Kids 500w fun bike keeps cutting out

[OcTag3n]

Hello, my son has a mini 500w dirt bike that has a problem of cutting out mid ride, usually after he has given it the beans it will decide to shut down. No lights, nothing. I can slide the battery out & back in again, still nothing. I can only get it to work again if I plug the bike into the mains charger for a second & that seems to bring it back to life again.

Any ideas what it might be?

Thanks in advance!

[Manbeer]

sounds like low-voltage cut out, chances are the battery is sagging when he's getting on it pretty hard And the BMS is cutting power

[OcTag3n]

New battery it is then.

At what voltage is it likely to be programmed to cut off at?

[Manbeer]

not knowing the brand i would guess 24, or maybe 36v. my theory would only apply to batteries that actually have a bms, so if its one with a sealed lead acid battery it could be another story...that being said, even if it is it is possible it has some type of lvc in the controller or something. As long as the battery isnt crazy expensive, that's where I would be starting

Next time it dies, unplug the battery and test the leads where they connect to the bike - If it's outputting Any normal-ish voltage then you'll know the problem is either on the other end of things (controller, wiring etc) Or just that the low-voltage threshold is above with the battery has left

[OcTag3n]

Thank you for your help mate, I do appreciate it.

So I opened the battery up. It was put together well to be fair.

Inside I found pink 18650 cells, no writing on the cells but a BMS is installed.

I tested the voltage at the time that it shut down, got 35v

It tested 35v all the way up to the controller, from there the voltage dropped to 3.8v. I left the multimeter where I had 3.8v reading & plugged the mains charger power lead in & the voltage bumped up to 40v, I unplugged the mains lead, 35v reading now from where it read 3.8v before I plugged the power lead in.

Bike works again.

So the problem is with the controller?

[Manbeer]

just speculation, but it doesn't sound as if the controller is bad, more like the battery is sagging hard enough under load to trigger the LVC on the controller. without knowing what the cutoff voltage is set to it could just be assumed that it is higher than whatever bms is on the pack. assuming its probably somewhere around 31v, it would be very feasible the pack drops to that under load

if you unplug the battery when it is dead, then plug it back in would it turn on? curious if a disconnect from power is needed to reset it.

any pics or links to the bike?

[OcTag3n]

No, pulling the battery & reconnecting it does nothing. The only way I can get it working again is to plug the mains charging lead in.

It's an old bike, the date on the battery is 17.08

The model is a Thumpstar Juice.

[e-beach]

Leave the battery to charge over night and then ride the bike again. Meter the battery voltage before you ride.

The point of this is to be sure all your cells are balanced and that one or two aren't triggering a low voltage cut off.

You can also find the balance leads of the bms and meter those to find the actual voltages of each cell group. Just be careful not to short the leads as they are small.

By charging it over night and then metering the bms leads and the full pack voltage, then repeating the process after the bike stops running again you can get a very good idea if the battery is in need of replacement.

[OcTag3n]

I'll give it a go, certainly worth a shot.

Thanks

[fechter]

Yes it sounds like maybe one of the cells in the battery is weak and triggering the BMS to cut off. Not sure how the battery is assembled, but next time it cuts off, you can try to measure each individual cell group on the battery and look for one that is low. The balance wires going to the BMS can be used to measure if you have a skinny enough probe, or if the cell connections are accessible you can go right to those.

Sounds like you may need a new battery.

[OcTag3n]

Okay, so I left the battery on charge over night. These are the voltage readings this morning...

Bank 1 : 4.21
Bank 2 : 4.26
Bank 3 : 4.17
Bank 4 : 4.16
Bank 5 : 4.21
Bank 6 : 4.11
Bank 7 : 4.20
Bank 8 :4.06
Bank 9 : 4.09
Bank 10 : 4.05
Total voltage : 41.56

[LeftieBiker]

Looks like the 4 lowest cells have to go. Maybe the 6 lowest reading. A better test, though, is to run the pack low or to cutoff, then test the voltage. That will reveal the weakest cells more accurately.

If those cells have really been in there since 2008 (which I seriously doubt) then they deserve an award for longevity.

[fechter]

If you can bring up the low cells, it might actually work OK. It's not uncommon for cells 1-3 being lower than the rest due to the BMS draining those more than the others during storage. Ideally use a bench power supply or a single cell charger and try to get them closer to the others.

[e-beach]

Looks like you might have weak cell groups. Leave on the charger for 24 hours to see of they fully balance. Then check them again and write down the results. Then, as suggested, run it to LVC and measure the cell groups again to compare the voltages of the cell groups when charged and at LVC. You will then be able to identify any weak cell groups.

[fechter]

Most chargers cut off way before balancing is finished. A bunch of shallow discharges followed by charges is about the only thing you can do with most. If you had a way to pump a constant 30mA or so into the pack and the balancing shunts will work full time.

[OcTag3n]

Massive thank you to everyone who was kind enough to give me their advice.

I'll try the options & hopefully rescue the battery

[fechter]

Another thing I've seen done is to hack off the end of a USB cable and use the 5v through a resistor to charge a single cell. Somewhere around a 10 ohm resistor that's physically large should work. The resistor will get pretty hot. You have to watch the voltage like a hawk to make sure it doesn't go over 4.2v. Placing a diode in series with the resistor would limit it to about 4.3v, which is higher than you want to go but not enough to ignite a cell.

[OcTag3n]

So my plan is that I want to learn how to build my own batteries.

I have ordered a pretty decent spot welder & everything else needed to build a battery.

I'm going to pull this 500w faulty battery apart, taking note of how it is all put together. I'll charge all the cells individually to the same voltage & reassemble it as it was before. It will be a good learning curve for me.

I'm going to hijack my own thread & ask any of the battery building wizards out there, how to connect either of the batteries below.
If one design is easier than the other, please tell me how to connect the easier version.

20S14P 72v 48ah using 280 high quality big branded cells (when I can find them, there's a worldwide shortage at the moment)

Thank you.

[OcTag3n]

Would doing it this way work or am I way off with my thinking?

I think I can get away with 20S16P

Black are negative
Red are positive
Current flow is green line

[OcTag3n]

This is what I am thinking about how to connect the cells together with nickel strips...



Can anybody see any potential problems with this?

[e-beach]

Seems plausible from where I sit. Make sure your nickle strips and other connecting wires can take the current.

[OcTag3n]

Thank you, at least now I know that I'm on the right path.

Somebody pointed out that 16P is a lot of current to flow through where I have only 3 cells connected to the next bank so I might have another move around of cells to try & eliminate any areas where this is a potential weak point.

Thank you for your advice, I really appreciate it.

[OcTag3n]

This is the version I will build...