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

10kWh LifePO4 DIY Battery Addition..... Here is my write up.

Options
1246750

Comments

  • Registered Users Posts: 299 ✭✭spose


    The time stamps are wrong. Not sure where it’s pulling them from but it’s not linked to the laptop. Looks to be ~7 hours off. At that time it was reading 65% soc. Lowest soc it made it to at any stage was 48%. Have it powered off now. Part of the oddity is when the solar is generating the alarms reset themselves and the me3000 works again. When no solar is generating the me3000 doesn’t come back out of standby



  • Registered Users Posts: 6,122 ✭✭✭championc


    What version of firmware are you running on the ME3000 ? Did you set it up for General Lithium ? If you go into System Information, on page 4, does it show the battery as a SHEnergy v0.0.1 ?



  • Registered Users Posts: 793 ✭✭✭reklamos


    That's a lot of errors. The errors suggest that you have some bad cells as you get 'cell differential' errors. I cannot remember exactly but the difference needs to be 0.3 or 0.5V for this to get triggered. Have you top balanced your cells? 'Output short' would suggest that someting on the inverter side but it could well be seplos.

    The best way to get to the root of the problem is to constanly monitor seplos, each cell and inverter stats.



  • Registered Users Posts: 299 ✭✭spose


    Thanks both. Settings are as you ask champion. Cells were balanced but did have a week before they were connected. When running difference I’ve seen is on the seplos is 30mv typically between high and low. Taking one example when showing high voltage alarm the highest cell reading is 3.5 lowest is 3.384. For other high voltage alarms the individual cells are all below 3.5. From watching the alarms are definitely appearing on the seplos first and then registering on the sofar



  • Registered Users Posts: 299 ✭✭spose


    Feedback from seplos…these alarm information will definitely be sent to the inverter, and it cannot be prevented unless your inverter can make the BMS discharge and charge alarm value not reach, that is, turn off the charging before the BMS is charged to the highest, and turn off the discharge before the discharge is the lowest. , you can take a look at your inverter manual to see if this function is described.



  • Advertisement
  • Registered Users Posts: 793 ✭✭✭reklamos


    3.5 and 3.384 is 116mv diference which is not bad but it seems to me that this not near the top of the charge as voltage is low. I have also some cells that start running away but this happens when soc is ~90% and the cells are at 3.45v

    Check waht diference you get at the top and bottom charge. You can configure logging on PC so it dumps data to file and you do not need to eyeball it.



  • Moderators, Home & Garden Moderators Posts: 6,036 Mod ✭✭✭✭graememk


    For LFP 3.384 could be anywhere from 50-95% though. The cells could be a little bit out.. or a lot.



  • Registered Users Posts: 299 ✭✭spose


    the fact that it can give high voltage and low voltage alarms within 20 mins doesn’t seem right. Reading the reality one log all the voltages look ok even though low voltage alarm says voltage was 0.2v on cell 1. In a closed box and checked cables and all seem fine.



  • Registered Users Posts: 299 ✭✭spose


    So powered it back up this evening to see how it behaved and ran for about 2 hours with only 2 alarms. There was no solar for charging during that time. discharging ranged from low couple of amps to long enough periods of>50amps. To make it awkward the time stamps on the realtime log and the alarm log don't match. Real time log is actual time and alarm log is 7hrs 11mins ahead. For the last alarm (06:50) on the alarm log shows low voltage on cell 1 (0.2V) cell 2 (0V). All the previous low voltage alarms seem to also be 0.2V. The realtime log shows the voltage stable across cell 1 and 2 at that time (23:38) at 3.2V. Don't want to just keep resetting the battery and inverters just to see how it behaves but I'm WFH tomorrow so will try it while there's a good bit of solar production going on too




  • Registered Users Posts: 6,122 ✭✭✭championc


    Check or reseat the BMS cell monitoring cable connectors onto the board's connectors



  • Advertisement
  • Registered Users Posts: 299 ✭✭spose


    I’ll try that later on. It’s running 4 hours now with only 2 trips. First another low voltage that reset itself. Second caused the bms to turn off. Have reached out to sofar as well with the error logs since a couple of the errors on it say to contact them if they are recurring.



  • Registered Users Posts: 299 ✭✭spose


    Ran the realtime logging for hours yesterday and had a number of alarms reporting low voltage on cells 1 and 2 but the realtime logging never shows a cell drop below 3.2V. Sofar support came back looking to access the inverter but I didnt install a wifi stick so will have to see if they give me any other alternative.

    I left the sofar off for the night and the battery turned on. Idea being if the alarms were due to loose cables or anything I might still get them even with nothing going on. No alarms overnight at least. Can I check something...I've changed the CAN protocol on the BMS to Sofar and using the cable that came with the Sofar to connect to the BMS. Is this correct and I shouldnt be using 485?



  • Registered Users Posts: 793 ✭✭✭reklamos


    BMS must match the protocol on inverter. CAN and RS485 are completely diferent protocols. Seplos has BMS with CAN or RS485 but not both. The ones we use with Solis is CAN since this is waht Solis supports. The protocol is fine otherwise I would expect inverter to complain about missing battery.

    Ignore this jsut seen the logs: As for cells, they do not need to drop below 3.2v, check the diference between the lowest one ant the highest one. Also do you get the alarms on when battery is being charged/discharged at high rates(50-60Amps)?



  • Registered Users Posts: 793 ✭✭✭reklamos


    Inspect your black terminal and white and yellow cables.

    I would do the following tests:

    1. Swap cells 1 and 2 with 15 and 16. If problem moves with cells, you have bad cells. If it stays do next test.
    2. Swap the white/black terminals arround but you will also need to reconnect the balancing cables accordingly on the cell side. If after swapping the problem moves to cells 9 and 10 then the problem is with terminal block/wires. If it stays, it is Seplos.

    If you do this, please do it safely and check everything multiple times.



  • Registered Users Posts: 6,122 ✭✭✭championc


    The Seplos CAN port should be connected to the Sofar CAN port with an appropriate cable.

    I think the RS485 port on the Seplos would be to extract the cells and other data into Node-Red or something like that.

    Similarly, the RS485 port of the Sofar is for reading Live data from a Sofar, or to remotely control a Sofar inverter (via a sofar2mqtt device)



  • Registered Users Posts: 299 ✭✭spose


    Thanks for the help. Will have to leave it now for a week. When I’m back I’ll try that suggestion and change cells 1&2. If it doesn’t point to the cells then I’ll probably strip it down and set it up simpler as a 16s1p and see if it behaves any different with half the cells. If behaving same then might bypass bms and under the default settings and see if sofar still alarms. At least that would definitively say whether the source of the alarms is the seplos or the sofar.



  • Registered Users Posts: 6,122 ✭✭✭championc


    Maybe there should be in a 20kWh thread 🤣

    Hopefully nice and snug for the winter, after repurposing the packaging

    Post edited by championc on


  • Registered Users Posts: 189 ✭✭connesha


    Nice.

    I see you've left some space between the cells, looks like even a few mm between some ones that don't have insulation?

    Is this in case they expand/bulge and put pressure on the terminals?



  • Registered Users Posts: 1,452 ✭✭✭DC999


    Wow, @championc great for a noob like me to see where we can go with this. More a comment in general - it takes a lot of space to have a lot of battery (ministry of the obvious of course). But I didn't think it would be that large. I've seen 5kwh like a puredrive, but not anything larger. You should post that on the solar setup thread :)



  • Registered Users Posts: 6,122 ✭✭✭championc


    The gaps were more the way the ended up after locking up the bolts on the busbars, and also the space I had in the foam coffin. So nothing deliberate going on here. In each 4, I have now put the temperature probe between cells 2 and 3



  • Advertisement
  • Registered Users Posts: 6,122 ✭✭✭championc


    Context is everything. Beside the chars on the left gives some impression of size. These cells are damn sturdy things and I bet several who got them didn't quite expect how big they actually are.



  • Registered Users Posts: 5 hwshaw


    Has anyone bought cells lately and have had a good experience of the seller. That is the cells arrived and were as described with correct capacity?


    I see OYE and PWOD store mentioned here as good. I was close to buying from OYE 16 x 200ah cells. The ribbed metal case style. Then noticed on the OYE store the smooth case version was more expensive. When you read the spec there is different operating specifications....does it matter?

    Also looked at Alibaba some good prices. You see a seller with on the site good review....google the seller and some discussion boards inside and outside Ireland state they are good then others slating them.


    Its a big investment so what to get it right or is it a case of roll the dice?



  • Registered Users Posts: 318 ✭✭E30M3


    In my experience if you find an issue after the 15 day Aliexpress window you are on your own. PWOD/OYE (Same seller) will not give two hoots.

    You could try paying with Paypal but you are rolling the dice with any seller really.



  • Registered Users Posts: 5 hwshaw


    Yes this is my thoughts.....I do question though on OYE for example as they have different technical spec for cells with the same capacity?

    To buy a capacity tester the cost on Alibaba is circa 185 euro for a 40A machine. What's the benefit\risk of diagnosing a bad cell?

    Does anyone have a capacity tester to be borrowed\rented and paid for in notes....or beer?

    I actually think OYE etc are little more than agent\re-sellers from an office they just buy in and sell out.



  • Registered Users Posts: 6,122 ✭✭✭championc


    I don't believe that a single buyer from PWOD or OYE have been left without their batteries. Getting the VAT off at the checkout and them arriving without any further charges is a huge bonus.

    Yes, people have had issues with the capacity , but I think it's more than compensated in the price, and the knowledge gained. I got my second set about 3 weeks ago.

    The CALB are 200Ah cells with a very strong plastic casing, and have large bolts and busbars. The flat cased cells all currently appear to be 280Ah cells, in Aluminium casings, and smaller terminals and busbars



  • Registered Users Posts: 261 ✭✭S'


    Finally getting around to top-balancing my cells. Just waiting on a capacity tester from AliExpress to make sure all is ok before I start into it🫣

    11.2kWp E/W split arra. Mayo



  • Registered Users Posts: 377 ✭✭munsterfan2


    I have a capacity tester, bought it to test my cells to prove to OYe they were crap, not that they cared. Just over 6 months now and one of the cells is gone completely. It did run for about 3 months with an SOC of 125, but now if current goes up to 40a cell 12 just collapses, then jumps back when the BMS tries to recharge. Unless I get a refund from the bank, I have applied for a credit card charge back, this is eur1200 down the toilet.

    Let me know if you want to borrow the tester.




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


    Only one cell looks bad tbh. If you don't get your money back you can always buy a single cell and replace cell12.



  • Moderators, Home & Garden Moderators Posts: 6,036 Mod ✭✭✭✭graememk




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


    I don't think so. Someone asked that in the https://diysolarforum.com The BMS needs to read 16 cell voltages.



Advertisement