Who wants to take a peak inside the Samsung SDI battery with me? Well come on over!

[TheCartBoy]

1. Remove busbar cables and tray covers.
2. Remove the Torx screws holding the case together. Cut the warranty sticker ❤️
3. Remove the internal B+ and B- busbars and fuse. Disconnect the data connector from the PCB.
4. Remove the 4 screws that hold the bottom case to the cells.

[TheCartBoy]

5. Pull the cell pack out and look at the BMS board. The cell groups are soldered to the PCB, so replacing with a JK BMS or other BMS would be easy enough to solder the new sense wires to these points.
6. Bonus round, peel off the insulation sticker to see how the cell groups are assembled.

There are 13 cells in each group and 14 groups. So this 14s13p pack contains 182 INR18650-29E cells! Nice.

13 * 2.75A = 35.75 Amps continuous discharge.
13 * 8.25A = 107.25 Amps peak discharge.
13 * 2.850 Ah = 37.05 Ah

Specifications:
Model: INR18650-29E
Size: 18650
Style: Flat Top
Protected: No
Rechargeable: Yes
Nominal Capacity: 2850mAh
Continuous Discharge Rating: 2.75A (8.25A Non-continuous rating)
Nominal Voltage: 3.65V
Approximate Dimensions: 18.4mm x 65.0mm
Approximate Weight: Max 48.0g

[davros]

straight Lithium Ion cell? some references to "safer chemistry" on the intertubes

[Volt_Ampere]

Samsung makes very good 18650 cells. These look like those. I use lots of these in my model planes. Tesla batteries are built in a very similar manner with thousands of cells. My 48V Ebike battery uses these Samsung cells.

[TheCartBoy]

I think we all know where this is going....

1) Remove the screws holding the PCB down.
2) Cut the wires from the thermistor closest to the PCB.
3) Desolder the PCB from from the cell module.

So this PCB is not a complete BMS, its really more of a cell balancing board. It has some resistors to bleed down cells, but no MOSFETs to disconnect them. The brain chips are LTC6804G, which means it would use isoSPI interface for communications to the master module.

At 30-31°C the thermistors that came with the JK measure around 8 kOhms. The two thermistors factory installed in the cell module measured the same resistance. What a nice coincidence, they may be the same part, in which case I will run the factory ones to the JK.
I'm thinking it is this part: TE Model 44031 Precision Epoxy NTC Thermistor
Will take another data point at a different temperature and see if it lines up.

[chuckinnc]

Excellent this answers all my questions, Samsung INR18650-29e @ 14s13p
my packs read 56.7v which is what the elite manual says it should be, then 56.7v/14 would be 4.05v per cell, that's good that there not charging them to the maximum [4.2v], maybe that's why there is a 10 year warranty on these packs.

[TheCartBoy]

Originally Posted by chuckinnc

Excellent this answers all my questions, Samsung INR18650-29e @ 14s13p
my packs read 56.7v which is what the elite manual says it should be, then 56.7v/14 would be 4.05v per cell, that's good that there not charging them to the maximum [4.2v], maybe that's why there is a 10 year warranty on these packs.

Yes, it also leaves headroom for regen current, that's critical on RXV and can be a problem for aftermarket lithium swaps that don't retune the controller settings.

[TheCartBoy]

Soldered in the sense wires for the JK BMS, ran them out the hole previously used for the data connector. The thermistors were in fact the same, took another temperature measurement to confirm, but ultimately used the new ones that came with the JK. The factory ones were not buried deep into the pack, so I just pulled them out and put the new ones in the same places.

Configured the JK BMS as Li-ION, 14 cell, 37 Ah.
Most of the default settings in advanced would be fine, but I opted to adjust the following:

• Charge and Discharge over temperature protection at 65°C, as the label on the battery said not to use above 65°C.
• Continued discharge current at 36 Amps.
• SOC 0% at 3.00V per datasheet recommendation. This required setting the cell undervoltage protection release just above this, so 3.01V. Then set the undervoltage protection to 2.95V.

I would have liked if the BMS could have supported an adjustable time limited higher current draw feature, that way I could run the cells at the peak currents and if it went on too long the protection would kick in. But they don't have that, or its unclear to me how I would configure that.

Datasheet:
https://eu.nkon.nl/sk/k/29E.pdf

[TheCartBoy]

Did a complete charge discharge cycle. Brought it up to 4.2V / cell, 58.8V for the pack. And discharged down to 3.0V / cell, 42.0V for the pack.
Measured Ah was 34.7 Ah. A bit less than the calculated 37 Ah, but to get the full capacity described in the datasheet, the cells would need to be discharged down to 2.5V, which I don't plan on doing.

So for an advertised 665954 - CELL MODULE, 30AH, 50.4V, I am pretty happy to see more than the advertised capacity.

[TheCartBoy]

Originally Posted by TheCartBoy

I would have liked if the BMS could have supported an adjustable time limited higher current draw feature, that way I could run the cells at the peak currents and if it went on too long the protection would kick in. But they don't have that, or its unclear to me how I would configure that.

When I set the continued current limit to 36 amps, this means anytime I pull more than 36 amps, it goes into the time limited (user adjustable) over current protection. The default is 300 seconds allowed, then it will shut off and has a default 60 seconds cool down.

The allowed over current limit is not adjustable, its the hardware limit of the JK BMS. I purchased an 80 Amp continuous, 150 Amp peak BMS (BD6A17S8P). And during testing, it would appear the peak current limit, 150 Amps, is what would be allowed for 300 seconds.

So I oversized the BMS for this cell pack, and that's not a good thing. A 100 Amp peak BMS would be more suitable for these cells as the datasheet indicated 107.25 Amps peak discharge.

I will be replacing it with a lower rated unit for this reason and more.