Dillenger 48V battery - tear down / repair

[bluess57]

Hey all,
Longtime lurker , first time poster...
TLDR; Dillenger batteries are chinese imported rubbish.

Bit of background:-
Bought a Dillenger 500w 48V 13Ah rear hub kit back in late 2019, fitted to a friday bike and ridden so far about 5-600km.
With covid lockdown it didn't get much use, and unfortunately I had the battery sitting idle for like 3 months , from a fully charged state.
Went to start using the ebike again, plug charger to the battery, check the status lights on battery... not good
2 faint colour, 1 green , 1 red from left to right (the last is a red led - photo doesn't show it)


put battery into bike, panel displaying 1 battery bar, quick ride and display panel quickly showing no battery.
Put battery onto charge and left for several hours... no improvement

Decided to open it up, it's out of warranty, and Dillenger advice is buy a new battery. - no thanks.
What I found / see does not impress. WTF - Rust over several areas of nickel strip and cell tops.
ATM I'm wondering if this was just some low grade nickel strip OR
who ever put this battery together spilt/dripped acid flux on the way to soldering points, wiped it up and threw in into the case.



Also IMO, the battery internal weather proofing is lacking.
DO NOT ride these batteries in the wet and I've never ridden this bike in the wet, bike is stored under cover, and I live in a dry environment area (approx 200km from coast/sea and 600m ASL)

This 48V battery has 18650 cells made by dynabat - MX18650-26P - 2600mah , looks to be your generic kinda chinese battery


I've put the multimeter across the cells and have found 5 batteries in parallel reading 0 volt. Already extracted these out.

Now the plan ATM is a budget fix-up. Replace those 5 cells with new 2600mah cells - looks like I can get my hands Sony/Murata VTC5A conveniently, and redo any of the nickel strip areas with corrosion.

BTW the BMS is

 

[amberwolf]

TLDR; Dillenger batteries are chinese imported rubbish.

This is unfortunately true for a vast amount of batteries out there. The lower the cost for the same capabilities, the less likely it is to be as capable as it claims, or to last as long as it should.

Paying more for a battery doesn't guarantee a better one, but it makes it more likely. When you have to buy a new one, I recommend vetting the seller, then the pack they sell that you're interested in. (most of them dont' even know anything about what they are selling, but will tell you anything you want to hear about the packs; they may also have identical ad copy to other pages about similar packs, or have a single ad that is for several different packs and/or has conflicting information on it, etc.

<snip>
Went to start using the ebike again, plug charger to the battery, check the status lights on battery... not good
2 faint colour, 1 green , 1 red from left to right (the last is a red led - photo doesn't show it)
dillengerbatterylights.jpg

put battery into bike, panel displaying 1 battery bar, quick ride and display panel quickly showing no battery.
Put battery onto charge and left for several hours... no improvement
<snip>
I've put the multimeter across the cells and have found 5 batteries in parallel reading 0 volt. Already extracted these out.

FWIW, with a cell group at zero volts, the BMS should have turned it's output *and* input completely off, and not allowed charge or discharge, and the battery meter should have read completely dead because of the turned-off output.

Since it didn't, it may mean that the BMS output FETs are failed (which most commonly fail shorted, so stuck on), so the BMS wouldn't be able to shut off discharge when cells run low or have some ohter problme during a ride, either, and you could end up with an even deader pack than before (or a fire hazard, since you won't know this has happened, as there won't be any indication of it without opening it and measuring cell voltages).

It probably *did* turn off the input, which is why the charger didn't charge that 0V group at all (which is good, because that can cause a fire; not a highly-likely event but it does happen to damaged and overdischarged cells, which is why a BMS doesn't allow charge after a cell has dropped below a certain point).


Regarding the 0V group, the two most common reasons this happens are a failed cell, which drains all the others in parallel with it, and a BMS balance channel that is stuck on, which continuously drains that group, all the time, during charge and discharge and just sitting there, at whatever rate the balancer is designed to work (usually less than 50mA).

So when you replace the cell group, and verify all cell groups in the pack are relatively balanced, you may want to let the pack sit, unused, for at least a day, and then recheck the voltage of that group every few hours. If it has dropped at all, keep checking it and if it keeps dropping, you probably have a bad BMS (stuck-on balancer).

To test if it's the balancer (a near-certainty), disconnect the BMS balance wire to that group (or the connector for all of them at the BMS), and keep measuring the group. If it doesn't drop any further, the BMS is failed and needs to be replaced. If it does keep dropping at the same rate, something is wrong with one of the cells in the group (not likely, since you'll be replacing them all with new ones).

 

[bluess57]

A bad BMS crossed my mind as well
Thanks for a good explanation on how to test the BMS
Might just drop $$$ and get a new BMS now as I imagine some time in the future I'll rebuild it with new cells
Ordered batteries today

 

[tomjasz]

Steel connections, not pure nickel. Nickel doesn't rust. Dillenger is shite!

 

[bluess57]

Took a peek inside the BMS, appears to have 4 power mosfets in TO-263 package, can't read the markings.
Multimeter in diode mode, they all read the same across G - S , so all seem to be in on state atm - fwiw.

 

[amberwolf]

Is that with the BMS disconnected entirely from the battery? If not, the reading may not be applicable due to other voltages present on the parts.

But if it is correct, then the BMS is stuck on in both charge and discharge modes; which doesn't sound like the previous symptoms (which soudned like the charge FET was still working but the discahrge was stuck on).

 

[bluess57]

My bad, didn't fully disconnect BMS