Li-ion Cells Salvage

Joshua R. T. Purba
Electronics

Last year I was given a dead battery pack of a laptop. I finally got around salvaging the cells in it.

The battery pack is model AL14A32, rated: 11.1V, typical 5000mAh / 56Wh, minimum 4700mAh / 52Wh. Manufacture date 2014.09. IIRC this was an original battery pack from an Acer laptop (forgot which model). This is old style laptop battery pack, with 18650 li-ion cells inside.

Typical laptop battery pack construction from its era.

Outside-facing part of the battery pack.

The inside facing part of the pack has the label with model number, rating, production date, compliance information, etc.

Inside-facing part of the battery pack.

The model number and ratings are clearly visible, along with safety warnings.

Close-up of the label.

When measured from the pack’s connector with a multimeter, it read 0V. Obviously the internal protection — probably the UVP (under-voltage protection) — was still doing its job.

To proceed I dismantled the pack’s external plastic shell. Inside the 18650 li-ion cells can be seen in 3S2P configuration. Measuring directly at the 3S2P terminals read 2.965V.

Purple-pink 18650 li-ion cells in 3S2P configuration still connected to its control/protection circuit. The broken pieces of battery pack's shell can be seen in front of it. Behind it was the lineman's pliers and wirecutter pliers I used to dismantle the shell.

The guts of the battery pack.

Past the pack's connector the voltage was 0V, but when measured directly on the series terminals it was 2.965V. So the UVP was still functioning.

Series cells' voltage.

After I cut off the nickel strips connecting them I measured each cells invidually:

Cell Initial Voltage
A1 1.238V
A2 1.239V
B1 0.998V
B2 0.999V
C1 0.731V
C2 0.736V

No wonder the UVP was triggered: normal li-ion cells usually considered defective below 2.0V. And for this pack maybe the protection circuit will refuse to let it be charged altogether when the voltage drop below 2.0V? Anyhow the laptop refused to charge this pack, hence it was given to me.

In the pack A1 ∥ A2, B1 ∥ B2, C1 ∥ C2, that’s why their voltages are similar.

I put the cells in my LiitoKala Lii-500 charger to try to charge them. That failed. Lii-500 increased their voltage by around 300mV and then stopped. If I repeat the process the voltage increase a bit and then stopped again. Probably a safety feature in Lii-500?

Lii-500 refused to continue charging when the A2 cell voltage was only 2.34V. This was after I repeatedly restarted charging. Now I need a way to continue charging.

Lii-500 refused to continue charging.

Next I tried to charge them individualy using my bench power supply. I charged them to 3V with 10mA current limit. The purpose was to see if the cells can be brought to the normal spent li-ion voltage level, so I can later resume charging normally with Lii-500. That succeeded, each cells was brought to 3V, but immediately started dropping after disconnected from the bench power supply. So I hurriedly put the cells into Lii-500 and … success! The Lii-500 was able to charge them normally. The charge setting I used in Lii-500 was the lowest one: 300mA current limit.

I put a cell on a 18650 holder and connected that to the bench power supply. One by one I brought each cell to 3V with 10mA constant current charge. Though evidently my bench power supply was pushing 14mA despite the 10mA setting. After that the cells were able to be charged normally using Lii-500.

Use bench power supply to charge.

After several hours charging was complete and I pulled out the cells and let them rest.

The next day I checked their voltage and it sill read around 4.1V, not bad for cells that had previously fallen to < 2V level. Then I load tested them each twice. The test was done using Lii-500 with setting: NOR TEST 1000mA. This is Lii-500 “normal” testing, which is: charge again until full (1000mA limit), discharge (500mA limit) while measuring capacity, and charge until full again (1000mA limit). I let them rest for at least one hour between NOR TESTs.

Each cells were tested twice using the NOR TEST 1000mAh setting. This made the charger hot, which was expected considering it had to dissipate power between 5.6W to 8.4W when testing 4 cells at once. I made it stood on its side and used a computer fan to cool it down.

Using Lii-500 to load test the salvaged cells.

These were the measurements by Lii-500:

Cell Test #1 Test #2
A1 1393mAh 1391mAh
A2 1466mAh 1463mAh
B1 1244mAh 1251mAh
B2 1236mAh 1330mAh
C1 1472mAh 1477mAh
C2 1501mAh 1531mAh

This is low considering when the battery pack was new each cells would have 2350mAh to 2500mAh nominally. And since each cells had reached < 2V level, I would not use this for high-current application. Still useful for low-current application such as Arduino projects IMO.

Anyway I’ll let them rest again and check their voltage after several hours. If they’re > 4.1V I’ll consider them usable. Then I’ll discharge them to around 3.7V for long term storage as recommended by Battery University.

Note: The cells doesn’t identify its manufacturer or model number, but each has this marking: SI4TA2C. Searching on Second Life Storage yield no result.

See you later folks!

  • 2023-01-22 Update: After the second NOR TEST ended I let the cells rest for 9 hours. Here’s the voltage of each cells now:

    Cell Voltage
    A1 4.168V
    A2 4.172V
    B1 4.160V
    B2 4.162V
    C1 4.172V
    C2 4.173V

    All cells are > 4.1V, so I think they’re still usable. I’ll discharge them to 3.7V now as I wrote above.