Bike cutting out when battery on 3 bars

[Nealh]

Off to work so no time for a proper reply, will do so later or tomorrow.
Some one else may chip in with advice in the meanwhile.

All will say for now is you will have to reduce C5 value to reduce the current draw load.

 

[matthewslack]

Thanks!
Here they are
The white bit on the side is what a guy on youtube seemed to test, but it's glued to the rest and I don't want to break anything.

Also I have used the bike from fully charged a couple times since and noticed it breaks down once the voltage drops below 44.0V (the voltage parked could be over 50.0 but once the load increases to drive it drops. I managed the load output today to keep it over 44.0 at all times and it didn't crash. No idea what this could suggest (I put C12 down to 1 which apparently should only cut the battery out when it hits 39.0, unless the BMS overrides this)

First, be extremely careful not to short out any cells, or you might have explosion or fire to deal with. Make sure your meter probes are insulated except for the very tip.

Your pictures show where to measure very well.

You can see 13 rows of cells, and what you want to do is measure the voltage of the group of cells in each row. But it is easier to measure working on one side at a time. When you measure voltage from one of those linking contact pieces to the next one, you are measuring two cell groups in series. All of these measurements should be very similar if the battery is well balanced. If it is fully charged, then two cells will be about 8.34 to 8.40 volts. Measure them all, write them down, see if any are significantly different. If you can get at the other side of the battery, do the same that side.

Report back with your measurements.

If you measure after using the bike, and pushing it hard, forcing a shutdown, then you might find one or more weaker cell groups.

But remember, be very, very careful.

 

[Bikes4two]

Have a look at my posts to see the advice I received.
https://www.pedelecs.co.uk/forum/threads/intermittent-fault-on-battery-causing-loss-of-power.35875/

This is a bit of a wild (maybe dumb-ass) thought in that on reading the link posted by Trevor Brooker, and in post #3 at that link by @vfr400 , he outlines the complexities of sensing wire configurations and the number of possible combinations that there are with different BMS.

So here is the wild (maybe dumb-ass) thought - the bike bought by @bikemerchant was a DIY build - did it include a DIY build of the battery and if so, maybe the BMS in the battery hasn't been wired correctly?

As I said, a 'wild' thought - I've almost zero knowledge of how BMS work especially in the 'mechanism' for Low Voltage Cut-off, so please see this post as a possible red herring.

Anyone with BMS knowledge and a view as whether this could be the OP's problem here?

 

[soundwave]

This is a bit of a wild (maybe dumb-ass) thought in that on reading the link posted by Trevor Brooker, and in post #3 at that link by @vfr400 , he outlines the complexities of sensing wire configurations and the number of possible combinations that there are with different BMS.

So here is the wild (maybe dumb-ass) thought - the bike bought by @bikemerchant was a DIY build - did it include a DIY build of the battery and if so, maybe the BMS in the battery hasn't been wired correctly?

As I said, a 'wild' thought - I've almost zero knowledge of how BMS work especially in the 'mechanism' for Low Voltage Cut-off, so please see this post as a possible red herring.

Anyone with BMS knowledge and a view as whether this could be the OP's problem here?

the bms manages the charging and balancing of the cell groups if one group goes out of balance then the batt wont fully charge and take a hit when you try and pull full load under power and get voltage sag.

the batt looks ok from the pics as if it was wired wrong would not change at all as the bms would say no and cut power but one cell bank could be low and not charging to full power.

id say it is either a cell bank gone out of wack voltage wise the bms has gone bad or the batt has been nuked and no capacity left.


problem is with these pre built packs is you never know if they are real cells or not as most fake the wrapper so only way to tell is to capacity check each cell, i use a imax charger 15-20 quid.

 

[bikemerchant]

So I want to hold both probes to this bump on the below attached photo to take a reading?

But remember, be very, very careful.

Anything specific I can do to be careful? The exposed end of the multimeter probes are about a fingernail long, I could put PVC tape on them to just expose a tiny bit?

@Bikes4two the seller said he bought the battery separately, I could ask him for the exact link if anyone thinks that would help

All will say for now is you will have to reduce C5 value to reduce the current draw load.

Thanks - this should make it go slower but for a longer amount of time I assume? If you can expand it would be great

 

[soundwave]

watch a few of his vids and you will get the idea just dont short anything or you will go kaboom or kill the bms.

 

[bikemerchant]

Thanks - seems I have to test this white bit but its stuck to whatever is next to it - should I try to rip it or cut something away?

 

[soundwave]

you can unplug it and then test each bank via the pins black is - and the rest positive, do not sort the pins with the multimeter probes!

 

[Nealh]

You only need to pull the White JST sense wire multi connector out for testing when direct access to the cells aren't possible. As both faces of the cells are accessible then the job is as easy as it gets and there is no chance of a fire or shorting of the probes as it just doesn't happen unless you are a fool, more scare mongering from the safety police.

 

[Nealh]

Lets make it simple, all the positive cell can ends which there are 13 all have the insulation card on that end and the negative can end is blank and do not require it.

I have edited a pic of your battery and written the cell groups order on it.
As the cells are top and tailed note they go minus/plus/minus /plus all the way along and the corresponding cells on the other side will be the opposite, starting plus /minus/plus/minus.
I have used 4v as the average cell charge to make it easy conveying the series increase in voltage. The top voltage figures is what one would see accumulatively on the positive reading and the bottom voltages one would see from that face on the positive reading.


Voltage is a like a river and flows from one end to the other, up/down/up/down etc,etc.

The Red numerals indicate the even plus end and the minus Black numerals the negative cell group end, so use a felt pen and mark as I have done on to your battery.
On the other face mark the cells opposite to what I have done and alternate the numbers so all the odd cells -1 cell is 1 red, -3 cell is 3 red and so on till you get to red 13 and the other cells become -2 , -4 , -6 etc.

 

[Nealh]

To simply get the 13 cell group voltages sit the battery on edge with the BMS on the top/upper edge, number #1- #13 on a bit of paper & set you meter to 20v dc.
Place Black probe on -1 as per the edited pic and Red probe on 1 on the opposite face, write down the voltage against #1.
For two do the same Black probe on -2 on opposite face and Red probe on 2 as in the edited pic, write down the voltage against #2.
For three do the same Black probe on -3 as per the edited pic and Red probe on 3 on the opposite face, write down the voltage against #3.
And then ditto for all the other ten voltage readings to see if there is a discrepancy in the cell group balancing.

One has a greater chance of causing a short by going via the BMS JST multiplug as the contacts are very close.

 

[bikemerchant]

Thanks very much!
#11 is the big discrepancy at 3.61V
#10 is 4.15V and all the rest are 4.16V

#1 4.16V
#2 4.16V
#3 4.16V
#4 4.16V
#5 4.16V
#6 4.16V
#7 4.16V
#8 4.16V
#9 4.16V
#10 4.15V
#11 3.61V
#12 4.16V
#13 4.16V

 

[Nealh]

Well done, good job.
#11 is where your problem lies it is the weak link. Either the cell group (6 cells)in that group are all faulty or one or two cells are toast and pulling the cell group down.
It is the reason why you are hitting lvc and seeing voltage collapse.

#10 is fine.

 

[bikemerchant]

Thanks!
So what next?
Is there an easy way to test each cell one by one in this group?
And will I then be buying a new cell/group of cells and replacing that by myself?

 

[Nealh]

You have a couple of choices go back to the seller to see if one can come to some agreement on terms.

Or as a last resort if no joy try to rectify the pack.

One can try checking for a false contact with the nickel interconnector on group #11, If you have a plastic tool or a wooden kekab skewer with a pointy end or ideally a flat thin end. Don't use anything metal and don't have anything metal nearby or wear any jewellery (no rings or neck chains) etc,etc.

Place the battery on a non conductive surface.
With the non conductive tool place the end gently under the metal interconnector (as per the mauve marking in the pic) and see if it lifts off each cell end on #11 on both sides on all 12 individual cells. If one has any bad weld joints the tool will slide under the whole end of the cell or one might see it completely lift a fraction of a thou.

 

[Nealh]

Replacing cells is only an option if you already have a spot welder, nickel and insulator's etc,etc. These items are a couple hundred £££'s to buy.

If the nickel is solid and secure with the welds then one can try to see if the group will balance manually, for this you need an old 5v 500ma phone charger with the charge jack cut off.
On the ends of the bare wires (determine the polarity and heat shrink the outer sheath with Red and Black heat shrink), then either solder on a pair of PCB/arduino pins from a pin strip and charge via the sense wire connector or solder each wire to a flat washer and use a magnet to attach said washers directly to the cell group to enable a slow charge.

One would need to sit and monitor the voltage up to 4.15v with a metre and stop the charge other wise one will over charge with the 5v charger.

 

[Nealh]

Either the cell group is faulty or the BMS is faulty and the bleed resistor is stuck open and self bleeding the voltage down. Once charged let the battery sit for 24 - 48hrs and see if the #11 bleeds down. If it remains stable try discharging the battery by riding and then charge up as normal to see if balance remains good.

The battery is encased in a case so it will have to be heat shrunk again before use, so if poss it might be a good idea to try and solder on an extra balance 14s balance JST lead which has the 1st wire for B0/B-. This then can be left accessible to check cell balancing externally for further issues.

 

[bikemerchant]

Thanks.
I can go to the shop to buy wooden skewers - I have a wooden fork and knife but neither can get under the metal interconnector - it doesn't seem loose.

I have no tools and your second last post seems advanced for me - I never heard of heat shrink, the pins you mention and I have not soldered anything since high school technology class.

Either the cell group is faulty or the BMS is faulty and the bleed resistor is stuck open and self bleeding the voltage down. Once charged let the battery sit for 24 - 48hrs and see if the #11 bleeds down. If it remains stable try discharging the battery by riding and then charge up as normal to see if balance remains good.

This I can kind of understand - I have already rode it a few times until it cut out (after it dropped below 44.0V) and recharged to full after, then rode again with the same issue.
The battery didn't seem fully discharged though, because I could pop it in and out and go at a lower power, it still showed 3 or 4 bars, about 49.0V when parked, and would be fully charged within 2 hours with a 3ah charger.

 

[Nealh]

Because #11 is only 3.61v it is at about 40% charge so is the common denominator, it matters not what the other cell groups charge to. Your battery is only at 40% charge so once that cell group hits lvc first then the BMS will respond and cut power.
The BMS doesn't balance the cells in use so as the other cell groups go from 4.16v to 4v the bad group will go from 3.61v to 3.45v and then eventually as the discharge voltage gets lower the cut out will occur. At approx. 3.45v tha lower cell group #11 will induce more voltage sag and also cut out hence why you still think the battery has good voltage. Normally voltage you see is good if all 13 cells groups are balanced but they aren't as their is a very weak link in #11.

If one hasn't the know how and skills to solder then the battery is scrap and a new one will have to be bought, the cost of a good 30a continuous battery is upwards of £400.

 

[Nealh]

The BMS are rudimentary safety devices, most cells keep good balance without help.
The BMS has two uses it is there to prevent deep voltage discharge so typically prevent cells from going lower then approx. 3 - 3.2v or lvc , once any one cell group reaches lvc is shuts the battery down.
It also manages charging and will bleed down high cell group voltage to balance others but this only occurs if the voltages are also to start with, this means within approx. 0.02v or so. The BMS certainly can't deal with the wayward voltage of #11on it's own, so #11 needs manually raising for it to have a chance to balance it out.