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

_dof_

I finally got my battery connected together after many months. I bought 32 CALB cells from PWOD store and top balanced them months ago, but only got them connected to the inverter though he Seplos BMS and the inverter wired to the consumer unit last week. It's a Sofar ME3000SP inverter. and I've wired the cells as 16s2p.

When I tried to discharge it, the seplos seemed to reboot, and it didn't discharge anything. The Seplos had been reporting around 35% full, which seemed a bit low, but I figured it wouldn't be accurate till a few charge discharge cycles anyay.

I noticed it was reporting around 100mv difference between the top and bottom cells, so I put the Sofar into charge mode to charge the battery at 400W. It didn't take long to report 100%, buy still with a 120mv voltage difference between the top and bottom cells, unsurprisingly, since there's not much real balancing going on.

The Sofar won't discharge now, as it's reporting battery low voltage, this is the data exported from the seplos:

I'll top balance again I guess, but is here possibly a bad cell or 2?

unkel

When did you top balance them, just now or months ago? I presume just now.

Don't get too worked up about the difference in mV, I wouldn't be concerned if the lowest was 3.4V when the highest hit 3.65V (250mV difference) but in your case the lowest cell is far too low with the 3.45V cell 99% or so charged and the 3.338V cell only about 85% charged (even though the difference is only 112mV). Could of course be capacity differences between the cells, but less likely in a 2P setup. You could of course capacity test every single cell and then match the parallel pairs based on capacity

_dof_

Thanks, I top balanced them months ago, and they've been reconfigured as 16s2p for a long time waiting to get wired up, so they must have drifted over time and I'll just reconfigure them in parallel and top balance again before trying again.

Must get myself a capacity tester, I think there was an AliExpress link earlier in the thread

slave1

I have the exact same setup, 32 200ah cells via Seplos and Sofar and also top balanced months before I got it all hooked up. Have you fully charged and then immediately fully discharged to allow Seplos figure out your capacity? Then just use them!

I'd to this before taking them apart to capacity check and pair off.

Mr Q

Has anyone purchased REPT cells before? I got an offer from Docan Power for some 280Ah versions which they say are closer to 300Ah in reality.

I purchased 280Ah EVE cells from them before which they said would be closer to 290Ah and they were actually slightly above this.

About €1,720 for 16 cells DDP from China.

unkel

@slave1 - "Have you fully charged and then immediately fully discharged to allow Seplos figure out your capacity? Then just use them!"

They are out of balance and Seplos can't ever fix that. You could of course just ignore it and be happy with the likely 85% of capacity the pack has and not have to bother spending 2 weeks testing the cells (and losing money coming into the best solar PV weeks of the year because of having no battery). I've done this myself in the past.

Alternative solution is to fit an active balancer. Either a smart NEEY or even a cheap €40 one from AliExpress that you just hook up for a week or so and then see what happened

_dof_

https://www.boards.ie/discussion/comment/120521870#Comment_120521870

I tried to discharge as part of the commissioning test, even though I wat the Seplos was reporting differing cell voltages and it thought it was only around 35% full, I thought it would discharge a bit until the lower voltage cells brought total voltage to the equivalent of 15% battery capacity (which should be around 3.2V*16=51.2V), then it should have stopped discharging.

I expected to be able to charge then until the highest cells reached the configured monomer high voltage level (from the file champoinc uploaded) of 3.6V.

But the first attempt to discharge resulted in the Seplos resetting and nothing being discharged. So I decided to force charge it using he inverter to 400W. It did start to charge, but very shortly afterwards it reported 100% full so it stopped charging. The Seplos reported the voltages I pasted above. I assume the max voltage got to 3.6V during charging and returned to 3.45 at rest, representing 100% for those cells).

So, I put the Inverter back into auto mode and with import from the grid it should have started dischagring, but it didn't and the event log showed an ID85 Battery Low error which is described in the manual as:

"

ID85

SOC <= 1 -DOD or Battery voltage is low

For example, if you set DOD to 30%, when SOC is less than 70%, you will see ID85 in the event list. ME3000SP won’t discharge the battery when ID85 is present. Or This is an indication of low battery voltage. ME3000SP won’t discharge battery in this case to ensure long battery cycle life.

"

The DOD is set at 85% in the inverter I think, so it shouldn't reach a battery low voltage until the battery pack voltage reaches (16X3.2v=) 51.2V, but it's still reporting 54.63V in the Seplos output above, so that's at rest. I'm wondering if the voltage drops below 51.2V when under load, but that would be a big voltage drop, and could indicate some faulty cells?

I'm a bit confused TBH, and trying to understand how this all should be working.

_dof_

https://www.boards.ie/discussion/comment/120521967#Comment_120521967

I'm afraid I can't ignore it, since the inverter won't currently discharge the battery due to the low voltage alarm.

I'm hoping a a fresh top balance will fix the issue, but if there's some faulty cells causing a voltage drop under load, then I could have a bigger problem.

I wash thinking about the active balancer option, but I was thinking that if I manually top balance again and all cells are equal voltage then the minimal balancing that the Seplos can do should be enough to keep it balanced, but maybe that's optimistic.

championc

Is the Seplos definitely integrated into the Sofar (i.e. is it setup as General Lithium) ?

And don't go back into configuration to check settings. You should go into the information menu screen 5 onwards

graememk

https://www.boards.ie/discussion/comment/120523130#Comment_120523130

15% capacity isn't 3.2 v it's more like 3v especially under any sort of load

_dof_

https://www.boards.ie/discussion/comment/120523181#Comment_120523181

Yep, it's configured as general lithium and is communicating correctly to the Seplos

_dof_

https://www.boards.ie/discussion/comment/120523328#Comment_120523328

Thanks, I was just using the voltages mentions at https://footprinthero.com/lifepo4-battery-voltage-charts.

Even worse then, it should only be reporting low voltage alarm at around 48V then, but it's recording 54V under no load.

When I was trying the commissioning discharge test and put it in auto mode, the grid draw at the time was only around 300W, so it should have only been discharging at that level so wouldn't expect a big voltage drop I guess, certainly not 6V?

_dof_

I'll try another discharge test later if I can, and monitor the battery voltage with the multimeter throughout

Sligobuck

A good insight into the very rare occasion a Lifepo4 battery goes bad.

https://youtu.be/-13tJHDfknI

unkel

https://www.boards.ie/discussion/comment/120524578#Comment_120524578

Those are EVE type cells. The CALB type cells that most of us are using (as per the OP) are far sturdier.

graememk

https://www.boards.ie/discussion/comment/120523169#Comment_120523169

The alarm has come from the seplos, but on the me3000 it won't discharge anymore until it charges the battery a bit first.

You'll need to be able to either log or watch the voltages to see what's actually happening(loose bolt? It can throw low/high cell voltage issues, has caught most of us out at some point)

randombar

Updates for those interested in my battery mishap:

Photo before I switched on the meter:

https://capture.dropbox.com/VMz4wggL9FcciELL

Next

https://capture.dropbox.com/h1SlKv1lL0GAlirZ

Next

https://capture.dropbox.com/k7sveIWGw4b9pFxc

Looks like it's nearly there, current has dropped way down so how do I know or what would people say the actual capacity is?

graememk

https://www.boards.ie/discussion/comment/120526035#Comment_120526035

Is the voltage reading on it really weird?

randombar

I think it shows low while draining. Maybe? I’ll check the voltage when it’s fully drained see what it says.

munsterfan2

https://www.boards.ie/discussion/comment/120526112#Comment_120526112

You dont want to discharge the cell below 2.5v generally. Check the voltage across the terminals with a multimeter

"Tip 1: Never over charge/discharge a cell!

The most common causes for premature failure of LiFePO4 cells are overcharging and over-discharging. Even a single occurrence can cause permanent damage to the cell, and such misuse voids the warranty. A Battery protection System is required to ensure it is not possible for any cell in your pack to go outside its nominal operating voltage range,

In the case of LiFePO4 chemistry, the absolute maximum is 4.2V per cell, though it is recommended that you charge to 3.5-3.6V per cell, there is less than 1% extra capacity between 3.5V and 4.2V.

Over charging causes heating within a cell and prolonged or extreme overcharging has the potential to cause a fire. EV Works Takes no responsibility for any damages caused as a result of a battery fire.

Over charging may occur as a result of.

1. Lack of a suitable battery protection system
2. Faulty of infective battery protection system
3. incorrect installation of the battery protection system

At the other end of the scale, over-discharging can also cause cell damage. The BMS must disconnect the load if any cells are approaching empty (less than 2.5V). Cells may suffer mild damage below 2.0V, but are usually recoverable. However, cells which get driven to negative voltages are damaged beyond recovery.

On 12v batteries the use of a low voltage cutoff takes the place of the BMS by preventing the overall battery voltage going under 11.5v no cell damage should occur. On the other end charging to no more than 14.2v no cell should be overcharged."

randombar

https://www.boards.ie/discussion/comment/120526460#Comment_120526460

How do you know then using a load tester when it's actually "empty", i.e. when to stop the test.

Plugged in the other one to test there, MM reading 5.3 when it starts running drops to 1.6

munsterfan2

The atorch tester I use has a setting for cut-off voltage which I would always set at 2.6v. 5.3v is definitely incorrect. ( Was that a typo for 3.5v ? ) See the capacity chart vs voltage below ( dont think this is necessarily 100% accurate but good enough )

randombar

https://www.boards.ie/discussion/comment/120526719#Comment_120526719

Nice one. Ya sorry getting all confused here. 3.5

Will sus out about the cutoff voltage.

unkel

In your third pic the voltage is showing as 0.67V? I know these cheap battery capacity testers under read the voltage (so under read the capacity) but did you check the actual voltage of the cell at that stage? What minimum setting did you program into the tester, zero volt? 😮 If a LiFePO4 cell goes below 2.5V it is getting damaged.

randombar

https://www.boards.ie/discussion/comment/120527179#Comment_120527179

Ya my issue is that I don't know what the true voltage is during a test. Drops way off as soon as I start the test.

This is also an expensive learning process for me I guess, not sure how to put the minimum setting in, also how does the device know what the true voltage is, does it stop every so often and check?

unkel

https://www.boards.ie/discussion/comment/120527242#Comment_120527242

No, it bases it on the voltage that you can see displayed. Which is indeed subject to a voltage drop as soon as you apply a 20A current. Will Prowse has a good video of how to program the damn thing

https://www.youtube.com/watch?v=3dmWqHR7b9w

randombar

So I’ve put in the voltage etc testing again there. Problem is voltage drops below 2.6 if I increase the current over 10a??

does that mean the battery is knackered?

graememk

https://www.boards.ie/discussion/comment/120528883#Comment_120528883

Voltage at the cell or the tester?

unkel

https://www.boards.ie/discussion/comment/120528883#Comment_120528883

If the cell is fully charged when you start testing and with just 10A (or even the max 20A the tester can do) it drops almost straight away to 2.6V, then yes the cell is knackered.

randombar

Reading 3.3 at the battery, tester saying 2.6

do I need thicker cables?

graememk

https://www.boards.ie/discussion/comment/120529123#Comment_120529123

Yes! Way thicker.

unkel

You measure 3.3V at the battery itself and 3.3V at the terminals of the tester at the same time as the tester shows 2.6V?

If that's the case, the tester is faulty. Nothing to do with your cables

spose

https://www.boards.ie/discussion/comment/120524354#Comment_120524354

I think you need to reboot the me3000 to clear that error. Connect it to a laptop and go in to the history log. Will see what cell alarmed low

randombar

So battery 3.3 other end of cable 2.6

what kind of cables do I need?

_dof_

https://www.boards.ie/discussion/comment/120530235#Comment_120530235

Do you mean to connect the Seplos to the laptop to read its history log or connect the me3000 to a laptop? Don't know how to connect laptop to the me3000 except using the logger wifi.

spose

To the seplos. Did you get a USB adapter with it?

unkel

https://www.boards.ie/discussion/comment/120530248#Comment_120530248

When there is no load?

randombar

https://www.boards.ie/discussion/comment/120530947#Comment_120530947

match up when there is no load, don't match when there is a load.

unkel

https://www.boards.ie/discussion/comment/120531009#Comment_120531009

Well try bigger cables for sure, but those flimsy terminals on the tester can't take that much of a cable, been there, done that. Try to fit as thick and short a cable as you can. Also before you test, make sure your cell is charged up to 3.65V and taking no more current.

graememk

I used 10mm2 when top balancing at 10amps and still noticed some voltage drop

bullit_dodger

You guys know this space better than I.

While I know that tester maxes out at 20amp, I think doing a 50+ amp test and looking at voltage drop would be more appropriate. Assuming a 3kw draw (ME3000 limit) and 16 cells, your looking at 190 watts per cell. At 3.4v you'll need to be drawing down ~55amps to get that kind of power. It's a fair real world loading that the cell would experience and have to take.

Still though if the 20amp draw shows up issues, then no hope for the larger 50amp load I guess.

Makes me want to play with this stuff :-)

graememk

https://www.boards.ie/discussion/comment/120531524#Comment_120531524

Yeah if the voltage was sagging that much at the cell it would be a big issue.

What's going on is that there is a (relatively) large voltage drop across the cables.

You would have the me3000 connected to the bank with 10 or 16mm2 cable.

The better testers have 2 sets of cables, one set to draw a load and the other just to measure the voltage of the cell itself.

graememk

Also wiring of the neeeeeeeey

https://youtu.be/__2Dgnas5mc

graememk

Talking about voltage drop

These are the 16 cells in my battery. (4p)

Can you see where I hit it with a 6kw load?, And when that load finishes and returns to 300-400w. Went from 85% to 35% ish.

Because I'm in a 4p at full load I'm hitting the cells with about a 30amp load

It's being charged at about 2kw now.

unkel

I'm 6P now, so about 800-900Ah usable given the CALBs aren't quite the full advertised capacity. While I can still see a voltage drop when people put a heavy load on (think kettle + oven + washing machine + rest of house), I no longer worry about it 😁

darraghsherwin

Has anybody else had issues with getting their batteries delivered from oye Store?

I ordered two batches of 16 cells 4 days apart, I had to collect the first batch from the DPD warehouse as they could not deliver them due to "lack of a manifest", according to DPD.

oye Store says about the other batch, "Hi friend, your parcel has been withdrawn. Please contact DPD to urge delivery. Local Contact number: +353 1 424 3918"

Has anyone else had this issue?

soram

Hi Guys,

   Firstly, I hope it is OK to post this here but please feel free to move. I thought it was a bit relevant to this discussion. I've been following the thread for many months but I thought I'd try out the LiFePOs on my camper first before jumping in to a full home install as I already have a 5kW Soluna.

So based on the recommendations I ordered 4 x 200Ah CALB cells from PWOD. I also got a JK active balancer BMS from elsewhere.

So the cells arrived a week or so ago. One had brown stains on the base and some dust still in the crevices. It measured a bit lower in voltage that the other three out o the box. I charged them anyway and left them a few days on the balancer. We used it in the camper but the BMS shut it down with nominally 40% left. Damn I thought so based on here I borrowed a bench power supply and charged all four individually up to 3.65V and 0 current. I then put a 10Amp load on and took the voltage readings from them all every hour. The BMS shut everything down again at 35% with three cells reading very close to 3.15V but the other was down to 2.5V. So unless I am doing something very silly the cell is, as I feared, not great. 

I've been on to PWOD all week and they reply promptly but their suggestion to me was a link to buy a new cell. As it is single it is almost double the price I paid.

I will be disputing the order but was wondering if anybody had any advice about how best to do this? What language to use or points to note. I will then have to make a decision on whether to buy a single one or maybe just sell the 3 good ones and buy 280Ah for a bigger battery elsewhere if there is a market for 3 cells here.

Apologies for coming in to the thread cold with this and as I said please feel free to move or delete as appropriate.

hexosan

Is there a current up to date shopping list for what’s needed for a battery set up.

Any recommendations (links) on batteries from Ali or nkon.

_dof_

https://www.boards.ie/discussion/comment/120530235#Comment_120530235

Thanks, was away for a while, but I've reconfigured the cells as parallel and top balanced them, while at the same time I had disconnected the seplos cables and measured the resistance from each pin of the white and black seplos connectors to the ring terminal of each one. Note, I had soldered on extensions to the wires and then crimped the ring terminal onto that wire, so I was measuring the resistance of the original wire, the solder connection, the extension wire and the crimp ring terminal.

All of the wires came in at around .3 to .4 ohms, except one which came in around .8 to .9 ohms. So, I resoldered and got it to measure .4 Ohms too.

I reconnected everything back to 16s2p again and turned on the seplos with the laptop connected, and cell 1 was still reporting lower voltage than the others, and when I turned on the Sofar to discharge, it caused shutdown of the seplos again and gave the Low Voltage alarm in the Sofar again.

I saved the History Log, so hopefully that can help to diagnose the problem, it's available at

https://docs.google.com/spreadsheets/d/1-CN2FddDeCHAB4xkOZihg-IWFEuegNRz/edit?usp=sharing&ouid=111517154881982763970&rtpof=true&sd=true

I'd appreciate any advice thanks.

spose

Pretty much identical failure I had. Always cell 1 or 2 dropping to 0 or 0.2v with no reason. Finally resolved by cleaning battery terminals and bus bars with wire brush drill bit and adding copper conductive paste to each when redoing the connections. Running no problem for months now