They are all above 3.4, very nearly above 3.5 so that is pretty reasonable. When you top balance you spend a long time in the last tapering charge and saturate the cells. Probably when charging if you tapered the current and voltage during the last hour you could get that extra 1 or 2%, but don't worry, looks pretty good to me.
Lishen 272 Ah Top Balance and Capacity Test
- Thread starter Inq720
- Start date
Inq720
Odysseus, expert on the Siren's call
I think I agree with you on the differences being unreasonable. It seems when most people show their results on this forum, they're touting differences down in the thousandths of a volt. However...
I followed the instructions from Nordkyn design especially the excerpt:
- Don’t waste time leaving cells sitting around connected in parallel. They don’t balance or equalise unless they are being charged in parallel at the end. I have specifically tested that. There is not enough voltage difference to keep driving current between cells until they balance out.
I was not diligent about this other excerpt:
- Bring all the cells up to 3.60-3.65V, disconnect and keep recharging this way until they all hold above 3.50V for at least 30 minutes. They are then full and balanced. Don’t leave the circuit closed and unattended under any circumstances; it would very quickly destroy all the cells.
I will say that after I do a top balance, cells will stay above 3.5v for at least 48 hours if not longer. All of your cells were well into the knee when high cell voltage disconnect was triggered, so I don't see much of a problem. I usually just charge each cell to 3.65v using my Riden supply, which stops automatically when amps in drop below .1 amps. So my cells are pretty saturated without going over 3.65v.
Inq720
Odysseus, expert on the Siren's call
I am not an expert in the field, but would consider Nordkyn to be. He says 30 minutes is sufficient and anything more is degrading the batteries. This is also consistent with of most Lithium based thinking in that cycling from 10-90 or even 20-80% SOC and keeping out of the upper knee as much as possible greatly increases cycle life. It is needed to get a balance state, but otherwise, its detrimental to leave or cycle into this area. This is how Tesla is getting such great longevity and are now claiming a million mile battery. This considering car usage is far more harsh than off-grid use typically and certainly more harsh than I'll be using them on the boat.
In my first top-balance, I only let the bench charger get down to about ~0.3 amps (it bounces around a lot) before removing, assembly of the battery/BMS and start the capacity test and the disparate cell voltages was the result shown in Post 17. With this in mind, and following Nordkyn's procedure, I did the following in my second top-balance cycle.
277.4 amp-hours
3543.3 watt-hours
(See yellow for cell results)
Following capacity test (See green above),
In my first top-balance, I only let the bench charger get down to about ~0.3 amps (it bounces around a lot) before removing, assembly of the battery/BMS and start the capacity test and the disparate cell voltages was the result shown in Post 17. With this in mind, and following Nordkyn's procedure, I did the following in my second top-balance cycle.
- Take the BMS off and put back in parallel.
- Set the bench charger to 3.5v and let them charge until the amps were around hovering below 0.1 amps.
- Set the bench charger to 3.65v and again let the charger continue till it was below 0.1 amps.
- Nordkyn's instructions then were to let them rest for thirty minutes. If their rest voltage is above 3.5v, they're good to go. Otherwise, re-do #3. After 30 minutes, mine settled to 3.61 to 3.63 volts. I'm assuming this being much better than the 3.5v Nordkyn suggests because the cells are new and A-quality.
- Assembled the battery/BMS and started a 0.2C discharge (54.4 amps).
277.4 amp-hours
3543.3 watt-hours
(See yellow for cell results)
Following capacity test (See green above),
- Fully recharge via BMS and bench charger till shutoff.
- Rested
- Did several high-current tests up to 130 amps. Although immediately after a test, the voltages varied by tenths of volts, after resting, they realigned.
Inq720
Odysseus, expert on the Siren's call
As I understand it, the test should be at 0.2C => 54.4 amps. My results were based on 0.2C.
But thanks... it's good to know, that it might be that high running at the lower loads. I expect my normal usage will be less than 15 amps.
But thanks... it's good to know, that it might be that high running at the lower loads. I expect my normal usage will be less than 15 amps.
I ran the same test against the EVE 280's and got 268 to 281 (mostly 270's). The Lishen's are kicking ass.
The CBA can do 40 amps, although I'm not sure I have the wiring for it. I'll try a higher rate at some point.
Inq720
Odysseus, expert on the Siren's call
I rigged my test like Will did in his video and I had an AC variable speed control that I could dial-in to get the 54 amps between the battery and the Inverter. I don't image everyone just happens to have one of those lying around. I only did the two runs above, but they were only 0.7% different results... like you, I'm real happy with the results. Just hope they live the hype of 3000 at 0.2C full discharge and the 10000 cycles at the low rates that I'll actually be using it.
I have not run individual cell tests on my 16 272Ah lishen's, but have run tests on all 4 of my 12v packs at a variety of discharge rates. The worst result I've seen is 276, starting from a top balance of 2.65v and discharging at 100a+. If I dial back the discharge to ~20a as the pack drops below 12v, the cell voltages surge up and I am able to get 280+ out of them without even waiting for low voltage cutoff. I can't say if a lower discharge rate for the entire cycle would add significant capacity, but lower discharge certainly extends the capacity near the end because the lowers amps seems to keep runners more in check.
Once I ran an initial test on each pack starting with a top balance, I proceeded to cycle each battery several more times at "real world" charge and discharge rates (40-50a). I charged them to 14.1v and discharged to ~12v. They are all hitting 272Ah capacity between 11.9v and 12v and I'm really pleased with that. I just don't see any reason to charge them higher or drain them lower for another ~10-15Ah.
One thing I noticed on the Lishen spec sheet is that their capacity test is defined at 1C discharge (if I'm reading it right). I don't have any way to run a 1C test with my BMS or my individual cell tester, but a 1C test is going to yield a lower capacity vs. testing these at ~.2C or even my initial test at ~.4C. It's not surprising that these Lishen's seem to be testing "better" than their rating when their rating is based on a 1C discharge test and most folks are testing at a much lower discharge rate. We're basically cheating on the test, but obviously the capacity is real if you are comparing to an EVE cell using the same testing approach.
I tested extensively from 10 to 40 amps and the results at those draws are identical. I now can test individual cells at up to 160 amps but haven't done so yet.
Yes, my Eve cells also test like that, the Lishen are obviously a better buy.
Eve specifies 1C as well. I suspect we'd get about the same at higher C rates -- but you need STRONG bus bars and clean connections for that kind of test to be accurate. None of my testing has shown that the capacity varies much by C rate (and I've gone to 220 amps on the EVE's).
What bus bars do you have? The no-insultation 20mm x 2mm rectangular (no rounded edges) ones like that ones I got via Basen?
I only see about 2Ah at the top of the knee but another 15 at the bottom of the knee. I had 273Ah before I went below 3v.
Makes sense. I didn't realize that EVE states 1C in their testing as well. It sounds like the lishens might have a more conservative rating or maybe these particular batches are just closer to spec.
I'm using the standard bus bars I got from Basen. I found them to be a little cupped, but I filed them a little and they seem to be working fine. I've got them doubled up and using noalox. My BMS's are limited to 120a, so I haven't run more than ~110a through them continuously. No heat build up on the buss bars at that current, just seeing the BMS itself getting warm.
I'm seeing the same thing you describe above and below the knees. Very little capacity difference between bringing the cells up to 3.6+ vs just bringing them to 3.525. On the bottom, I've been stopping the tests when I hit 272Ah. That's usually right at 12V or slightly below.
I think my tests with packs in series (vs. individual cells) is probably playing into my low end also. If I'm discharging the pack at 100a+, cell voltages start to diverge below 3v and once we get into 2.7v territory, something is going to quickly drop to 2.5v and cause shut down. As soon as I back it off to 20a discharge, voltages recover quite a bit and converge again and I get another 10+Ah at the lower discharge rate. You may not have that same phenomenon with individual cell testing. I don't know enough about how these cells interact to know whether a test on the pack is really a good measuring stick for the lowest cell or if other cells might be compensating a bit. I'd like to test all my individual cells, but everything seems to be working well and I hate to introduce the risk of damaging a cell through human error or a malfunction from my cheap tester.
DerpsyDoodler
Solar Addict
- Joined
- Jan 10, 2021
- Messages
- 2,247
For sure. just don’t lay them on their sides.... ?
MrM1
I'm Here, But I'm Not All There
Can someone connect on the included busbars and orientation? Will the same busbars u would use to parallel a group of cells in a row front to back work to connect in series off the side? Studying the spec sheet diagram it looks like if I come off the side I might need a longer busbar.
I want to build an 8s pack in this orientation (see attached), but I think the one off the side might have to be longer.
I want to build an 8s pack in this orientation (see attached), but I think the one off the side might have to be longer.
Attachments
Yes, you will need a longer busbar for that one connection.
Cell 2 results: 286.4 (tested 1 day after charging it to 3.65 so it had settled to ~3.45). The charge graph is hilariously the same as the first cell.
I have actually been tracking a lot of charge and discharge cycles, and except for some deviations due to temperature, 95 to 98%. That is hard to beat for efficiency of energy storage. That magic 98% is what got me interested in LiFePO4, fully 50% of my electric bill is 3 hours a day, 5 days a week. Just 3 hours of load shifting can save me $150/month in the summer, and the summers are quite long.
Anyway, I count amp hours in and amp hours out and it really is good.
Your updates have inspired me to go ahead and individually test my 16 lishens. My capacity tester will only run at ~6.6a, so it's gonna take a couple days for each cell, but it shouldn't take much actual effort. I started the first cell today.
I was a little nervous about messing up a cell by over charging or discharging (with cheap power supply and battery tester), but I've got my Victron system set up to send me an alert as I get close to top or bottom. I am using the "starter battery" feed for my Victron smart shunt and I set up alerts on min and max voltage of 2.7v and 3.5v. That should give me a head's up to watch the last bit of the charge and discharge processes to make sure nothing goes into the danger zones. My cerbo is uploading to VRM every minute, so I get alerts in real time (without depending on bluetooth) and I can keep an eye on voltage as much as I want through VRM.
Similar threads
