Seplos BMS

spose · 19-01-2023 11:15pm

If you go in to real-time record and turn that on it will give you a reading every few seconds and not just when there is a status change like in the log. Might help see if it is really a step change or not

noelocon · 20-01-2023 3:30pm

So I hooked up the mm when that battery was discharging at 99% and once again after the substantial SOC drop was recorded.

I checked the individual cell voltages and they all seem to be 3.28-3.31+

It seems strange that the BMS is reporting an SOC of 99.95 with a pack voltage of 53.01v and an SOC of 10.12 with a pack voltage of 54.08 volts.

Does this look like a BMS issue or is the pack stuffed ?

PXL_20230120_134728554.jpg

PXL_20230120_132442936.jpg

unkel · 20-01-2023 3:44pm

https://www.boards.ie/discussion/comment/120134502#Comment_120134502

Never believe SOC values from a BMS, they are mostly bullshit, unless the BMS has had a good chance to calibrate itself (and even then I would never trust those values)

BTW, how is the actual cell level voltages with your multimeter compared to what BMS reports?

noelocon · 20-01-2023 4:15pm

The BMS is around 0.02 volts ahead of the multimeter...on each cell

graememk · 20-01-2023 4:23pm

https://www.boards.ie/discussion/comment/120134795#Comment_120134795

Now do it under load.

noelocon · 20-01-2023 5:43pm

I started discharging 24.14 amps and the pack voltage was 52.18 9.96 SOC

The cell voltages according to the BMS were:

1 =>3128

2 => 3136 (*)

3 => 3134

4 = >3158

5 => 3117

6 => 3111

7 => 3185

8 => 3120

9 = >3126

10 =>3149 (*)

11 =>3103

12 =>3116

13 => 3108 (*)

14 => 3095

15 => 2982 (*)

16 => 3075

This only lasted for a short while as the inverter then moved to start charging the pack. (*) indicates marked cells from the short circuit.

graememk · 20-01-2023 6:12pm

https://www.boards.ie/discussion/comment/120135183#Comment_120135183

Yeah because if the sofar me3000 detects a very low SOC it will charge the pack no matter what.

What we are looking for does the seplos bms voltage readings tally with the multimeter readings from the cells directly. When under load. If it doesn't match theres a connection issue. If they do match. The cells are damaged.and it looks like across the board.

noelocon · 20-01-2023 6:47pm

I managed to get the realtime events as the charge/discharge/charge cycle happened. I don't understand how the pack can go from full to less that 10% and then full again in such a quick period of time. It seems to be happening quicker now.

If the cells are damaged is there anything I can try or is it just best to turf them out and order a new set.

Hisotry data 20230120_1729.xls

Grouped data 20230120_1727.xls

graememk · 20-01-2023 7:19pm

https://www.boards.ie/discussion/comment/120135457#Comment_120135457

Cells are full at 17:21:15, Starts discharging.

40 seconds later the cells are hitting 3.1v, Seplos throws a fit, and slashes the SOC. Discharge stops, Cycle repeats. A full battery should sit at 3.3/3.2 until nearly the end.

If you can get a multimeter onto one of them cells when it is hitting 3.1v (on the seplos), we can verify if the seplos is reading it correctly. (although it shouldnt be throwing a fit at 3.1 more like less than 3, 2.8ish v)

If the multi meter does tally, id say they they are done.

graememk · 20-01-2023 7:21pm

you can see the voltage bounce back quite quickly, Maybe loosen the voltages (ie let it go to 2.6v or something), Id say there is damage to the cells, and are likely to sag a good bit under load,

If they sag, but hold at 3.1v You might get away with it

noelocon · 20-01-2023 7:31pm

Cheers, I'll try that

Just to be sure do you know what parameter(s) to change on the seplos ? Something like min discharge?

graememk · 20-01-2023 8:23pm

https://www.boards.ie/discussion/comment/120135635#Comment_120135635

I don't know my way around the seplos @championc would you know?

unkel · 20-01-2023 8:42pm

Personally, I would disconnect the Seplos to rule it out as a cause of the problem. Your cells seem out of balance, so I would take the pack apart, parallel all cells hooked up to a charger providing 3.65V for as long as it takes for the amperage to go down to near zero. Then re-assemble the pack and connect it directly to the ME3000SP. Charge it up fully, then apply a load, if you can a constant load to make measuring easier. And watch the cell voltages and pack voltage drop. If the cells drop below 3V before you have at least taken about 6kWh or so out of the pack, then your cells are compromised.

I wouldn't presume the cells are bad, until I had proof like that to back it up. I certainly wouldn't take the word of a potentially dodgy BMS for it.

noelocon · 21-01-2023 3:04pm

I disconnected the seplos and started a 3000w discharge from 80% to 20% (according to the Sofar). This 60% drop took 1h 51mins.

I measured the battery voltage before starting and it was at 53.05v.

Once the discharge started the multi-meter measured pack voltages dropped to 47.68 and over the time increased slightly to 48.26 before finally stopping at 47.26 when the sofar detected the battery at 20%.

When the battery was discharging the cell voltages ranged from 2.96 to 3.05.

I don't know how to interpret the voltage values but the 60% SOC drop in 1h:51mins suggests to me that a 80% drop would take 2h:45m at full discharge rate.

Does this make sense or have I completely lost the plot ;-)

Screenshot from 2023-01-21 13-40-57.png

Screenshot from 2023-01-21 13-59-06.png

Screenshot from 2023-01-21 13-58-46.png

graememk · 21-01-2023 3:11pm

https://www.boards.ie/discussion/comment/120138580#Comment_120138580

So you got nearly 6kwh out of the battery (nearly 2 hrs at 3000w) ?

Did you set the SOC at 80% in the seplos?

It did Sag to 50v which would be lower than normal, Possibly some harm was done, But as long as you know that, its water under the bridge really.

3v/cell at 20% is close enough, Run with that for a while and see how you go.

championc · 21-01-2023 3:44pm

You could lower the monomer (single cell) and overall warning and stop values which might then stop the Seplos from alerting and recalculating

slave1 · 21-01-2023 3:49pm

You had to clean up the torched terminals, I wonder what quality contact you're making with the bus bars? It annoyed the fuk out of me with my terminals as they had stickers on them and they left a horrible residue that took ages to get 100% off.

unkel · 21-01-2023 5:02pm

You started at 53V, this is about 80% and you went down to 48V which is about 10%, so you used 70% of the battery

So you got nearly 6kWh out of this 70% of the battery. Your cells are perfectly fine and have pretty much full capacity as when you bought them (as I expected)

The problem is with the Seplos, maybe make some changes to the settings, let it charge and discharge the battery a good few times and see does it settle.

noelocon · 21-01-2023 6:25pm

Hopefully you're right. I still think there's a number of variables in play. The sofar is reporting a voltage value 2 volts above the seplos and multimeter. This stops it from charging higher than 80% as it thinks it's hitting the max charge voltage.

I'm going to leave it for tonight as it's beer o'clock and reset the sofar tomorrow.

My plan is to leave the sofar on normal battery settings and timed mode until seplos comes back from holidays and then get support from them.

Once again thanks to everyone for their help and encouragement.

unkel · 21-01-2023 6:59pm

@noelocon - "The sofar is reporting a voltage value 2 volts above the seplos and multimeter."

That doesn't make sense. Take a pic of the voltage the ME3000SP is showing and immediately after take a pic of the measurement of your multimeter at positive and negative poles of the battery itself (not at the Seplos) and show us the screenshots.

noelocon · 21-01-2023 11:26pm

PXL_20230121_170300006.jpg

PXL_20230121_170231832.jpg

Roughly around 1 min apart between pictures

championc · 22-01-2023 12:04am

And with a multimeter on the first and last cell ?

unkel · 22-01-2023 12:09am

https://www.boards.ie/discussion/comment/120140277#Comment_120140277

That's not a multimeter. Take a measurement with a multimeter at the positive and negative poles of the battery itself (not anywhere else, not on the Seplos) immediately after taking a picture of the Sofar.

noelocon · 22-01-2023 11:23am

it seems to be an inverter issue, when under-load the battery is reading 49.5v, the seplos is reporting 49.8v and the voltage into the inverter is 49.3v.

Inverter and battery are all read directly off the poles.

The inverter reports a voltage of 52.1V on it's information screen and via the stats it outputs.

unkel · 22-01-2023 12:26pm

Those readings of 49.5, 49.8 and 49.3V are only a few hundred millivolt apart. That's par for the course. Unlike the difference of several volts you claimed to see yesterday.

I'm not sure what you mean with the last sentence with the 52.1V

noelocon · 22-01-2023 12:57pm

Yea that's the issue the inverter is reporting 52.1 on the info screen even though it's only getting 49.3 into it.

unkel · 22-01-2023 1:00pm

https://www.boards.ie/discussion/comment/120141242#Comment_120141242

Can you take two pictures immediately after each other (preferably when the system isn't under much load) and post them here as we asked?

noelocon · 22-01-2023 1:31pm

oops sorry, here they are. The inverter in in waiting mode and here are the voltages.

PXL_20230122_122738441.jpg

PXL_20230122_122717106.jpg

unkel · 22-01-2023 2:09pm

Yeah only a few hundred millivolt between them. 50.9, 51.0 and 51.2V. That's fine. That's just the normal margin of error between different measuring tools.

Not sure why you keep saying there's several volt in it

championc · 22-01-2023 3:26pm

We know under load that the DC voltage will be pulled lower. Obviously, the higher the load, the lower it goes, but stop drawing and the voltage will go back up.

So the drop is proportionate to the load. Therefore, the Sofar is "estimating" what it feels is the real voltage.

So since you seem to fairly regularly pull 3kW, and the DC voltage dropping down so low, you'll need to set your Seplos below those levels to stop the Seplos from thinking the cells are empty.

Seplos BMS

Comments