Xuba Dongguan LifePO4 low voltage disconnect value

[loverofpeace]

Hey guys, having some capacity issues with my custom 280ah lifepo4 Dongguan cells. Get about 220ah out of them currently with the xiaoxiang bms with a Low Voltage Disconnect set at 11.5V (2.875v per cell). When I contacted Dongguan they told me the correct LVD should be 10.0V (2.5v per cell). Given the Lifepo4 discharge curve I can’t see how I could possibly get another 20% out of these cells if I try that.

What is yours set at? Have any of you set it this low? Thanks.

 

[loverofpeace]

 

[Dzl]

Hey guys, having some capacity issues with my custom 280ah lifepo4 Dongguan cells. Get about 220ah out of them currently with the xiaoxiang bms with a Low Voltage Disconnect set at 11.5V (2.875v per cell). When I contacted Dongguan they told me the correct LVD should be 10.0V (2.5v per cell). Given the Lifepo4 discharge curve I can’t see how I could possibly get another 20% out of these cells if I try that.

What is yours set at? Have any of you set it this low? Thanks.

I agree, I don't think the difference between 2.875 and 2.5 would explain the 60Ah discrepancy (though maybe it could account for half of the issue).

Some initial questions that might help narrow things down:

1. Have you balanced and capacity tested the cells?
2. How are you defining a full charge?
3. When your BMS disconnects what are the cell voltages of the other cells in the pack.
4. What approximate temperature and C-rate are you discharging at?

 

[loverofpeace]

I agree, I don't think the difference between 2.875 and 2.5 would explain the 60Ah discrepancy (though maybe it could account for half of the issue).

Some initial questions that might help narrow things down:

1. Have you balanced and capacity tested the cells?
2. How are you defining a full charge?
3. When your BMS disconnects what are the cell voltages of the other cells in the pack.
4. What approximate temperature and C-rate are you discharging at?

1. Yes top balanced when I got them.
2. 100% on the bms so above 13.5v or so after settling back down following a charge. I have an aili two way battery monitor as well that shows that there should still be 26% charge left currently. Which is a bit concerning.
3. Haven’t checked what they were under when it disconnected as it happened overnight. Now two are at 2.95v, one at 2.93v and one at 3.0v.
4. C rate is very low avg is about 0.05 with a high of 0.1.Temp is 26C during the day.

 

[Steve_S]

Well what the vendor Dongguan said is not incorrect.

LFP 0% SOC = 2.50V and 100% = 3.65V and yes the BMS should force OFF at either of those voltages to keep the cells within their specs / safety margins. Now many of us tinker with the BMS settings and play for the margins giving us either 90% or even 80% of actual capacity to stretch the lifespan / cycles as much as possible. Really, 10 or even 20% isn't much in the grand scheme and you can dip into it if needed, unlike FLA for when you take them below 50% you are hurting them.

Now take 280AH: -20% = 224AH. -10% = 252AH and your getting 220AH. Your BMS setting "Designed Capacity" which you have set for 2800mAh (or 280AH) but that assumes 2.50V to 3.65V. That should be adjusted according to settings. * Any monitor that you add on like the Aili 2-Way also has to have the same numbers to match what your setting to. If you take a 280AH pack, derate it to 250AH, then your settings should reflect 250AH, not 280AH. BattleBorn 100AH Batteries are actually 110 to 115AH but they derate them even though they are matched sets to ensure 100% compliance to their stated output.

Technically, the BMS should not be doing cutting off as that is not its job unless it is to protect the cells/pack. It is the job of the SCC/Charger & Inverter to cutoff for Hi/Lo Voltage conditions. As for the Out of Temp conditions, the BMS certainly has to have that control for the cells & pack, and the SCC / Inverter systems should be able to deal with it too, unfortunately, not all are capable for various reasons. My own Midnite SCC & Samlex Inverter can't use their temp sensors for lithium system, they are for FLA & Temp Compensation only. The samlex Temp sensor can be used as a disconnect relay input though.
* BMS NOTE: Some BMS' take a few charge/discharge cycles to get prperly set internally, especially if they're reading Internal Resistances of cells.

Because we play with the BMS and fudging around with the voltage per cell settings, that will cause some issues on readouts & accuracy. This applies to many BMS'. Ideally, it's best to let the SCC & Inverter do it based on what you want. So if you want 2.85 for Lo & 3.55 for HI cutoffs then enter those values into the SCC & Inverter X volts per cell X number of cells.

You can charge your cells till you are blue in the face and amperage in is zero, the moment you stop, they will settle for the next hour or so (pending on cell size) and will more than likely end up around 3.55V per cell.

With some chemistries, it is easier to determine the difference in balance from voltages than others. With iron phosphate in particular, it is difficult to determine a difference in state of charge from voltage alone. The difference between 3.2 volts (open cell voltage / no load) and 3.35 volts (open cell voltage) is far more significant than the difference between 3.35 and 3.65v. 3.2 volts (open cell voltage) may equate to around 30-40% while 3.35 volts may equate closer to 80-85% state of charge. This sometimes causes confusion as to whether or not a battery pack is significantly out of balance or not.

source: https://www.orionbms.com/general/pre-balancing-cells/

There is also a lot more in other reference materials related to this entire issue of reading voltages & determining SOC and what your getting in real terms.

I hope some of this helps.
Steve

 

[Dzl]

1. Yes top balanced when I got them.
We can rule this out as a cause
But you didn't answer whether you did an initial capacity test after the balance, if so what was the observed capacity?

2. 100% on the bms so above 13.5v or so after settling back down following a charge. I have an aili two way battery monitor as well that shows that there should still be 26% charge left currently. Which is a bit concerning.

Let me reframe this question, how does the charge terminate, what signals to the charger (or BMS if that is how termination happens) end of charge? What voltage and what if any absorption time? 280Ah capacity rating is based on full charge/discharge based on standard conditions as defined by EVE (Charge to 3.65V disconnect when current tapers to X, Discharge to 2.5V, disconnect immediately). Tweaking these parameters will affect usable capacity, as will imbalances at the top or bottom (you top balanced so you are god good there, but at the bottom mismatches may contribute to reduced capacity).

3. Haven’t checked what they were under when it disconnected as it happened overnight. Now two are at 2.95v, one at 2.93v and one at 3.0v.
See the paragraph above


4. C rate is very low avg is about 0.05 with a high of 0.1.Temp is 26C during the day.
We can rule this out as a cause

 

[Steve_S]

Oops, I forgot... darn seniors moments...
Remember these battery cells WILL diverge regardless of what you do. 1mv per AH is not unusual at the top & bottom ends, so if teh BMS is too sensity it will trip and cutoff. These settings believe it or not, especially on commodity cells, needs to be "Loosy Goosy" and you can easily allow for 250 to 500 mv even because it will all settle fairly quickly.

 

[loverofpeace]

Thanks @Steve_S and @Dzl for taking the time to reply and help me. I appreciate the thorough responses.

To answer your questions: I did not do a capacity test initially as I trusted the merchant based on their good reviews here. Perhaps a mistake but until recently when I had my solar disconnected for a week I don’t really need anywhere near full capacity in even 3 days of use in my van (with no sun).

So if I am understanding this correctly if i reset my specs to 2.5v and 3.65v, fully charge the cells and then discharge them fully I should be near 280ah correct? (Barring any balance issues with the cells)

 

[Dzl]

Thanks @Steve_S and @Dzl for taking the time to reply. I appreciate the thorough responses.

To answer your questions: I did not do a capacity test initially as I trusted the merchant based on their good reviews here.

Its not just about trusting the merchant (though my perspective is none of the merchants--even the ones popular here with good track record--should be trusted to the point that you don't verify and test yourself, at the end of the day they are all grey market resellers selling discounted cells that they don't really test themselves beyond a few quick checks (and can't be expected to for the price) beyond a few quick checks)).

One of the primary purposes of capacity testing other than verifying you got what you paid for, is to determine and set a baseline for your specific cells/system, so you have a reference point going forward.

So if I am understanding this correctly if i reset my specs to 2.5v and 3.65v, fully charge the cells and then discharge them fully I should be near 280ah correct? (Barring any balance issues with the cells)

Correct-- I think. This would also serve as capacity test. Try to pay attention to differences in voltage near the bottom and the top.
For capacity testing I'm not exactly sure how/where people are cutting charging in practice. EVE defines cutoff as when current tapers to 0.05C (14A) at 3.65v. Many people capacity test immediately after the top balance so charging is a non-issue. But if not, I suspect most people either cut charging as soon as 3.65v is reached or let current taper at a slightly lower voltage. In either case, it will give you a decent estimate of your full capacity if you charge to 3.65 and discharge to 2.5 at <0.5C

 

[Steve_S]

EVE defines cutoff as when current tapers to 0.05C (14A) at 3.65v.

Quite correct DZL. There is a point lurking Right There in front of us though. Some SCC's & Inverters can read the current resistance and act on it for charge cutoff. For example a Midnite SCC with WizBangJr on the shunt can be programmed so that when it reads the resistance at a certain programmed amperage it cuts off. This is obviously better than just going by voltage alone as that can fool a system fairly easily. Some might even recomend the Amp read could be allowed to drop to 5A or lower, but of course this all depends on the charging device.

If cells aking 3.65V per cell at 20A they most certainly are not full yet, even though they hit 3.65V but when all they allow is say 2A or even 5A well they are pretty much full. The bugger in the soup, as that some cells will hit that before others and then the game begins with the BMS and Balancing (if that's there) and more. This is where perfectly matched & binned cells like those used in EV Packs (which are rigourously tested & verified into matched sets) shine bright.