battery disassembly day.
Still applying the famous Australian motto "Don't turn it on, take it apart"
Doing so mostly because I'm changing the BMS for a speedict neptune 15 (a very cool piece of kit
) but also to check the assembly for potential issues...
So this is the 48V 21Ah model, 13S6P with supposedly
LG MJ1 3.5Ah cells.
Casing is made by Reention (known for the "Dolphin" & "Whale" models), model is Dorado ID-something (their website
is unbearingly slow these days...)
first thing I'm worried with is that even though there are sealing lips provisioned on the plastic end caps, they are of no use with the aluminum body, so it's not great to prevent water ingress.
Is Reention using this part on other tubes with better design ? Or was the mold already done and they cheaped out on the body ? We'll never know but this sucks nonetheless.
If any serious rain riding is intended, adding some good electrical tape (like 3M 33
) on the outside will help a lot.
A better version could be designed and 3D printed too. I might look into that when i'll run out of spares and need to change the stock connector sets.
Amongst the wire bundle, the main discharge leads are retained by selastic, all the others are free moving. Not a major issue as most are connector terminated. But, the 48->5V USB converter input is soldered in and quite poorly: wires are thin and have a 90° bend at the joint, with vibrations and wire movement, they will
Things are pretty well insulated with all this cardboard around so a dead short is unlikely, but still. They could have secured it as they have done for the others surely. Or use a connector
Also, my negative converter input lead is damaged, was trapped in a screw post apparently, wtf
Anyways, I'm ditching this module so I'm not too worried, but you guys might want to check yours...
On the other end of the battery, we get the charge connector. Same remark about water ingress.
But here, the wires are held in place with selastic and use a connector for easy disassembly (more likely for faster production but hey ^^). Cool.
As a side note, the connector placement is quite strange on this battery.
The discharge port is located at the top of the downtube, whereas the charge one is at the bottom. This means the discharge cable runs the totality of the downtube to reach the motor. Voltage drop. Cable weight & cost. Poor design choices made by Reention there IMO.
Onto the BMS,
Looks quite decent. Controls charge and discharge on 1 lead. Measures 70 * 52mm. Around 8mm thick, but a much thicker one could easliy fit. Has a thermal cutoff sensor which states 90°C. PCBA is conformally coated, not perfectly but good enough. There are thermal pads on the alu spreader. Mosfets are SkySilicon, a Chinese brand, but with available specs
. I did not looked much more into it as I'm swapping in the Neptune...
Now, onto the cells.
It was a MASSIVE PIA to get the cell array out.
A jig is probably used at the factory to force push the bundle in the casing.
First I removed all the selastic on the BMS side of the battery, to allow it to slide out unobstructed.
Then I cut a piece of wood approximately battery shaped, onto which I sticked a thick dense foam pad. I did not want to hammer directly on the cells, even through this makeshift padded shim.
Instead, I sat the battery on a pvc pipe I had laying around and repeatedly hammered (with a rubber mallet) on a side corner of the alu casing. Effectively pushing the casing down on the battery.
I also roasted the battery in the sun for a few minutes to allow the casing to expand (the cells themselves, with their high thermal mass remained cool). That helped quite a bit. Once 1/3 of the battery was out, this went much quicker.
Overall it took 30min+ of hammering
And, oh surprise. No LG MJ1 cells as advertised. Instead, we get Samsung 35E. 78 of them. In plastic holders.
I mean, they are good cells but I was sold a MJ1 pack... I don't like that. Asked Ivy about it, waiting her response.
If a change was made in production they should have told us and updated their product page.
From the research I've done
, capacity is around the same but the 35E have a higher cutoff vs the other 3.5Ah cell (GA, MJ1) at 2.65V vs 2.5V
. --> Set the motor LVC accordingly
Voltage sag looks worse on the 35E and it seems they are quite bad in cycle life
if fully discharged (dead at ~300 Cycles @ 100% DoD). So avoid like the plage fully draining your batteries...
It also happens they are ~15% cheaper
than the MJ1...
Last, about all this green cardboard wraped on the cells. It's too much. This is what made the removal such a massive headache.
As can be seen on the below, the cardboard got scored by the screw holders within the tube, indicating again too much material. Plus if it's goal was to protect the battery it's moot as it's not covering the whole array. And there is some random white cardboard added, wtf
Anyways, I'm going to rewrap it properly and will secure it with EASILY removable shims
Overall, I would say the build of the array itself is OK. The spot welds appear strong enough (6 points per cell).