Samsung robo vaccum battery/BMS issue (not e-bike!)

[max_m]

I'm hoping some of you know a bit about appliance batteries. My Samsung robot vacuum sat in a shed for a few weeks, including during a cold snap. When I brought it inside, it wouldn't turn on, and wouldn't charge from the docking station

Some observations. The cells have some charge (>3V) and are balanced, and I can charge them directly (bypassing the BMS).

Voltage on the main output pins on the BMS (P+, P-) is much lower than the battery pack, and doesn't accept charge. There are two additional pins: PD1, PD2. They appear to be pulled to ground. No idea what they do.

I assume the BMS has entered some sort of protective mode. I disconnected the cells briefly from the BMS to try to reset it, and that brought the BMS output voltage up, but as soon as I tried to use it in the vacuum, it shut down again.

Any ideas?

 

[amberwolf]

Some observations. The cells have some charge (>3V) and are balanced, and I can charge them directly (bypassing the BMS).

The BMS may have a fairly high cell-level LVC, that could be higher than the voltage one or more cells are at.

Do you know which specific cells it uses? (at least, which chemistry) This should help us find out what a safe LVC actually is for those cells, and what conditions they are designed to handle.

If it got cold enough to freeze any liquid in the cells (electrolyte, etc), and this caused physical damage (expansion, fracturing, etc) inside them (probalby not visible), they may not behave like they should under load (but could still look ok without a load). If they *are* damaged, then directly charging them could cause a fire, so be careful.

Voltage on the main output pins on the BMS (P+, P-) is much lower than the battery pack, and doesn't accept charge. There are two additional pins: PD1, PD2. They appear to be pulled to ground. No idea what they do.
Those are probably data pins from the BMS to either the charger or the device's control system. That lets the battery have actual capacity data, usage data, and/or problems, stored in an internal memory, and when asked for this info by the charger or device it can let them know if it's still healthy and safe to use or charge.

Some of them store this data permanently, so that if the BMS ever goes into a protect mode for certain reasons, it "bricks" the BMS so it cannot be charged or used anymore, to protect the user from a potential fire.


Since if you reset the BMS it does start to work, it's probably not that, but since it still shuts down under load, it sounds like the cells are no longer capable of handling the load of the system.

It is possible they're simply too discharged to do it.

If you reset the BMS the same way, and then try to charge it rather than use it, does it then begin to charge? Or does it give an error, or do nothing?

If it starts to charge, then you might stay with it and let it finish a charge cycle, and see how it performs.

If it does not, and gives an error, that may help determine if the pack is toast or if it is just bad cells that can be replaced.

If it does nothing at all, and shuts the pack down again, then the BMS is probably trying to protect you against dramatic failure modes.

Replacing all of the cells may fix this case. (it will probably fix all cases, because the BMS is probably fine, but needs testing to be sure).

 

[max_m]

Thanks. The BMS would not allow charging even with the cells at 3.3V. I tried putting voltage straight from the charger on the P+ and P- and nothing happened. I've brought the cells up a bit higher, but still need to disconnect the BMS again to see how it behaves (it involves unsoldering a connection, I don't want to do it too many times).

I can also try to load each cell group and see how it behaves. Or try with different cells. Unfortunately I don't have a spot welder.

These batteries are tricky to find here in Canada. If I knew the BMS was working (not bricked) I'd get someone local to rebuild with fresh cells.

 

[amberwolf]

Thanks. The BMS would not allow charging even with the cells at 3.3V.

It's still possible the BMS LVC is higher than that (not usually, but possible, since that's still nearly empty.

I tried putting voltage straight from the charger on the P+ and P- and nothing happened.

Depends on what the P+ and P- are. If those are the connections for output from the BMS, then the BMS has probably turned them off. If they are where the cells connect, it might be that the charger has to communicate with the BMS to turn on, etc., via the two data lines remarked upon before.

If the charger does have data lines from it to the charge port, it is even possible there is a bad connection on one or the other, or even just the regular + and - pins, so the charger doesn't get what it needs from the battery to startup, or the battery doesnt' get anything from the charger.

I can also try to load each cell group and see how it behaves.

That's probably the simplest way to test for saggy groups. Shouldn't require disconnecting anything.

Keep in mind that at only 3.3v, assuming 3.7v nominal cells, the cells are nearly empty, so they will sag a lot, which is probably why the system shuts off when you reset the BMS and try to use it.

You can try to bring up the cells more first, by charging slowly with a CC-mode adjustable-current power supply or charger, at 100mA to 250mA to be on the safe side (in case the cells *are* damaged, it is less likely to cause any kind of dramatic failure).

 

[max_m]

Charged cells up to 3.6V and reconnected to the BMS. Still no voltage on the BMS output. Parking this project for now. Thanks for the help.