Trouble shooting a broken bike

[newguy]

Hey guys,

Had an awful day with the ebike yesterday, fell off in the ice, broke the pannier... replaced it.....then the bike stopped working!!!! I was travelling slowly uphill (to avoid slipping in ice again) at the time and the motor cut out. Need help identifying the source of the problem.

I'm running a 350w Bafang motor 36v with a KU90 controller, running motor sensor-less. Got the multimeter out when i got home, and checked the recommended usual things.

• Battery resting at 40v
• Thumb throttle is powered at the 5v+/-, and responding correctly to throttle moving.
• Checked continuity of wires from controller to end points, all fine
• Checked continuity through all variations of phase wires, all fine

(I'm only using a battery, controller, throttle and motor)
I haven't opened up the controller box, but im getting the impression it might be the problem. I read that i might have blown a mosfet?!! i think that is just a transistor, right? How do I test it? How could i replace it and would it be worth my while? Is the KU90 a cheap controller that was always likely to fail anyway?

Thanks

 

[Deleted member 4366]

It'll be something a lot simpler than a blown mosfet. You didn't do anything that could cause one to blow, so why would it? First question: Is your controller definitely 36v and not set to 48v? Next question: Is the thin wire definitely and securely connected to 36v? Third question: Why are you running sensorless?

 

[Jeremy]

Bad luck. To test for a blown FET disconnect the controller from the motor and battery and, with your meter on the resistance range, check the resistance between each phase wire at the controller and the positive and negative controller power leads in turn. You may need to do this twice, once with themeter red lead on each phase wire and the black lead alternatively on the positive and negative controller wires and once the other way around.

If you get a very low resistance (less than an ohm or so) for any of these measurements then it indicates a blown FET.

The controllers are cheap, but they are generally pretty reliable if not abused. The worst condition for a controller (or, more specifically, the FETs) is the very condition you describe, going slowly under a fairly high torque load. This is the very condition where the FET current is at its absolute maximum as a rule.

FETs can be replaced, but it's an hour or so's work to strip the controller down, remove the FETs and fit new ones, and replacement FETs often cost as much, or more, than the cost of the controller.

 

[newguy]

D8veh,
1. Its a 36v controller, well thats what the sticker says, i thought myself 40v sounded high and maybe the controller thought i was running 48v and cut off? dont know if controllers are that smart.
2. Thin wire is also reading 40v
3. tbh there isnt really a good reason why im not using the sensored mode, i felt wiring it in my set up was more hassle than its worth, wasnt enough connections in my connect:
PX0551 Flex Mount Male Circular Connector : UK and European Mains Power Connectors : Maplin Electronics

Jeremy, thanks, ill try this out once i rule everything else out, I read that i woud see and smell smoke, this didnt happen and the controller didn't even appear to warm really.

I should also mention that the bike was playing up a bit just before it cut out, jumpy starts and short moments of no power. I a FET was to blow, would the bike be lifeless, or would it still kick occasionally?

I suspect and hope that D8veh is correct and it something simple.

 

[Jeremy]

From the added info of the bike playing up before this happened, I'd suggest it's a dodgy connection somewhere, rather than a blown FET. FETs can blow with no smoke or sparks, or with the controller getting warm, but more often than not there is some indication, like a pop or maybe a blown fuse.

I'd double check every connection. As you have a meter, one thing that you can do to see if power is getting to wires with all the connectors made, is to use pins (the dressmaking sort) as probes through the wire insulation and into the copper core. In really stubborn problems I've done this to see if power is really getting to where you think it is. The tiny holes left in the insulation after the pins have been removed don't usually cause a problem, but can be sealed with a smear of Vaseline or maybe sealant.

 

[Deleted member 4366]

Pull the phase wire connectors apart. Crimp them down a bit and re-connect them. Those bullet connectors often give problems,

I have heard of a few dodgy solder connections in the controller, so if all else fails to give an answer, have a look in there three screws on the side and 4 in the wires end, and you can slide out the contents.

My money's on that Maplin connector.

 

[newguy]

From the added info of the bike playing up before this happened, I'd suggest it's a dodgy connection somewhere, rather than a blown FET. FETs can blow with no smoke or sparks, or with the controller getting warm, but more often than not there is some indication, like a pop or maybe a blown fuse.

I'd double check every connection. As you have a meter, one thing that you can do to see if power is getting to wires with all the connectors made, is to use pins (the dressmaking sort) as probes through the wire insulation and into the copper core. In really stubborn problems I've done this to see if power is really getting to where you think it is. The tiny holes left in the insulation after the pins have been removed don't usually cause a problem, but can be sealed with a smear of Vaseline or maybe sealant.

Checked all wires and connections and they are fine, opened the controller and all wires to the PCB are fine. Ran the test you described, and bingo, yellow phase wires resistance < 1 ohm. Thanks

Appears to be 9 FETS, three for each phase wire, if one blows i assume all three blow as they appear to be in parallel? That is some hefty soldering job, i would need to be holding my soldering iron on that for an hour just to get them out!!!

D8veh, the wires didnt look like they were holding up great in there tbh, could poor connections contribute somehow to the FETS blowing?

So I guess I should buy a new controller then? Could you recommend any?

Thanks again guys

 

[Deleted member 4366]

Checked all wires and connections and they are fine, opened the controller and all wires to the PCB are fine. Ran the test you described, and bingo, yellow phase wires resistance < 1 ohm. Thanks

Appears to be 9 FETS, three for each phase wire, if one blows i assume all three blow as they appear to be in parallel? That is some hefty soldering job, i would need to be holding my soldering iron on that for an hour just to get them out!!!

D8veh, the wires didnt look like they were holding up great in there tbh, could poor connections contribute somehow to the FETS blowing?

So I guess I should buy a new controller then? Could you recommend any?

Thanks again guys

It's very easy to lift the pads when unsoldering FETs, so probably best to get a new controller. I've never had a problem with a KU93 from BMSBattery.
For a FET to go like yours did is a bit unusual. Are you sure something didn't short out in that Maplin connector?

 

[Jeremy]

Checked all wires and connections and they are fine, opened the controller and all wires to the PCB are fine. Ran the test you described, and bingo, yellow phase wires resistance < 1 ohm. Thanks

Appears to be 9 FETS, three for each phase wire, if one blows i assume all three blow as they appear to be in parallel? That is some hefty soldering job, i would need to be holding my soldering iron on that for an hour just to get them out!!!

D8veh, the wires didnt look like they were holding up great in there tbh, could poor connections contribute somehow to the FETS blowing?

So I guess I should buy a new controller then? Could you recommend any?

Thanks again guys

My guess is that running slowly up a hill was the most probable cause, as that's particularly tough on the FETs (and the motor windings). Funnily enough, riding up hills faster is far less arduous on the electronics.

Don't try unsoldering the FETs while they are still on the board, just snip the leads on the duff ones and then you only need to unsolder the tiny wire stubs left in the board, which is a great deal easier. You will also need to remove all the heavy solder build up on the tracks, either using a solder sucker or desoldering wick. I've repaired a few dozen controllers now and haven't yet damaged a board or PCB track, but the secret is to use a powerful soldering iron with a fairly large bit and do the job as quickly as possible.

When fitting new FETs, assemble them to the heat spreader bar first, so they are in exactly the right position, and only then solder the legs to the board. The board then needs to have solder added to build up the trace thickness again.

TBH, it's probably cheaper and easier to get a new controller than to faff around repairing the one you have, unless you're a dab hand at electronic repair.

 

[Geebee]

Totally useless to help the op, but glancing down the threads I read this one as Trouble, shooting a bike
Good luck with the repair.

 

[Jeremy]

Totally useless to help the op, but glancing down the threads I read this one as Trouble, shooting a bike
Good luck with the repair.

Ah! Shades of "Eats shoots and leaves"........................

 

[newguy]

Im 100% sure its the FETs blown, another test i tried was forcing the motor in reverse with and without the controller attached (no battery in either case). And there is significant difference in the resistance when the controller is attached (im not entirely sure how that works, i assume its generating electricity which powers another coil that happens to resist the movement)?!?!?!

Are these FETs easy to get from the likes of maplin? i think i will buy another controller anyway, do you guys just buy from ebay? any recommendations?

 

[OldBob1]

Hi if I was to repair somethink like this I would replace all the FETs. but my recomendation would be a replacement, BMS is one of the sites members seem to use.

 

[Jeremy]

The best bet for FETs is either RS or Farnell, but they do tend to be fairly expensive (the majority of cheap Chinese controllers use very, very cheap knock-off (i.e. counterfeit) FETs, as counterfeiting parts like this is rife there).

Ebay is a bit of a minefield for controllers. There are a few good sellers (people like e-crazyman, Sheijjie Ge, who is excellent) but quite a few dodgy ones. BMS/GBK sell the Wuxi Xichen Micro-Electronic Co controllers (which they both re-badge as "KU" controllers), Greentime sell the same Wuxi Xichen Micro-Electronic Co controllers, but only via Alibaba I believe, and the prices for these are generally lower than for most other "known to be OK" controllers, but postage tends to push prices up. The other well known manufacturer of controllers is Xiechang, which make the controllers that e-crazyman sells via ebay and also those sold by Ed Lyen and cell_man (Paul in Shanghai, who happens to be English). Paul's online shop is now up and running, here:EM3ev

 

[newguy]

I've sent BMS battery an e-mail, as the kit is not even 3 months old. See what they reply with, but I'm not too hopeful. That failing i will go with e-crazyman. Just so i know what im buying is correct, does it matter if the power rating is say 500w for my 350w motor?

Does anyone know why the controllers aren't fitted with fuses? Would have saved me a head ache.

thanks

 

[Deleted member 4366]

Don't waste yoour time emailling BMSBattery. They'll only refund yoou if you send it back, which will cost more than it's worth. Sometimes, they'lll send you a free one with your next order if it's clearly a faulty part, but the probability is that you damged it, so you need to find the cause. Anything that causes a short on a phase wire could cause the FET to blow.

 

[103Alex1]

I've sent BMS battery an e-mail, as the kit is not even 3 months old. See what they reply with, but I'm not too hopeful. That failing i will go with e-crazyman. Just so i know what im buying is correct, does it matter if the power rating is say 500w for my 350w motor?

Just a thought - having seen the shipping prices for orders maybe if someone on the forum is next ordering from China they could add in a new controller for you and distribute on receipt. The shipping cost on single items seems very high pro-rata to the marginal weight on a larger order - and also might get one quicker if they are using courier.

I never thought to ask if anyone needed stuff on my 1st order with BMS but would do so if I order from China again. PayPal means it is easy to pay each other these days with minimal costs using personal payment option.

 

[Gdread]

Hi, I got a controller, throttle and PAS from this company http://electric-hybrid-bikes.com/ the chap that runs it is very helpful his name is Spencer, he also has an eBay shop. Good luck

 

[newguy]

I'm sure your right about BMS.

The timing of the slow hill climb and the motor cutting out is too much of coincidence to ignore i think, the FET must have overheated. The wiring is dead simple on my set up, 3 throttle wires and 3 phase wires and i've went over them all, i dont see much scope for a short to happen. I'll know not to labour the motor that much again.

 

[newguy]

Hi, I got a controller, throttle and PAS from this company electric-hybrid-bikes.com the chap that runs it is very helpful his name is Spencer, he also has an eBay shop. Good luck

Thanks for that link, prices seem good and its a UK supplier! Sent him an email there