Help me diagnose and possibly resurrect a salvaged e-scooter

redmouse

10 W
Joined
Aug 3, 2017
Messages
65
Hi everyone, I happen to have this broken down scooter in my backyard, and I'm wondering if there is a chance to make it running again?

More pictures: https://imgur.com/a/sAbrs

The owner said 2 years ago, he rode it up a very steep hill and it stopped running suddenly. He knows nothing more, he just towed it back home and parked in in the backyard where it has been sitting ever since.

The big questions are probably:
1.) what broke down? (I'm guessing he probably overheated the motor and cooked the windings? Or is there something else that typically breaks down in heavy load conditions?)
2.) is there a chance at all the batteries are still good? They appear to be prismatic LiFePo, but it's been 2 years outside with zero maintenance. If I turn on the keys, nothing lights up and power meter says empty. I reckon the batteries make bulk of the vehicle cost, so if they are bad, there is no point in trying to fix it.

I have some basic knowledge of electricity, I have a charger for the bike, a multimeter, but a very little practical experience. Are there some specific steps I can do to identify the problems? I will be happy for any tips, I plan to look into it this weekend and I can report back what I found.

From what I found, the motor appears to be 13" QS motor 60V 3500W

__________________________________________________________
UPDATE 1: I tried charging it, but the charger doesn't produce any current - I tried measuring the output cable while the charger was plugged in (red light was on, fan was on) and there is 0 volts. I tried measuring my ebike charger, and it has 54V, so I assume there shoudl always be some voltage even when it's not plugged into the motorbike? Picture of charger: http://i.imgur.com/06Kw50l.jpg
__________________________________________________________
UPDATE 2: I've identified the scooter as X-Treme XM-3500Li http://www.monstermotorscooter.com/xm-3500li.html
__________________________________________________________
UPDATE 3: I tried hooking it up with my ebike charger to no avail. Then I disassembled the covers and measured individual cells. There seems to be only residual voltage - like 0.1 V. So I'm guessing the batteries are toast.
But I will try one more thing - attach the entire pack to 12V car battery charger and see what happens. (inb4 RIP)
__________________________________________________________
UPDATE 4: MAJOR SUCCESS! Got it charging with original 72V charger! I used procedure recommended by Dauntless: Took a 12V car battery charger and connected it to 3 cells at a time, and let it charge up to 9.6V. I did this with all cells until they had at least 2.9V charge each - too around 20 minutes for one block. After this, I connected the original charger and it's charging!
Also to my surprise, when I turned on the key, the lights lit up and even the blinkers work, not bad :) Didn't try runnign the motor yet, will leave it charging for couple of hours.
__________________________________________________________
UPDATE 5: Left it charging for 3 hours, let it sit overnight and checked voltage in the morning - according to onboard meter, it dropped from full to 3/4. So probably not a good sign. I hope I will be able to at least use the half-dead battery to test the other components. I already tried twisting the throttle, but it's unresponsive. Gonna dig in further and try diagnostic tricks from Assstray http://www.ebikes.ca/learn/troubleshooting.html
__________________________________________________________
UPDATE 6:
Got it running using a new ebike controller https://youtu.be/ZS76oelta74
 
Number one is to figure out if the batteries will be usable. I want to say number two is to replace them once you realize they're no good, but occasionally you'll get lucky. No, not all chargers give voltage when not connected. Not always a good idea to just hook up the charger, sometimes you need to "Prime" it by connecting to another battery of the same voltage. east enough when you have 12v and you can connect positive to positive on you car battery, but these you might want to go 3 cells at a time to it. Doesn't have to be there long, but it makes a huge difference on a dormant battery.. Are you even sure there isn't a break in the connecting between the battery and the charger plug? Problems can be anywhere.

Then you crack things open and look for some sign of, say, the controller housing was hot, something odd about the motor, etc. I still have one that I haven't resolved but haven't kept messing with. It would run fine with the wheel on the ground, but was real weird when I tried to ride it. I wondered at what it might have done when it failed. The batteries were SLA's and quite swollen, I guess the guy thought he just had to keep charging them

Good lawrd, a 3500w scooter would be well worth a few hundred in batters to fix. It looks good in the little pictures.
 
Dauntless, so if I understood correctly, you are saying I should take a 12V car battery charger and conmect it to 3 cells at a time, and by this give them a little "forced" charge? Won't that cause some dysballance?

I also have a 1A 48V Li-ion ebike battery charger that put out 54V unconnected. Might that be a better idea than a car battery charger?
 
I tried hooking it up with my ebike charger to no avail. Then I disassembled the covers and measured individual cells. There seems to be only residual voltage - like 0.1 V. So I'm guessing the batteries are toast.
But I will try one more thing - attach the entire pack to 12V car battery charger and see what happens. (inb4 RIP)
 
MAJOR SUCCESS! Got it charging with original 72V charger! I used procedure recommended by Dauntless: Took a 12V car battery charger and connected it to 3 cells at a time, and let it charge up to 9.6V. I did this with all cells until they had at least 2.9V charge each - took around 20 minutes for one block. After this, I connected the original charger and it's charging!


Also to my surprise, when I turned on the key, the lights lit up and even the blinkers work, not bad :) Didn't try running the motor yet, will leave it charging for couple of hours.
 
Yeah, I had a little trepidation telling you to do it with the volatile lithium which I'd never tried it with. But it's the only way to be sure. You're lucky you have a car charger that works without any charge in a battery. I was saying hook them to the car battery itself. You'll need to give it maybe a half day charge before you decide if the batteries are usable or not. You still might discover it'll run your motor for 90 seconds, etc. I'm still doubting they'll work out, but you feel better being sure, right?
 
Yeah, sometimes one has more luck than reason :) YOLO, I say :)

After 3 hours of charge and leaving it overnight, the voltage dropped - according to anboard charge meter, it was full in the evening, and 3/4 in the morning - probably not a good sign. But I think I can at least use the half-dead battery pack to test other components. Going to strip it more and examine the controller, and possibly test hall sensors on the motor? I've been recommended this http://www.ebikes.ca/learn/troubleshooting.html by Assstray.
 
The drop could be that something is still "on" and it drained some of the charge from the pack. Most of those scooters have a DC-DC to make "12v" for lights and stuff, and if that stayed connected even with the lights off it'd still drain some power. So would the controller itself, if left connected, as would any power meter displays it has.

If there's a keyswitch make sure that's actually in the OFF position. If it is then the drain is from something else, probably the cells themselves (internally from age or damage) or possibly corrosion between the cells.


It could also be from a BMS attached to the cells. (damaged, faulty, or even just balancing them normally because some of them overcharged a lot).
 
The key was in off position. I assume there must be a BMS, but there is nothing connected to the individual cells, so the BMS is probably just monitoring overall voltage of the entire pack. I'll dig in further and see what I find.
 
Ok, looks like I found the culprit. It was the controller, seems like two MOSFETs burned out, and possibly something else. What do you guys think?
Pictures: http://imgur.com/a/IS2jl

Can I possibly order the mosfets somewhere and replace them, or is the entire controller toast?
 

Attachments

  • 20170806_131333.jpg
    20170806_131333.jpg
    418.2 KB · Views: 3,281
  • 20170806_131243.jpg
    20170806_131243.jpg
    417 KB · Views: 3,281
  • 20170806_131320.jpg
    20170806_131320.jpg
    225.8 KB · Views: 3,281
  • 20170806_131600.jpg
    20170806_131600.jpg
    61 KB · Views: 3,281
That's definitely a fried FET. It's also possible others are fried in the same phase with it, so you can test them all via the info here
http://www.ebikes.ca/learn/troubleshooting.html
but you may have to unsolder some to do that. (the known-dead ones, for instance)

That particular FET p/n is a pretty commonly used one. Mouser probably has them (or Digikey, etc).

Sometimes other things like gate drivers can be damaged when a FET blows, depending on what happens inside the FET.
 
I have what looks to be that exact 60v 8000w controller sitting around brand new if you're looking for one. I bought it for a project it turned out to be too big to fit..

P.s. yours looks like it got water infiltration... And the picture showing the underside with the main traces looks like it shorted to the case at some point.
 
Thanks Voltron, though I'm in New Zealand, the postage wouldn't make it worth it. I think the water damage is secondary, after it has been sitting outside. It was sealed pretty tight when I took it apart. Guess I'll see later.

Also thanks Amberwolf for advice.

I desoldered the visibly bad mosfets, but turns out there were much more bad ones. Had to desolder all of them in the end, to identify which are good. 12 out of 24 were shorted.
So I ordered more replacements and gonna attempt to solder it all together. Will probably fail, but I wanna try :) Plan B is a new controller.
 
I have one question, and would be happy is someone can verify:

On my original controller, the mosfets were mounted to a heasink through some kind of soft rubbery strip. What's up with that? Should I keep it, or do I better just mount the mosfets directly on the aluminium block?
 
Welp, that was disappointing.

I replaced the faulty mosfets, turned on the key... and the controller exploded :cry:
If I messed up something, or there was some other fault somewhere else, I don't know. But I'm done trying to fix it - next time I'll replace the entire controller and battery, if the owner decides to pay for it.

R.I.P.
 
It's probably gate drivers and/or other related parts, turning FETs on that shouldn't be, creating a full short across the battery thru the FETs. :(

It's not that uncommon for a FET that fails to do so in a way that takes out gate driver components, unfortunately. I have a box of controllers that have such issues that are used for parts to fix other less-resource-intensive damaged stuff. :/
 
Update after a long time: Got myself a spare controller from an ebike, so I hooked it up to the scooter and it works! It's just 35A, but good enough for testing. Gonna try riding around for a bit and see how the battery holds up. So far it can move under it's own power in my backyard, so that's progress.

Video here :)
https://youtu.be/ZS76oelta74
 
Sorry for resurrecting this - I managed to buy this scooter and now I'm planning a build.
Would you experienced people advise me how many amps I can run this motor on?
More info here:
https://endless-sphere.com/forums/viewtopic.php?f=30&t=99709
 
Back
Top