I messed up a custom ebike build please help.

[JosephShotts]

I am building an ebike using these parts and a mountain bike: https://www.ebay.com/itm/364149505027, https://www.ebay.com/itm/175567304674?var=474907899771. I had stuff working and then made a terrible error. I have some experience with small scale electronics but this project is well out side of my normal scope. Any help or advice would be really apreciated.

I first dropped the battery during a test drive where the connector ripped out and it shorted. Then when I went to test the battery by plugging it into the bike I flipped the posative and negative terminals which cause a capacitor to blow up inside the battery case. The capictor was on a circuit board wired from the battery output terminals to an auxillary usb port pictured below. The BMS also wasn't working correctly after this mistake.

I replaced the BMS with https://www.amazon.com/dp/B09FTKN97L?psc=1&ref=ppx_yo2ov_dt_b_product_details, and got the battery to charge and show 54v at it's terminal.

The problem I need help with is connecting the controller to the hopefully fixed battery. I was getting a massive spark when I plugged it in so I did some looking around and tried to make a precharge circuit using this https://goldenmotor.com/SMF/index.php?topic=5874.0 and two 50Watt 6 Ohm resistors in series I already had.

This stopped the spark but I can't get anything to turn on when I connect the main power. The throttle and the display are the two main things that would. I did try a new speed controller as well but I bought one with differant connectors than the original. It did have similar issues though so I'm thinking that I'm doing something wrong with the connection between the speed controller and battery.

 

[HumanPerson]

Welcome to the forums mate!

We have some battery people that might chime in here soon.

Have you tried another battery at all?

 

[JosephShotts]

Thanks! I ordered one but it doesn't come in until monday.

 

[ElHegpah]

You can buy pre-built, XT90 anti-spark connectors on pigtails, easy to splice on. These have been in use for many years and are inexpensive. No need to re-invent the wheel, these are a known quantity.

The spark itself is not real harmful, but over time can do some damage. Assuming you match the colors, this is important.

Connection to controller from battery is simple, red and black, nothing else. Be careful to match the colors. The controller may have been damaged here, what "similar issues" did the replacement controller have, you seem to be saying there are no display, no lights, and no indication whatsoever of power. Do you have an extra display as well, and are there any other indications of power, on controller, battery, anywhere?

When you say "speed controller", I am thinking throttle, which connects to the controller. Leave it off and just connect display to controller, and battery power to controller. Nothing else. See if it lights up. If not, display or controller (or both) is bad. Then disconnect power to the controller, re-connect throttle, connect power to controller last, and test. If throttle has power level indicators, (bad idea) it may be blown by the reversed power cables.

Make certain the motor is secured to the frame before testing throttle.

You may have damaged the battery by dropping it while running, the controller by reversing the pos and neg, the display thru that same error, and possibly the motor at the same time. If the motor is loose at all and rips clean off the frame, ripping the wires out of the axle, you may have been able to destroy all the important pieces of the kit.

Watch a few youtube vids and be more careful.

 

[JosephShotts]

Thanks for the help. I know I really screwed up this build in one aweful moment just still trying to figure out where. Sorry if these posts are unclear or hard to follow.

I wasn't referring to the throttle at all earlier or the LCD screen, I was talking about the controller (referring to the enclosed metal box with all of the connectors) itself. I keep getting the same problem testing different things one by one.

To test it I have been simply connecting the battery first through the 12 ohm precharge resistors. Then disconnecting the resistors and connecting the battery directly to the controller. When the controller is connected through the precharge resistors a large amount of current flows through the resistors (they're rated for 50w and heat up) but the battery doesn't shut off. When the controller is connected directly to the battery, the battery shuts off.

I know the battery shuts off because there is a LED voltage indicator on the battery housing. It shows 0v when either controller is connected directly (but full voltage normally). I also have been measuring the voltage across the pos/neg terminals on the controller. It shows approx. 1.9v when the precharge resistors are connected and approx. 0.7v that drops to 0v when the controllers are connected directly to the battery.

With the old controller (the one that came with the kit) I tested it with just the motor and battery, with just the display and battery, and with just the throttle and battery. In all cases it had the same result described above.
With the new controller I tested it on its own (with just the battery hooked up to it) and it had the same result. The new controller I bought has different connectors and doesn't connect to a display. I didn't want to damage it as I intended to return it so I didn't connect anything to it.

It's great to know that there is an actual product for avoiding a shock and I will use those connectors in the future instead of the jank setup I used to test the controllers.

I also should mention I replaced the BMS on the battery a second time. I took a more careful look at the first BMS I purchased and figured it could have been of low quality or illegitimate (I had literally bought it bc It seemed exactly the same as the original bms that was on the battery). So I purchased this one: https://www.amazon.com/dp/B0876NQ88P?psc=1&ref=ppx_yo2ov_dt_b_product_details. This BMS checks for over heating and is more water resilient so I feel better about having it in the bike anyway. After installing it I retested everything and had the exact same issues.

All that to say I'm guessing that at least the old controller is bad and possibly the screen and throttle as well. I'm wondering if the new controller I bought had these issues because I was testing it with only the controller hooked up to the battery?

At any rate I ordered a third and hopefully higher quality controller (https://www.amazon.com/gp/product/B0C2TPTNRZ/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1) along with a new screen, throttle (the old throttle had the battery level indicators you described earlier), and those connectors. Hopefully the new setup will fix this issue.

 

[ElHegpah]

That is not good and this may get to be an expensive lesson. Most of the best ones are.

What you need to do is some major battery testing. Seems like you get correct voltage with no load, but add a load and voltage drops to zero.
This is very, very bad. Old controller and display could be an issue, but a new controller, no display, no damaged components, this should not cause a shutdown.

Ideal would be to rig a load, like a lightbulb, so you can test voltage with the load on.

Next would be to test the series strings at the BMS.

You use the multipin connector with all the thin balance wires.

You test neg to pin one, then neg to pin two. Each is adding one cell value, 4.2V or less. These add each new value to the previous total, so you need to write down the reading and subtract the previous one. 4.2, 8.4, 12.6, etc. except they will not add exactly 4.2 each time.

What you likely have is one or more low or damaged cell groups. They will appear to take a charge, but they have little to no capacity, as soon as you apply a load, their voltage craters and causes the BMS to shut down power output as a protective measure.

You may need to do this test with the main power thru the BMS disconnected. Do not load the battery for any significant time in this condition as cells at a low voltage become unstable and can burst into flame, the heat igniting the others around them, the fire must be smothered and water makes it worse. You could just test main voltage and observe the drop, or test the balance wires and find the one or more damaged cell groups.

With a fairly new battery, it is feasible to replace individual cell groups. You need a spot-welder.

You may find a broken or partially broken spot weld to just one or a few cell groups. probably more likely an individual connection problem rather than the cells themselves, but no way to be sure without testing and checking all welds.

 

[JosephShotts]

I measured the voltage of each of the cell groups (I measured directly on the cells themselves while wearing gloves, glasses and on a piece of stone that I could move if necessary) they all seem normal and the values are below. I measured them twice and got the same values. I don't see any visable damage to the spot welded conections.

I don't know that I have a suitable light bulb to do a load test. I should say again that when I had those 50w 12ohm resistors wired in series they heated up significantly (so the battery was at least supplying some current?) I also was measuring the voltage after the resistors so there should be some voltage drop (probably not by that much though?) and that the LED voltage indicator on the battery housing showed full charge. It only showed no charge when I connected the controller directly with no resistors.

I also only did only one test with the new controller as I didn't want to break it. All the other testing has been done on the old controller.

Is it possible that one 18650 cell is bad and not an entire cell group? Could the BMS be detecting something else like a short in the controller or a poor connection maybe?

I really apreciate your help and yeah this is become quite the lesson in patience and doing more careful troubleshooting with expensive componenets.

Cell Group Number	Expected Voltage	Measured Voltage
1	4.2	4.19
2	8.4	8.39
3	12.6	12.60
4	16.8	16.81
5	21.0	21.02
6	25.2	25.23
7	29.4	29.27
8	33.6	33.53
9	37.8	37.8
10	42.0	42.1
11	46.2	46.4
12	50.6	50.6
13	54.8	54.9

 

[ElHegpah]

First, YOU HAVE TO DO THE MATH. Third column, each reading, subtract the previous TOTAL reading to get the total for that group only. Do NOT make me do this again.

Next, DO IT AGAIN. This is wrong.

Anything over 4.20 is DANGEROUS.

Any group more than .03 different is bad.

Cell groups 3 thru 6 are all 4.21. That is TOO HIGH.

Group 7 is 4.04. Very bad. Look real hard for a bad connection, discolored cell, etc. The number is not bad, the difference from all other groups is.

Group 9 is 4.27 !!!!! RUN!!! GET IT OUT OF THE HOUSE!!!!!

Group 11 is 4.3. Write out the second digit. 4.30 is impossibly high, and if correct, I am serious about getting it out of the building.

The expected voltage on a fully charged battery is 4.20V, every single group. Small variation OK.


Triple check each reading. Keep trying till you get THREE IDENTICAL READINGS.

Each cell in a parallel group will have the same voltage. This is usually just one bad cell.

I am NOT an electrical engineer, but AFTER you get some reliable readings, ONLY AFTER THAT, and in fact, post them here, do the math on that last column, then you will hook up those resistors and check total voltage and each cell group. Just one of those drops, the BMS should shut down the battery. Controller only sees total voltage.

BMS does not see the controller. Total voltage, parallel group voltage, both monitored for LVC, voltage IN controlled for HVC, and amps out with a limit. That's it, for the vast majority.

OK, now on the new controller. Most of them have the on/off switch on the display, and will not work or only work in very basic mode without the display connected.

I am thinking the old controller is blown. Now, IF it has a dead short, AND IF there is a fuse on the BMS, it may have blown this fuse. Check and replace. Connecting the new controller, WITH DISPLAY, should do no harm. Do NOT repeat NOT connect anything that touched the reversed polarity situation. Fruit of the poison tree. Battery, avoid the resisters, connect quickly but WITH CORRECT POLARITY, red and black. So, Battery, New controller, new controller Display.

 

[JosephShotts]

I measured their values more carefully when I got home. I tested around in multiple spots around each cell group (at least 3 spots) If the last place was jumping between values I listed both below. Once the multimeter gets past around 33v it loses precision and rounds which is why some cells are calculated to be 4.3v. I tested across some of those cell groups and got a value of around 4.25v.

Regardless of there exact measurements it's pretty clear that many of these cell groups are overcharged and dangerous. I think this is the point where I call it quits and get a new battery (definitly an expensive lesson).

Right now the battery is in a bucket in a bucket sitting outside in my driveway. I am going to look around for a way to dispose of it. Do you think they will be safe to handle and transport if I let them sit for awhile? I would definitly apreciate some advice or ideas to handle and dispose of them at this point.

As to how they got overcharged I'm guessing when I replaced the BMS and charged the battery the BMS wasn't right for this brand of batteries. When I charged the battery this brand of 18650 cells didn't have overcharge protection? Just a guess.

Thanks again for your help and the warning. It's just me in my home right now and I don't think my family would apreciate it if I burned down the house.

Measured Voltages	Cell Group Value
4.15	4.15
8.31	4.16
12.49	4.18
16.65,16.66	4.17
20.83	4.17
25.01, 25.02	4.17
29.01	4.27
33.31, 33.32	4.04
37.6	4.29
41.9	4.3
46.2	4.3
50.4	4.2
54.7	4.3

 

[ElHegpah]

No 18650 has overcharge protection. That is why you have a BMS.

I would get a better voltmeter. That one is simply not good enough.

Group 7 the math is wrong. I get 4.00. That is the only overcharged group until the meter begins rounding. Likely need to change the voltage range on the meter dial. Also that group was the way low one on the first round. Note that the NEXT group is now way low. Each measurement affects the next one, errors are cumulative.

I would run the whole test AGAIN. Batteries are expensive.

ONCE is bad luck. TWICE is coincidence. THREE TIMES is enemy action.

Take your time. Go slow. Double-check the math.

If you do need to dispose, they are considered hazardous waste. IMO you are not there yet.

The womenfolk do tend to get annoyed if you burn the house down. Something about having to re-decorate.

 

[JosephShotts]

I just waited for the new battery and everything from the original kit works with that new battery. The old battery could likely be fixed with a spot welder and some more testing but I was pretty done with it. Thank you so much for your help with this project I couldn't have done it without you. Here is some pics of the bike:

 

[Snoop]

Is that a tartigrade?

 

[ElHegpah]

It is quite likely that the original battery is still useful as a spare, or for a second bike for a backup, friend, or spouse.

Also, overcharges are fairly rare. Nailing down the cause and mechanism for this could be very useful to the community.

Batteries are not cheap, either.

At least, before you chuck it, tear it open and carefully examine all the welds. Also attachments for the balance wires.

IMO, an accurate check of the readings direct to the cells, and an examination of the balance wires plus the voltage check at the BMS connectors, should reveal some useful info. My educated guess would be that some of the balance wires, particularly those going to the overcharged groups, must have a poor connection.

I will lose sleep at night and possibly go into a severe downward depressive spiral if you do not take advantage of a rare opportunity to investigate a possibly repairable serious safety issue which could actually save lives in the future. (seriously, it actually could)

Your ride looks fine, not terribly different from mine.