Got a weird BAFANG 43V swing in battery with some sort of BMS like problem.

[NatetheGrate]

The battery is a BAFANG (made by PYTES) BT F07.450.C. 10.5AHr 43v Swing In Battery , although it reads often 48V on the outer slit contacts. It has Samsung cells. Got it used from a donation place, no key (picked), cleaned the battery contacts 5X. Still problem. (see below)
Link Cty e1.1

1. This battery has a problem. If I ride the bike (1-8mi) the bike tanks out, and the handlebar display blanks. It does not fail on 30% grades, but at sort of random times, like smooth pavement. But the on-battery display still shows 5 bars.

2. I work with lithium batteries. So I (grrrr) cut (very very carefully) this thermo welded guy open (most cases have screws) and started measuring voltages and the like. Samsung 18650 lavender cells ganged in plastic cages. The thing seems to charge up to 5 LED bars OK.

3. All cells (ganged) read about 4.1V in the pack with my voltmeter and on the information screen fully charged (when working). but when it tanks, the voltage on the black and red main wire/pins goes to some sort of low voltage ~12v drifting down value of leaking power gate FETs of the BMS of 14V-200mV.

4. So I pulled all the little connectors off the BMS. Waited, and then maybe after an hour, a mysterious 48V reappeared on the big red/black leads. Replugged. Closed the case, and hit start, BAFANG screen lit, and whammo, the bike runs. It can take 30% grades and no fail, load 3-4 level. I am off for about 4 miles with the bull before it dies. The battery indicator indicates 5 bars still. I wait about a minute restart and can see the BAFANG screen and ride home. The error code on the display seems to be zero.

Any idea what is happening? A cell tanking or a defective BMS pcb? Maybe it is a loose connection to the sensors on the battery, fuse, or thermal overload defective sensor?

5. How to pull the battery bank out to get at the BMS. It has all these little tangs, no bolts. Cut? This baby is really glued or something. How do I pull off the battery pack in plastic cage to evaluate the BMS. It sort of works? Suggested replacement BMS if not loose? What to look for.

Thanks,

Nate

 

[Apartunis]

It seems like you may have one punished cell group in the pack. Most cheap BMS runs off of just one or two cell groups and will squash that fraction of the pack when left unused for a long time. If all twelve cell groups read the same voltage on your multimeter then it is likely a hinky BMS.

 

[NatetheGrate]

All cells groups that I can measure (they are sort of welded in pairs of 6?) seem to read beautiful voltages at rest, and they are very well balanced, all reading 4.0V at the moment on the controller. For example I am at 4 blue bars, and I read d1-d12 on the Bafang display 3.8V all cells. Fully charged all d1-12 12 read 4.09V, and the same on the multimeter opened up.

The problem is this battery pack is really secured into the case. I cannot pull it out. I guess I have to cut these side tangs to just remove the pack out to get at the BMS?

One last thought. I have a Bafang CR S104.250.FC 3.0 controller in the frame. Could it malfunction the battery and cause this symptom?

 

[NatetheGrate]

For the next person: it was a real creepy PITA, opening this guy. I used a degluer, Goof Off (flammable) very very carefully on the edges to soften the glue, but maybe Goo Gone would be safer. Then using an insulated a tire iron, I was able with great effort to remove/break bonds of the battery pack to the case which has no protective sleeve. The 5 conn battery pack/BMS was glued into the case. The Samsung? (lavender) battery pack outputs appears to be soldered by side tangs (white glue) to 12 parts of the BMS circuit board, , so there is no ribbon cable. BF-CAN ready says. The 5 blue level LEDs (still 5 bars) are on the BMS PCB- the ribbon cable is 2 wires to the battery level button on the outside. The BMS is epoxy coated. There may be a white fuse? 500$ for a new one. A little back up battery?. A tiny LED blinks periodically. Part: BF-12S F07 & F08CU T1
Q1. What can I check/replace on the BMS if it goes bad after about 100ft, but battery levels happy?. FUSE? Nothing looks cooked to the eye on first impressions.
Q2. How do the tangs match a typical battery BMS diagram?
Q3. What would be a good replacement BMS to use given these tangs? I guess I solder the wiring harness to these. Link please?

 

[fechter]

My guess would be one of the cell groups broke a parallel connection and now behaves like a lower capacity cell.
If you run it to cutoff, then measure the cell voltages, you should be able to determine which one it is. I've seen something like this when a bad spot weld broke off some cells in a pack.

 

[NatetheGrate]

OK, I put it back together (done this a couple times), carefully seating connectors.

Couple things.
Start 1: I noticed in Information starting up it did not display the version # of the battery. Tried to use the Bafang app but nothing talking that I could find like with Bafang in the bluetooth. No failure though as I did not ride it yet.
Start 2. Restarted: I went 3 miles to the grocery store on it, running it up to 450W, on the indicator, .... nothing happened. Display Information now shows battery version #. Voltages at end on the cells via indicator are beautiful..all D1-12 =4V, started D1-12 =4.1V.
I hate intermittent bugs.

 

[NatetheGrate]

20 more miles, no dropouts, and all is well? Battery is happy. Now to really seal it back up in a way I can get at it if needed.

Still not sure the cause...oxide on middle connector to power controller? Or serial communication?
Not sure if CANBUS type. Battery is stated inside CANBUS compatible. I did not fiddle with switches, as it appeared have programming.

Thanks for your thoughts on the problem.

Nate

 

[Apartunis]

If it is CANBUS then that is a serious problem even when it works.