7S4P Li-Ion battery failure

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
This battery, little used, mostly stored, refused to charge and had only about 5V output voltage. A replacement is very expensive, circa £400 so I decided to see what what was going on.
I dismatled the battery and checked the voltage of each group of four parallel cells. They were all between 3.4V and 3,7V so that seems OK. Does this mean that the problem is in the BMS?
It uses an SH367107X chip and six HY4306 N-channel MOSFETs. I cannot find an english datasheet for this chip so checking it is difficult/impossible.
Help please!

[Edit] The cells are 18650 marked as 2500 mAh
 
Last edited:

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
Thanks. I can see that Q2 and Q3 form a bidirectional switch but wahat is Q1 doing?
Would it be a good idea to use a single cell Li-Ion charger to charge each group of four parallel cells so they are balanced?
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
Here is 'english' translation of datasheet - it might help in conjunction with datasheet

Q1 is load detection.. determines if external load connected to pack.
OK, thanks.

I have made a little PCB to allow me to connect a single cell charger and a voltmeter to each set of parallel cells without risk of shorting anything. It has just started on the third set, only another four after that. I'll let you know what happens.
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
Nope, still not working.
I have now charged and discharged each group of four cells, measuring the capacity, and they are all similar.
I was going to replace the SH367107 but I can't find a supplier, not even China.
I have ordered a BMS (China) which I think will fit in the space and I will replace the BMS and see what happens then.
 

Martin_R

Joined Aug 28, 2019
137
If I understand correctly you can charge and test each group of 4 X 2500maH cells. Do they charge ok and discharge to their rated capacity, or near enough?
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
If I understand correctly you can charge and test each group of 4 X 2500maH cells. Do they charge ok and discharge to their rated capacity, or near enough?
Yes. I tested from fully charged down to 3.4V (I didn't want to go too low). The four parallel cells, each supposedly 2500mAh, gives around 7100mAh and all seven groups are very similar.
 

Thread Starter

AlbertHall

Joined Jun 4, 2014
12,345
So the cells seemed OK and therefore I suspected that the problem was the BMS. I could not find a source for the control chip on that board so I looked for a complete board which would fit in the battery box. I found this one which is marginal on the width but should just about fit. It arrived yesterday and I have connected it up temporarily and it looks good. It charged normally and I went for a trundle round the town on the scooter it powers and all was well. I put it on charge again and that was normal too.

I was a little concerned that this module has a metal heatsink which is going to be hidden inside the battery box but even after a good spell of max speed running I could feel no warmth in that heatsink so it would seem to be no problem putting it in the battery box.

So a £9.17 fix for a battery that would have cost £300 to replace. I call that a win!
 
Top