Bike shuts off when riding. Whats going on?

[eNdo]

I have a Wilderness Energy BL-36 eBike that I have replaced the SLA batteries with LiFePO4 cells.

The motor is 36v 600w
The LiFePO4 battery is 36v / 20ah

Whenever I lift the front wheel (with the motor) off the ground and turn the throttle, the motor kicks in and runs fine when there is no friction on the wheel.

However, when I get on the bike and ride it, the controller shuts off after about 10-20 seconds of riding.

I'm really not sure what is causing this. Any ideas? When I first purchased the batteries and got everything connected, I was able to ride my bike all the way to work and back in one day. I rode to work 16 miles, charged the batteries during the day, and then rode home without a problem. That was the one and only time that I was able to ride the bike. Everytime I tried to ride it afterwards, the power shuts off within seconds of me getting on it.

Is the controller being overloaded? How can I fix this?

Like I said, the motor runs perfectly fine when it is lifted off the ground and there is zero friction.

Thanks for any help!

 

[needWheels]

I bet the BMS is bad, unless there is a bad cell in the pack.
However I will let the experts give you a better answer...

counting down to dogman helping out in 3... 2... 1... :wink:

ps. do you know how to check the balance of the cells at the little wires?
it will help you make sure there's not a problem with that

 

[eNdo]

I bet the BMS is bad, unless there is a bad cell in the pack.


ps. do you know how to check the balance of the cells at the little wires?
it will help you make sure there's not a problem with that

I don't know how to do that. I'll search the forums here to find out though. Unless its quick and you could explain?

The BMS did get dropped once but I didn't think it was that hard of a fall.

 

[needWheels]

Your battery probably connects to the BMS via a little connector with 12 wires (or two connectors of 6 wires)
By the way is your BMS very thin like the one on the right here:
http://www.superetrade.com/images/cammycc/bicycle battery/pic/cc2.jpg

If you disconnect the connector you can VERY CAREFULLY measure the voltage at each one of the connectors to get the voltage at each cell group. I only recently learned to do this myself. The meter positive goes to the little points and the meter negative goes to the battery negative.

Once you have the 12 voltages you can then use my little voltage difference calculator to see each group and make sure they are somewhere around 3.3-3.4v and fairly close to each other
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=10242

What you are trying to make sure of is that there isn't a single group that is way too low.

Also I've been told not to deep discharge a pack when it's new, 16 miles may have been a bit much for a new pack?
But that's not necessarily true, just what I've read.

ps. who made the battery? Is it an ebay china special?

 

[eNdo]

Yes, its an ebay China special, lol!

And the BMS is exactly like the one in the picture you sent.

Hmm....

 

[dnmun]

did this lifepo4 battery pack work before you dropped the BMS? do you have picture of the BMS where it hit?

you should be able to measure the voltage across each cell on the BMS while it is still connected through the plug. put the probes on each end of the shunt resistor. or each one sequentially. record each cell.

 

[needWheels]

Ah I found your original thread when you first got it, looks like a CammyCC pack?
Or was it another vendor?
http://www.endless-sphere.com/forums/viewtopic.php?p=124616#p124616

So you have the better BMS on the left, not the one on the right in the picture I showed you?

By the way, what was the weight on that battery in the end, you mentioned it seemed far too light?
A 36v 20ah pack made of 168 of 18650 cells should weigh about 16.5 pounds at minimum.

 

[potatonet]

the bms output is too low for the amount of power you are putting to the motor with the controller.

i had this problem with my battery last week.

I have a 48V 25ah battery so I just bypassed the BMS and use the BMS to charge the battery.

I use a separate grounding wire soldered to the same point that the BMS ground is soldered for the controller.

 

[dogman dan]

Heh heh, a little slow today. I rarely work after lunchtime so usually I'm here on the sphere by noon. But the Giro De Italia is on the web, so I have to watch todays stage first!

It does sound like a battery problem, particularly a bms problem. But first, I'd try the sla's again to be sure it's not something else, like a throttle or controller malfunction. Hopefully you can just replace the bms and be good to go. I'd be very cautious about bypassing the bms though. Do that and cammy would not be needing to care about fixing your battery under warrantee if it is the cells that is the problem. There is no reason that particular motor should be drawing enough amps to trip a good bms, so I suspect something is wrong with the bms.

 

[eNdo]

Heh heh, a little slow today. I rarely work after lunchtime so usually I'm here on the sphere by noon. But the Giro De Italia is on the web, so I have to watch todays stage first!

It does sound like a battery problem, particularly a bms problem. But first, I'd try the sla's again to be sure it's not something else, like a throttle or controller malfunction. Hopefully you can just replace the bms and be good to go. I'd be very cautious about bypassing the bms though. Do that and cammy would not be needing to care about fixing your battery under warrantee if it is the cells that is the problem. There is no reason that particular motor should be drawing enough amps to trip a good bms, so I suspect something is wrong with the bms.

Thanks. I'll try reconnecting the sla's and see what happens. However, the sla's have been just sitting there in my garage for quite a while so I'm not even sure if they are good anymore.

 

[eNdo]

Ah I found your original thread when you first got it, looks like a CammyCC pack?
Or was it another vendor?
http://www.endless-sphere.com/forums/viewtopic.php?p=124616#p124616

So you have the better BMS on the left, not the one on the right in the picture I showed you?

By the way, what was the weight on that battery in the end, you mentioned it seemed far too light?
A 36v 20ah pack made of 168 of 18650 cells should weigh about 16.5 pounds at minimum.

I believe it was another vendor since "CammyCC" doesn't ring a bell.

Yes, I have the one on the left in the picture you sent.

16.5 pounds sounds about right for the weight of the batteries I have.

 

[cammycc]

Hi,
This is cammycc,I have learned something happen.
I think it is the BMS problem.or the connector problem.

Do you let the battery full charger before you use at first time?

Hope I can help you resolve this problem.
Thanks

 

[eNdo]

Hi,
This is cammycc,I have learned something happen.
I think it is the BMS problem.or the connector problem.

Do you let the battery full charger before you use at first time?

Hope I can help you resolve this problem.
Thanks

Hello cammycc,

Yes, the battery was fully charged before I used it for the first time.

 

[eNdo]

Heh heh, a little slow today. I rarely work after lunchtime so usually I'm here on the sphere by noon. But the Giro De Italia is on the web, so I have to watch todays stage first!

It does sound like a battery problem, particularly a bms problem. But first, I'd try the sla's again to be sure it's not something else, like a throttle or controller malfunction. Hopefully you can just replace the bms and be good to go. I'd be very cautious about bypassing the bms though. Do that and cammy would not be needing to care about fixing your battery under warrantee if it is the cells that is the problem. There is no reason that particular motor should be drawing enough amps to trip a good bms, so I suspect something is wrong with the bms.

Thanks. I'll try reconnecting the sla's and see what happens. However, the sla's have been just sitting there in my garage for quite a while so I'm not even sure if they are good anymore.

I reconnected my old SLA batteries to the bike and they appear to be running fine.
Next, I will test the voltage on the cells in the LiFePO batteries. I'm hoping the problem is just with the BMS. If so, where is the most reliable place to purchase one?

 

[eNdo]

I let the LiFePO batteries charge for well over 24 hours, re-connected them to the controller and was able to ride the bike around the block for the first time in quite a while. I really didn't do anything differently except let the batteries charge for a while.

Weird.

I'm going to attempt riding the bike a longer distance this weekend and see what happens. I just ordered a pair of Anderson PP45 connectors. Currently I'm using some cheapo radio shack connectors. Maybe that was the problem...not enough contact or something.

I never thought that building an e-bike would lead to me practically getting a degree in electrical engineering! Sheesh!

Also, the front wheel where the motor is mounted seems to be getting severely wobbly. Maybe trueing the wheel will help some.

 

[dnmun]

there is nothing wrong with getting a degree in EE, but don't worry, you are a long way from that.

i suspect the battery was not balanced and still had some low cells when you used it initially and when it finally got all the cells charged up, then you no longer were having low voltage cutouts from the BMS, which was functioning normally.

 

[nineball]

I'm going to attempt riding the bike a longer distance this weekend and see what happens.

Ya might what to do a few short runs and then 24 hour charge cycles to break that puppy in.

 

[potatonet]

I still think the BMS cant put out the amperage needed.

I have a 25ah pack and it needs to discharge 80A peak (50A continuous)

dont think there is a BMS that can handle that.

I need to upgrade the FETs on the BMS to see if that would help.

 

[dnmun]

pnet, what kinda pack and BMS are you using? the output FETs on the BMS can be bypassed, and use the gate driver for the FETs to sink the brake line to shut off the controller when you get to overcurrent and for LVC control. you can then use the controller current limiting or get the CA to do the current limiting.

if you show us a picture we can show you how to solder up the shunt on the BMS to keep it from shutting down the output from overcurrent limits.

 

[potatonet]

I know how to solder up the shunt =-)

I will find a pic....

voila:

the second BMS was a comparison between the 2 that I have available, the top one is being used.

 

[potatonet]

dnmun, you have seen the pack before, you just didnt know the specs of the battery.

there are 96 4.125ah cells in a 16S6P arrangement. each cell can discharge 4C bursts 2C constant.

found a part number on a cell and looked it up.

 

[dnmun]

is that mr pings BMS? i think they are 20A packs, so you will need alotta solder.

you could just jumper over the shunt with a wire too, maybe use some old desoldering braid which is already full of solder and put it across the shunt, or solder it from one end to the other on top of the shunt. that way you can save the LVC function from the BMS, and still charge normally. that is a lot of power. can your controller take it?

 

[potatonet]

2C is not a high discharge rate...

There are prismatic cells that discharge well over 30C used for RC cars...

dont know what you are talking about...

just found out these are 5ah cells, they can discharge 3-5 C each continuously and being that this is actually a 16S5P arrangement that gives me a working amperage of about 5*5*3 =75 - 125 amps continuous and 8C peak so 200A peak.

the controller is a 50A CL controller, I can set the amp limit, right now its set at 80A its got a few fans on it too.

Im going to get an infineon controller soon and get more batteries.

 

[eNdo]

Update:

The last couple of days I've been able to ride the bike for 10-15 miles without a problem. I usually let the batteries charge for a good 18+ hours before riding again. Everything seemed fine up until today when I took it for an extended ride somewhere around 18+ miles and minimum pedaling. I basically ran the bike between 75% - 100% throttle for quite some time. The bike performed great until I turned around and started heading back home. On flat ground, power to the controller seemed to have just cut out. The power meter on the throttle instantly went from the green light to the empty red light and I was no longer getting any power to the motor.

I pulled over to check everything out and see if I could figure out what the problem is. It turns out that the batteries seem to be fine. In order to get power back into the controller and up and running, I simply shut off the controller, disconnected the positive battery cable, reconnected it (causing a spark) and powering the controller back up. Every time I do this, the power meter shows a green light and I am able to ride the bike once again without problem. Although after the first time the power cuts off, I am no longer able to ride the bike either at full throttle or up a hill. It seems that I am no longer able to run the motor at full speed.

Again, every time I disconnect the battery and reset the controller, everything seems fine. Because of this, I'm pretty sure there is nothing wrong with my batteries. Could it just be that the Wilderness Energy BL-36 kit is not suitable to be modified with LiFePO4 batteries? Are they somehow overpowering the motor and/or controller to the point of automatic shut-off?

The controller is a 50amp one. I am using 36v 20ah batteries.

Any suggestions on what to do next? I'm thinking I should look into a different controller now but thought I'd ask here to get some advice first. It seems like there are quite a few other people who experience this same problem of needing to reset the controller. I'll search the forums and dig around to see if I can find any answers...

Ps. One thing that I found strange was how the red LED light on the power meter manages to stay on until I come to a stop at which point it slowly fades out. Does riding the bike without any throttle applied manage to generate a small amount of power that gets routed back to the controller???

 

[eNdo]

// cue crickets