Help with battery fault finding.

[bram.biesiekierski]

Hi.
I have a Dengfu e22 battery pack that has developed an issue. The battery is a 48v, 17.5ah, 840wh. 13s5p config. It has a "BestWay" or "BatWay" ??? 30a BMS. With LG INR18650 MJ1 cells.

The battery is not particularly old, and hasn't had more than about a dozen charge cycles. I use the supplied charger. I try and charge the battery the day before a ride. I am mindfull to not leave the battery plugged into the charger too long. I have never discharged the battery more than approximately 20% on the Bafang Display.

The symtoms I experienced are a reduction in felt motor power while riding for last 1 or 2 rides. Maybe a 5% reduction at first, then a 10% reduction the next time, where the main problem became apparent. The bike rode fine for the first few KMs (although with the slight reduction in the felt motor power output). As it approached approximately 50% discharge on the display (a little too soon aswell), the motor output became erratic. The battery percentage on display then rapidly fell down to very low. Like i could literally watch the percentage fall as i stoped to investigate what had happened. The motor and display shut down. This was mid ride so all I could do was try turning it on and off, And try removing and re-inserting the battery pack into the cradle. When I turned it back on, the display read back to approx 50%. When I tried riding, I would get about 5 seconds of good power output, then the motor and display would shut down again. I have not charged the battery since this incident.

I thought it may have possibly been a high temp cut-off. As I was riding it pretty hard. The motor was hot to the touch. But that's pretty normal since I've had the bike. I let everything cool down while I was stuck on side of the trail, and it still had same problems.

When I got back to the car, I swapped in a 2nd battery pack, and the problem disappeared. So I believe the problem is in this battery pack, and not the bike/motor/display etc.

As some of you may know, I ride this bike pretty damn hard. And this time was no different.

I am now trying to diagnose the fault. I have removed the inner battery from the metal case, and inspected it. I can not see any obvious fault, burns/melts, or other damage. I have not un-wrapped and completely dissasembled the battery yet though, so can only see what's visible form the outside.

I have also measured the cell voltages, by unplugging the BMS balance plug and performing measurements there. They are all exactly 3.72v each. (See attached test results)

When I measure voltage across the 4 main battery pack mounting system terminals, I only measure 42v, And 21.8v across the 2 small terminals. (See pic) Although, I have not got it plugged in, or connected to the power switch in the bike frame.

I suspect perhaps the BMS maybe faulty. I was hoping some of you guys with more battery knowledge could help me confirm? Or suggest what to test next etc.

Thanks.

 

[bram.biesiekierski]

On the back of the BMS PCB, there are 2 small wire connections labelled as "SW". These go to the 2 small circular terminals in the mounting system connector. I believe these then go to the power switch on the side of the bike frame.

When I jump these 2 wires together, the voltage at the maim power connection in the mounting system go up to 48.3 volts.

 

[bram.biesiekierski]

I am going to temporarily wire the SW terminals to on, and do a test ride. To try and rule out a problem with the main bike on/off switch, and associated wiring and connections.

I am doing this, as on my good battery, I don't have to even turn the main power switch on to power the bike up. So I presume, it must be hard-wired inside to bypass the on/off switch circuit.

 

[patdam]

All your deductions are correct.

The 42V on main conectors is the remaining tension (condensator ?) when the power switch on frame is off

The 2 little connectors '21.8V) are for the frame power switch

and true as that you said, if you link the 21.8V contact directely on battery, you will have permanently 48V on main contacts. BootsDave had allready this problem and it solve it with the same solution

 

[bram.biesiekierski]

All your deductions are correct.

The 42V on main conectors is the remaining tension (condensator ?) when the power switch on frame is off

The 2 little connectors '21.8V) are for the frame power switch

and true as that you said, if you link the 21.8V contact directely on battery, you will have permanently 48V on main contacts. BootsDave had allready this problem and it solve it with the same solution

So did BootsDave just join the "SW" terminals together? Or did he join them to the 48v power wire?

Just trying to confirm if this is the correct way to power the BMS up.

Is there any issues with bypassing the On/Off switch?

 

[Waynemarlow]

You will need to “load “ the battery some how whilst testing the cell groups voltage, there’s one or more groups that has a low capacity compared to the others. You could try measuring the groups after discharging the battery as fully as you can but I would suspect that if say three cells are OK, then they will mask the other two. But at a deep discharge you should still see a different voltage between groups.

The other concern would be the actual battery welds, you can get perfect looking welds on the outside, but with little or no contact with the cell below. Under high load that small contact heats up and a high resistance means a low voltage on the connecting bar. The BMS then reads a low voltage on that group. As soon as the bar cools you have a normal voltage. Thats pretty hard to fault find unless you can load the battery.

You can build a tester relatively cheaply using large resistors that are built just for this purpose, fully thermally connected to a large piece of Ali that you can direct a fan on. It’s going to need to dissipate a lot of heat and take probably 800W’s so don’t go there unless you think are capable.

,

 

[bram.biesiekierski]

Making a big resistor should be no trouble for me. I am a heavy duty auto electrician. I jut have not ever had anything to do with lithium battery's.

I've actually got a carbon pile load test box at my work. I reckon I could probably make that work, or make some other device to load the battery.

It should be easy to hook the battery up to a load, and back probe the BMS balance leads to test voltage. I guess that would be a good test next. I haven't had a chance to test the battery after bypassing the in/off switch yet.

 

[patdam]

So did BootsDave just join the "SW" terminals together? Or did he join them to the 48v power wire?

Just trying to confirm if this is the correct way to power the BMS up.

Is there any issues with bypassing the On/Off switch?

i think that bootsdave will be explain to you when it's reading the post, i think also it have explain that somewhere in M600/500 subject.

About on/off switch shunt. The generic batteries purchased have not this external switch, i using one in my lightcarbon frame and i haven't any problem at this time

 

[vext]

I looks like a lot of dielectric grease on the terminals. Am I right?

I say that only wondering if you're actually getting intermittent connection due to the grease..?

I've read that you don't want any dielectric grease on the actual terminals, only around them. I could be wrong.

 

[bram.biesiekierski]

I looks like a lot of dielectric grease on the terminals. Am I right?

I say that only wondering if you're actually getting intermittent connection due to the grease..?

I've read that you don't want any dielectric grease on the actual terminals, only around them. I could be wrong.

Its never been an issue for me. I put senson grease in all the connectors from day 1.

 

[bram.biesiekierski]

A little follow up with the battery.

So I managed to go for a ride today, and I tried using the battery again. I have not charged it since the issues started. The battery voltage has not fallen since my testing above. And it had been a couple of weeks. (I was crook with covid and had other mechanical issues with the bike to resolve.)

The bike powered up, and showed approximately 50% charge on the display. The bike felt good, and didnt shut down at all. Although i dont think it was as powerfull as it should be. But it was half discharged aswell. The power quickly fell off to approx 20% on the display after some full throttle on a long uphill fire road. It seamed to fall off quite quick considering it was at 50% on the display when I started out. I'm not sure if this is to be expected, or indicates a further issue.

But the bike didn't shut down and such like it was before. So I don't know.

Im going to recharge the battery, and try again and see what happenes from a full charge.