Help needed by New Entrant to eBiking

[PK969]

Hi all,  I am a new entrant into ebiking. 

Some time ago I purchased a second hand project bike, and am now having problems with power supply.  Can anyone help?  I have several issues.

Description of bike set up:  Golden motor front and rear drive motors, external controller (no markings but looks same shape as BAC-282), GM 48v12AH battery (LFP4812S), thumb throttle, cruise, regen. brake controls.  Charger is HP8204L3 - no output indicated, but measured at 54v (which is strange as most specs. say 36v for that model).

Initial use - the power kept cutting out after a few minutes of powered riding.  I suspected the battery, but then concluded that the power draw could be too much for either the controller, or battery, so removed the front GM drive, and replaced with non powered wheel.  I do not need wheel spin on both drives off road... This seemed better, and the power did not cut out. Used a few times like that, but was always concerned about battery life, as the LEDs on the thumb throttle did not work.  I live in a VERY hilly area - hence my purchase of an eBike.   

I have now lost all drive to the rear wheel motor.  Battery charger indicates fully charged, but battery output measured at 45v only.

First question: How do I identify what Magic Pie version I have - there are no markings on the wheel motors or controller?

Saw that the 3 pin Power lead connector to the rear GM drive motor had suffered excess heat on the neg pin. Continuity still present though.  Initial thoughts were I suspected the controller.  Does not beep when Hall sensor connector disconnected, or supply power to the motor connectors when the battery is on, and throttle pressed.  Checked regen. brake switches, they work.  Likewise cruise button.  Can't think of a way to check the throttle control, so bought a replacement with LED volt display - yet to fit.

Tried online to find UK GM agents for advice, but most not operating due to lock down, and some no longer agents.  One helpful email response, from one no longer an agent, suggested that the battery could be at fault, as 45v may be too low to activate the external controller.

Second question:  Does anyone know a UK current stockist for GM products?

Decided to investigate the battery to look for obvious problems.  Took off the aluminium cover and exposed the cells and B.M.S.  There are 39 unmarked orange cells, which each measure 26.15mm diameter, length approx 65mm.  Assume these are 26650 LiFePO4 cells?  Cells are linked in 3s parallel, and 13 groups in Series.  Assume cells are each 4AH capacity, hence 12AH battery. 

Third question:  If the battery has 39 cells, (13S3P=54v @ 12AH) why is it marked as LFP 4812S?  The B.M.S also has only 12 thin white leads to the PCB connector, not 13 or 14, as seen on advertised 48v 13S B.M.S's.  I can't even see a 12S B.M.S advertised apart from 36v?

Studying the B.M.S board, I noticed that one of the fuses/resistors/diodes? (i'm not an electrical engineer) had broken, and was not continuous.  It is flat, and marked 101.  See attached photo.  There are 13 pairs of these on this board, and underneath on the board is marked R37.  Possible cause for low battery output?  Reducing cells in effect to 36 (12S) and thus limiting charger input?  Cells currently measure 49.9v, but only 45v being output from the B.M.S.  Am I correct in assuming that each one is a fuse for that group of 3 cells?

Forth question:  Does anyone know if it is possible to purchase these components (I could solder it on) of if not, where to source a replacement B.M.S suitable for this battery.

As you can see from the above - I am fumbling my way into the subject.
Any help would be greatly appreciated.  Regards, PK

[Bikemad]

Hi PK andto the forum.

The battery you have is uses 4Ah 26650 Lithium Manganese cells not LiFePO₄.

Each group of three cells should be charged up to 4.2V where balancing takes place, hence the 54V measured charger output (13 x 4.2V = 54.6V).
LiFePO₄ only charge to 3.65V per cell.

This battery must be quite old as GM superseded them with the 48V10Ah LiFePO₄ packs over 8 years ago.

The newer 48V packs are labelled LFP-4810S-LT as shown here:

I suspect that the cells have deteriorated over this time (especially if they have been powering dual Pies) and this has caused the failure of the BMS resistor.

You could try replacing it with another 100 Ohm resistor but I don't know how successful that will be. 

I'm guessing it will be a 3 Watt resistor, but it might be smaller, so check the physical size and see if it matches the one in the link above.

As your motors both have external controllers I suspect they will be Magic Pie II's:



But if they don't have the offset spoke holes they could be original Magic Pies that have been modified to use external controllers.

Check the battery voltage when the throttle is applied to see if it drops noticeably (because the BMS is cutting the power).
If it doesn't drop, try the other controller to see if it is a controller problem. The external controllers should run on 24~48V and were often set on 24V even when used with a 48V pack as many users said it produced more torque that way. 

Alan
 

[PK969]

Alan,

Many thanks for your reply and help.

Both GM hub drives were wired into the one external controller, until I disconnected the front one. 

Do you have any idea why the GM battery is Labelled LFP 48V12S, if it is in fact Lithium Manganese cells wired as 13S?  Most confusing.  Why would a 13S battery have a B.M.S with 12 wires going to the PCB connector?  I had assumed that each 3P set of cells, would need a wire to the B.M.S to allow monitoring/control.     Also, if I replace the B.M.S (if my repair does not work), am I correct in thinking that B.M.S's are specific to cell type?  i.e. I could not use one specified for the more commonly used 13S3P B.M.S's for 18650 cells, without having to adjust the charge/control parameters?  Could I just set the max charge and cut off values to suit the 26650 cells? 

Regards,
PK

[Bikemad]

Both GM hub drives were wired into the one external controller, until I disconnected the front one.

That sounds like a recipe for disaster as brushless motors cannot run correctly from a single controller unless they are physically locked by a chain, belt or gear to maintain the exact same rotational position relative to each other. The timing of each motor must be controlled by a separate controller.

Is it possible that the front motor you removed had its own internal controller, or did it only have the same three thick Phase wires and five thin Hall sensor wires like the rear one?
Does the battery only connect to the single controller or did it connect to the front (or rear) motor as well?

Do you have any idea why the GM battery is Labelled LFP 48V12S, if it is in fact Lithium Manganese cells wired as 13S?  Most confusing.

I don't know why they use the LFP prefix, but the 12S signifies 12Ah not 12 series cells.

Why would a 13S battery have a B.M.S with 12 wires going to the PCB connector?  I had assumed that each 3P set of cells, would need a wire to the B.M.S to allow monitoring/control.

It only has 12 thin balance wires because it uses the thick battery + and Battery - wires for the other two connections required to balance all 13 cell groups:

Also, if I replace the B.M.S (if my repair does not work), am I correct in thinking that B.M.S's are specific to cell type?  i.e. I could not use one specified for the more commonly used 13S3P B.M.S's for 18650 cells, without having to adjust the charge/control parameters?  Could I just set the max charge and cut off values to suit the 26650 cells?

The BMS is specific to the cell chemistry as they use different charging and low cut off voltages. Some can be adjusted by the manufacture, but others cannot.
The number of series cells and the maximum and continuous charge/discharge current requirements will also dictate the choice of BMS.

Yes you can use the more commonly used 13S3P BMS for 3.7~3.8V 18650 cells.
The physical cells sizes are irrelevant as the BMS only monitors the voltage and current. The combined capacity of the paralleled cells (Ahs) may require different drain currents to be used to balance the cells.
A BMS designed for a 10Ah battery might struggle to properly balance a 100Ah battery as the drain current would probably be too low.

Hopefully some of this this makes sense.

Alan
 
EDIT: Balance lead diagram added

[PK969]

Alan,

Again thanks for clarifying those queries.  As a matter of interest, the front wheel motor has 2 heavy leads (Black and Red) in one connector, and the remaining 9 wires are varying colours, but very much smaller diameter, in another connector.  From the picture you attached, the motors look like the Magic Pie IIs.

Hopefully, with a replacement BMS board, I will get it back up and running again - although I expect the battery may need to be changed in the near future, if it is over 8 years old, and possibly heavily used with two 1000W motors.

Once again, thanks for your help.
Regards,
Peter

[Bikemad]

Hi Peter,

The front wheel definitely has a built in controller and the Black and Red leads would have been connected to the battery:



This explains why there was only one external controller. 

Check out this post to hopefully confirm which motor you have.

Alan
 

[PK969]

Alan,

I have been trying to source a suitable replacement BMS, and one supplier has asked what the cut out voltage would have been on these 26650 Lithium Manganese cells.  Your previous reply indicated that they would be charged to 4.2v, but do you have any idea what the low voltage cut out would have been set at?   If a newer 18650 cell 13S BMS is used, will the upper Charge and cut off voltages be different?

Regards,
Peter

[Bikemad]

The voltage range of your 26650 cells should be the same as most non-LiFePO₄18650 lithium cells (3~4.2V per cell).

With a 13S BMS I would expect the Low voltage cutoff to be 40~42V (3~3.25V per cell).

As long as the BMS is for a lithium Ion battery (not LiFePO₄) it should be ok.
The original BMS was probably rated for ~30A current.

Alan