Kn@ckered controller

[peerjay56]

Being my usual bull in a china shop self, I have managed burn out the controller on my wife's urban mover um44sl. I was trying to fit a throttle (on the first day of a week in Derbyshire to enjoy some cycle routes:$). I simply plugged the throttle into the spare connector (it is a throttle specifically for this bike). When I switched the battery on, I got smoke and fizzing from the throttle.
Now the pedelec setting doesn't work. All lights show on the display, there's just no action:-/
I have had the controller apart, and there is an obviously fried resistor, but I assume that other components are also likely to be jiggered.



Will I be able to repair/replace the controller, and will the throttle be fixable (I don't really want to lash out on another)?
I probably aren't going to be able to sort it for now, due to lack of resources, but I would appreciate some advice for repair/replacement.

Sent from my LT26i using Tapatalk 4

 

[John F]

Being my usual bull in a china shop self, I have managed burn out the controller on my wife's urban mover um44sl. I was trying to fit a throttle (on the first day of a week in Derbyshire to enjoy some cycle routes:$). I simply plugged the throttle into the spare connector (it is a throttle specifically for this bike). When I switched the battery on, I got smoke and fizzing from the throttle.
Now the pedelec setting doesn't work. All lights show on the display, there's just no action:-/
I have had the controller apart, and there is an obviously fried resistor, but I assume that other components are also likely to be jiggered.
View attachment 6892
View attachment 6893
View attachment 6894
Will I be able to repair/replace the controller, and will the throttle be fixable (I don't really want to lash out on another)?
I probably aren't going to be able to sort it for now, due to lack of resources, but I would appreciate some advice for repair/replacement.

Sent from my LT26i using Tapatalk 4

Tough luck Phil - hope you get it fixed for your visit to the Flatlands
Off out tonight for a moonlit ride

 

[Cottonpickers]

Does that say D3 next to it? Likely a diode if it is. Might be worth trying a replacement - often the same components are used on boards elsewhere so you may be able to get details off another one. If it's a low voltage one I may have one here

 

[Deleted member 4366]

I can't see how you can burn that resistor by plugging in the throttle. It might just appear burnt because they can run quite hot - especially if you over-volt them. Were you using a 36v battery? To me, it sounds like you've connected the 36v wire for the throttle LEDs to the throttle hall sensor, which will blow it. It can then feed 36v back to the 5v regulator, which will blow that and every 5v device, which includes the CPU and pedal sensor, so I wouldn't bet on being able to repair it.

The power comes in at 36v goes through that resistor to the 12v regulator. The 12V branches to the 5v regulator, which powers the CPU and the sensors. Follow that path and check whether you get the 12v (could be a bit higher) and the 5v. You can get a 5v regulator from Maplin for not a lot. If you have the 12v, but not the 5v, it might be worth replacing the 5v regulator to see what happens. It's a 78L05

The 12v regulator is the upright component with a heatsink bottom right of the middle picture. The 5v regulator is the black D-shaped three-legged one next to it. The middle leg is 0v, one outer leg is input, the other output. The 12v one is the same, so follow the volts from 36v into the 12v one, and the 12v out of the 12v one into the 5v one, etc. Keep your black probe on the battery negative.

 

[trex]

The resistor that burnt: I think it's a 150 Ohm 2W resistor that feeds the 14V regulator from 36V rail. The 14V output feeds the 5V regulator, which is shorted by the throttle - reverse polarity. With a little bit of luck, replacing the resistor is enough to get the controller going, the throttle will have to be replaced.

 

[peerjay56]

OK,
Thanks for the suggestions so far. I'll see if I can put some gear together to try out the things suggested.
There are no leds on the throttle I was supplied Dave. Because i was in a rush, and foolishly thought as it came from Urban Mover it would 'just work', I failed to take even the most basic of precautions.
I'll let you know how I get on.

 

[Deleted member 4366]

The throttle wires only have 5v on them. If no 36v was involved, the worst thing you've done is short the 5v regulator. I'm not sure that's possible regardless of mis-wiring combinations. It's possible that a wire came off in the throttle and made a dead short. Whatever, you have to do the same anyway: Check the 12v and 5v regulators. Hopefully about 50p will get you going again.

 

[peerjay56]

Well I've managed to borrow a multimeter and some leads, and it's bad I have indeed managed to send 36v to the throttle. A 4pin connector - 4 wires in, three out. 1 of those in was 36v and I managed to send it to the throttle, thanks to not checking connectivity. There is no response to probes with the multimeter beyond the first capacitor (the big one in the second picture), which seems to be the next component in the circuit from the 36v in. The fried resistor (150 ohm) is showing 200 ohm resistance, presumably as a result of being fried.
I still had all 'normal' leds at the display. Does this mean the display is ok? Also, is there a test I can do on the torque sensor to confirm it's viability? Will the brake sensors also be fried? If those parts work, I'll consider ordering a new controller from Urban mover. If not, I am thinking about refitting the bike with the alien ocean kit that I now have lying around - it would mean swapping the torque sensor for a pedelec sensor, I presume, and also presumably swapping the display.
Meantime, I shall be spending the week riding my wife's non-operational ebike, whilst she rides my claud butler
We have a patchy btwifi signal where we are staying, poached from a neighbouring cottage, and no mobile connection at all, so no response from me to any suggestions anyone may make is not me sulking; it'll be me on holiday, pedalling hard!

 

[Deleted member 4366]

The 200 ohm you measured doesn't mean the resistor is faulty. It can be a measuring error because your meter hasn't been calibrated. It's in the right range, so is probably OK. You need to check what's on each side of the 12v regulator.

The brake sensors are normally switches, so should be OK. Some torque sensors are 12v and some are 5v. You need to follow the power wire (normally red) back to see whether it connects to a track that goes to the 5v or 12v, then you can test it by giving a supply from a 5v charger or 12v battery to test the signal wire. I just remembered, some torque sensors have two 5v signal wires. How many wires does yours have at the connector?

 

[peerjay56]

Ok, thanks. It will all have to wait till I get home. I simply don't have the resources here. Plus, I'm on holiday so I'm going to leave it alone for now. Off pedalling!

 

[peerjay56]

Ok, thanks. It will all have to wait till I get home. I simply don't have the resources here. Plus, I'm on holiday so I'm going to leave it alone for now. Off pedalling!

I lied! I had another mess last night
Firstly, the torque sensor has three wires running to it, from a three way connector.
I've managed to get some readings at various points in the controller.
I don't really want to go prodding about blindly without some further guidance.
On the large resistor: 41.7v in 9.0v out. Running very hot, (as you said, Dave)
On the 12v regulator? (far right with seperate heatsink):
P1 9.0v }
N 7.5v } Getting very hot (May be normal, no doubt why it has a heatsink!)
P2 6.8v }
Whilst being hot may be normal, I'm guessing the pin readings are not. Shouldn't N be 0v?
On the small D-shaped one next to it:
P1 2.3v
N 0.00v
P2 0.3v
In both cases, I've taken P1 as being the one nearest the end of the board, as viewed in the side on picture. I've assumed they are PNP, rather than NPN.
I've not tested beyond that, other than to check 3 easily accessible diodes on the back of the board, where the connections for the throttle are. Using the multimeter I've borrowed, which has a diode test setting, two are showing 1.02v, whilst the third shows 0v.
Whether that means anything, I have no idea!

 

[Deleted member 4366]

Did you test with the throttle and torque sensor connected? If so, test again without them. Those results show that there's a dead short somewhere. It could be the 5v regulator, or, if your throttle or torque sensor is connected, it could be one or both of them, so it's over-loading the regulators. I still think it's worth having a go at replacing the 5v regulator if it's faulty.

 

[ghost]

Warning... ''e-bike Ray Mears'' at work! This is for fun but I would probably do it because I am a mad hero type who hates to lose.

What I do know is if I was on holiday wanting to e-bike without a spare controller or soldering iron or anything I would probably be mad enough to do the following assuming I had some spare time.

Firsly if its a brushed motor I would ditch the controller and run it direct with caution and a huge makeshift switch probably using a mousetrap or something! Assuming not...

If the bike was non functional electrically then there is nothing much to lose there by playing as long as you can know what you are doing enough to avoid being a numpty and frying it more.

Give up on throttle dream until you get home, try to fix it back to how it was to get you going. If you have ever driven in the outback you may understand how you have to think when you dont have a shop to go to.

If one assumes the resistor is lunched then its got to come out anyway so I would crunch it with pliers to salvage the maximum tail length, snip it in the middle to leave the longest tails to attach to possible.

As you have a multimeter you need to build an external remote resistor! Either salvage some electrical wire from somewhere or design it around using the probe wires from the miltimeter although Ideally they need to be long enough to reach your bike bag. Alternatively in a pinch I have used plastic coated garden wire, (quite good actually because you can pre-form it, - think twin & earth) or use some wire borrowed from the bike like the dynamo lighting wire.
Anyway using the multimeter you just need something with the correct resistance. Either an actual resistor borrowed from something in the car or bike light or get inventive with the garden wire reel, or the element from an old toaster snipped down, or try using clean white buff (not shiny whatever its called) card and making blobs and tracks (or a patch) with a pencil lead and probing the patch/line to get the resistance you need, then insert he wires through the card and tape them, then cover the patch with tape. Basically you need to be MacGyver.
To join to the controller I would twist or bind the wires together using strands of wire first the best I could and seal by setting fire to the end of a cable tie and dripping the molten plastic onto the joint, this insulates it & holds it in place. Any hard plastic will do just avoid breathing it in, works better in long strips. You can usually get it off later on quite easily. Candle wax doesn't work its too soft!
Obviously if its more than just a resistor gone then you are fibered but I would blooody love having a go at that! Sometimes I wish I was worse prepared when I travel and that things would break more often so I can actually use some of my crazy repair skills!

 

[peerjay56]

Thanks Ghost!
I did indeed give up on the throttle, in the hope I could restore it to 'before j!ggered' mode. I like your inventiveness. I guess living in a place where things are more readily easily available has removed the necessity to be inventive

 

[peerjay56]

Did you test with the throttle and torque sensor connected? If so, test again without them. Those results show that there's a dead short somewhere. It could be the 5v regulator, or, if your throttle or torque sensor is connected, it could be one or both of them, so it's over-loading the regulators. I still think it's worth having a go at replacing the 5v regulator if it's faulty.

I've had a more comprehensive look around now. I've tested the 5v regulator alone, and it's OK (12v P1, 0v N and 4.8v P2). Neither the throttle or torque sensor were connected when I did the earlier tests. The hall sensor in the throttle has completely vaporised, and a seperate test of the torque sensor seems to confirm it is still working (12v supply, 0v on black lead and between 0.68v - 4.10v on the blue lead, when I apply a load to the crank). The pcb in the controller has small additional boards placed on top, and following the route of the 5v supply is just not possible, as far as I can see, and I can't see another obvious short. The only other component which I am certain is faulty is one of three zener diodes which shows 0.001v in both directions (the other two - apparently the same, read about 1.03v in one direction only). Unfortunately the diodes don't have a marking on them. All i can say with certainty is that they are not IN0041's (I have some of those, and they read about 0.8v in one direction only when tested on the multimeter).
So, I think I can say the torque sensor is OK, the display is functional, the brake switches should be OK. On this basis I think I will try to get a new controller from Urban Mover. Unless anyone has a better idea?

 

[peerjay56]

Just a thank you to all those who tried to help me fix my foolishness last week, in particular to D8veh, who tried to improve my electronics capabilities. Sadly, I was found wanting
I've fitted a new controller (paying particular attention to connections and polarity) and all is now working. I'm not trying to fit a throttle again just yet! Assembling electronic kits is one thing, but trouble shooting some complex circuitry is a whole different ballgame.
The only point with the new controller (which shows a date of manufacture of 7/2009!) is that the brake switch connectors have three wires, rather than two as in the old ones. I've used the old 2way jst connectors and just connected the 0v and 4.4v wires (white and yellow) as per the original, and isolated the 4.8v red wires. All seems to be working as it should.