charge & discharge ) are 10mv away from the other cell group and have a range of 22mv from top to bottom, Cells 7, 6, 16, 5, 13, 8, 1, & 2.




    Second chart is the cells and the discharge current (dotted line right hand axis) just as the SOC drops from 40% to 10%, which causes the inverter to charge back up to 15%



    I think looking at this that I need to redo the top balance of the cells, especially the lower 8. If that doesnt fix then it looks like 8 of the cells have a different impedance and I need to talk to OYE. Does that make sense ?



  • Registered Users Posts: 1,874 ✭✭✭garo


    First check your connection between the top 8 and bottom 8. You need to screw the busbars on pretty tight. Also make sure your BMS sensors are above the busbar and not between the cell terminal and the busbar.


    How are you guys pulling in the data from the Seplos? I am very interested in replicating your charts.



  • Advertisement
  • Registered Users Posts: 377 ✭✭munsterfan2


    Hi,

    Yup all busbars under the sensors, and nice and tight. the 8 behaving cells 3, 4, 9, 10, 11, 12, 14 & 15, with the misbehaving cells are 1, 2, 5, 6, 7, 8, 13 & 16. For the graphs I just click realtime in the BMS monitoring software and then save that data out to excel every morning manually. I will automate it when I have time, try get the csv data into influxdb then to grafana.

    The graphs really do highlight that some of the cells are OK and some are not which was hard to see just looking at numbers.



  • Moderators, Home & Garden Moderators Posts: 6,042 Mod ✭✭✭✭graememk


    A loose connection will usually throw one cell off (low on discharge and high on charge)

    When fully charged do you get all cells above 3.4 (I usually aim for above 3.5) but getting them above 3.3x takes the longest.



  • Registered Users Posts: 430 ✭✭Geeyfds53573


    Modbus over RS485? oh do share when you pretty it up 😀



  • Registered Users Posts: 793 ✭✭✭reklamos


    It is modbus on esp8266. Have second battery hooked up and been looking at the dashboard 24x7 :D. Only pulling stats but would like to pull alarms also.

    The solution is not clean or refined. There are way too many dependacies for it to work.



  • Registered Users Posts: 377 ✭✭munsterfan2


    Just an update, redid the top balancing, wired all 16 cells together rather than 8 & 8. Ran until current was down to 0.04a, disconnected and let sit for 16hours. Then ran a discharge at 50a (.25c) and immediately 8 of the cells started to diverge. The same 8 cells as above, although now in different locations. Discharged about 110ah before tripping the low voltage alarm. The 8 good cells were still reading 3.1x, looking at the labels on the batteries the 8 good batteries are labelled 124-### while the bad ones are labelled 79-### so looks like all the bad ones came from the same batch and have a capacity approx 120ah.

    I've contacted OYE so will see how that goes.



  • Registered Users Posts: 6,130 ✭✭✭championc


    Unfortunately, the response is likely that they will be happy to replace them - if you return them to them in China



  • Registered Users Posts: 377 ✭✭munsterfan2


    Yes, there are a small few complaints on aliexpress store with similar issues, no resolutions. BTW, in the meantime I guess I should drop the battery capacity in the SEPLOS to 130ah, it was reading 84ah when it cut out due to low voltage and dropped to 20ah



  • Registered Users Posts: 6,130 ✭✭✭championc


    The Seplos should "learn" the capacity with a full charge and discharge. I think if it already did this previously, I think you need to tell it to do it again.



  • Registered Users Posts: 377 ✭✭munsterfan2


    It didnt manage to learn previously, would jump from 140 to 200ah on charging then from 80 to 20ah on discharge. I know @reklamos mentioned you needed to be within 20ah for the BMS to learn. I think I'm well outside that, hence the issue. Have asked seplos.



  • Advertisement
  • Registered Users Posts: 6,130 ✭✭✭championc


    Bad time of year too now, to be trying to use the whole capacity



Advertisement