shorted sensor repair and phase rewiring (updated: WORKZ)

hillzofvalp

hillzofvalp

100 kW
Joined
Dec 25, 2010
Messages
1,887
Location
Somewhere over the rainbow, Canada
I just hooked up my nine continent 2807 to my 12 FET extreme lyen controller, not knowing that blue and green phases and halls needed to switched. It seemed to work, but I thought it would have more torque. Being my first build, I rode it thinking that it was supposed to be like this. :lol:

Running 14s5p A123 M1 pack.

After 5 minutes of use, the 11.5Ah pack was at half its capacity. The motor quit, and was extremely hot. I could've fried perch on it. The controller was pretty hot as well, but probably acceptable.

I have charged the battery up, tested for continuity between phases.. I get continuity between all phases. I get 4.5V from sense line +/-. I switched the the blue and green wires.

:evil: :evil: :evil: :evil: :evil: :evil: :evil: :evil: :evil:

Please tell me I can fix this easily. Here is a video. I think it was pretty slow.. should've done 25mph but did around 18mph. THE RACK MOUNT IS TEMPORARY (lol).

http://youtu.be/HdaD1AkUjQU
[youtube]http://youtu.be/HdaD1AkUjQU[/youtube]
 
Disconnect everything. It could have fried a lot or it could be a relatively easy fix. The incorrect wiring combo had the motor running at an extremely advanced timing and virtually no torque, so while it did run, it was pushing max current and whatever caused the stoppage could have damaged the motor and/or controller.

A quick check of the controller would be to take the 4 screws off of the rear plate of the controller, the end without wires, and give it a sniff as you open it. It it doesn't smell burnt you're not out of the woods, but a burnt smell is bad. Also check for continuity between the negative battery lead and each phase wire, as well as the positive lead and each phase wire of the controller alone.

To check the motor phase wires, make sure they aren't connected to the controller and not touching each other, and turn the wheel with it still on the bike. It should turn somewhat freely. Then touch 2 of the phase wires together, and it should be harder to turn and somewhat jerky. Then touch all 3 together and it should become even harder to rotate, but smoother.

Report back with results, and in the meantime order yourself an ebike tester from Lyen. Even if it doesn't arrive in time for this problem resolution, it's a tool you need in your arsenal, because it makes diagnosing problems a snap.
 
Open circuits for all phases..

Checked motor by turning it.. feels just as you say. tried different combinations. very smooth with all three... bumpy when in pairs..

I can't really tell if there's a burnt smell. There's a smell... but not sure how to describe it.

I just tested my throttle using a spare M1 3.3V cell.. At idle, the throttle voltage reads .8V.. Full reads about 2.6V. Shouldn't this become 0V no matter what input voltage? Could this be my problem? I will say this: I drillled a hole in the housing to mount a momentary switch for the brake.. I wonder if that ruined it? Pic in a sec
 
I suppose the pressure from braking might've messed up the pot?
6024229895_d0123d3c2f_b.jpg
 
Ok, good your motor phases didn't get so hot that they melted insulation and shorted.

Controllers have a smell, but typically something burnt is obvious when you open it. By open circuit on all phases, did you mean between each power lead and each phase? If no continuity there, then that also is good, because burnt mosfets result in a short in one of those pathways.

Now it comes down to the hardest thing to test without the ebike tester, and that's the motor's hall sensors. I've never burnt a hall, so you'll have to do a search on how to do that, but essentially you provide 5V or so to the red and black to neg and turn the wheel to see if each hall is firing.

WRT wiring, you ran on what I call a false positive combination, ie it spun the wheel in the correct direction but was the incorrect combo. To get the correct one you will need to swap 2 and only 2 phase wires, and then find the right combination of halls (only 5 remaining combos, so it's easier to find the right one than it is to search for a good combo online). Just use a tiny allen wrench or screwdriver to get the male spades out of the motor side hall connect for easily trying different combos of halls. When trying combos always use just small throttle pulses until you have one running nice and smooth. Get the wheel off the ground, and take the chain off if it's a rear motor so the pedals don't whack you if the motor runs in reverse.

When you stopped did you try again after a cool down and recharging the battery? Does your system have a fuse that could have blown? If not, put a fuse on your to-get list too. Do you have some kind of switch on the controller power wire to the brain of the controller, another mandatory safety item to be within reach while riding to shutdown in the event of a runaway condition like a broken or shorted throttle? How about a precharge resistor? 200-300 ohms is good, and that's to prevent the spark when you connect power leads to the controller. I disco mine so rarely that I just charge the caps manually with a resistor as I make the 2nd connection to the battery.
 
Regarding the throttle...maybe, but it should work even at 2.6V, and most of the controllers accept a voltage range that is not 0-5V with say 0.5V being some amount of go, but yes you should get at least something above 4.5V for full throttle. I just use my ebike tester for testing throttles now.
 
Yes. I did do the tests you are listing. The pack fully charged yields the same problem. continuity among all FETs

The issue to me is that the throttle is .8V at rest.. Shouldn't it be 0V? If this were the case, then the controller must have a safety mechanism that detects stuck throttle.. which is probably unlikely.

Now that I think of it. When it died I had just had a little down hill run.. and I tried to regen with the ebrake. I'm not sure that I felt the regen braking all the time... and when it died I remember squeezing the ebrake hard which might've damaged the throttle, since the ebrake button is perpendicular and embedded in it... Does anyone have a picture of the inner workings of a throttle? I'd like to take mine apart but it seems pretty fragile...
 
hillzofvalp said:
The issue to me is that the throttle is .8V at rest.. Shouldn't it be 0V? If this were the case, then the controller must have a safety mechanism that detects stuck throttle.. which is probably unlikely
Click to expand...

I would think so, but not 100% sure. Definitely make sure the wheel is off the ground once you're ready to give things a try in case that's a voltage above the minimum threshold for the controller. These controllers aren't like the more advanced mature industry ones like on a golf cart, which have something called a "high pedal disable" where the controller won't come on while there is some throttle. Ebikes can be at full throttle when you switch them on and things get ugly in a hurry with most controllers.

So yes it's something for you to address, but it's not your no-go problem.

Yours looks like a half twist. I've had a half twist where the end cap touches the twist part and it didn't fully return, so the bike wanted to go when I turned it on. It just so happened that's the only good grip I had on the bike when I went to try to check why it didn't want to fully stop when I parked. I needed video, because when I switched it on I'm holding a 10kw ebike by the throttle and the saddle with the bike bouncing up in the air and coming back down on the spinning wheel at WOT, 70-80mph no load. I finally corralled the bike still at WOT, but wheel off the ground, and yelling for someone to come help me by turning the switch off. Luckily my 5 year old came to the rescue before my arms got too tired, because I couldn't let go with either hand to turn it off.

The moral of that story is make sure the twist part has some space between it and the end cap, because the handlebar end can get bumped pushing the end cap on a bit further, and giving you a sticky throttle. It could even be what's giving you the .8V when you think you're at 0
 
You lost me on the FET continuity. With controller off and disconnected from battery and motor, you should have no continuity between each phase wire and each battery lead. A short there indicates blown FETs. If no short, the controller may be fine, and likely is fine, since the motor phases didn't short, which quickly blows controllers.

There are other tests, requiring opening it up, but I'm not the right person for that guidance, just like I'm not the one to guide you through the testing of halls. Do a search here in the technical section for how to test halls. You're starting with more electronics knowledge than I, so it should be a snap for you.

As part of your process of elimination, unplug everything from the controller except throttle, battery and motor. Brakes, 3speed switch, cruise control, CA, etc. should not be connected. Get down to the bare essentials first.

Does it make any sound at all, or it completely a nothing result when you twist the throttle?

Watch out for bad connections. eg a real hair puller is if one of the end spades slides off to the side or isn't fully clicked in for that motor side hall connector. It can seem like it's plugged, but one spade is off to the side.
 
Hall throttles don't go all the way down to zero or all the way up to supply voltage, unlike a pot throttle, so probably yours is fine.

If the motor "judders" then it is possible it isn't giving position info to the controller, and a hall could be disconnected or dead. The solder could've melted and the wire fallen off if it was hot enough; phase wires unsoldered themselves on one of Dogman's when it was heated like that but it could happen to halls, too.

It's also possible that the wire colors are simply mixed up; unlikely but possible. I thought I had broken my setup for a while, but I had just accidentally mixed up the blue and green halls when redoing the wires inside the motor. :roll:

If the controller, when completely disconnected from everything else, reads essentially open or shorted from each FET pin to either power or ground (except for the ones that are actually soldered directly to the power or ground traces which you can see), there's probably damaged FETs. A 12FET has 4 FETs on each phase, two upper and two lower, so that pairs of them are wired in parallel. Sometimes only one is damaged, but wired in parallel you cant' tell that until they are removed.

Typically FETs in-circuit will read 10Kohm to 100Kohm or more between any pin and power or ground. Blown ones can read short or open, depending on what happened inside.

Hall testing can be done with simply-made circuits, like one Gordo recently posted (whcih I haven't tested myself, but see no reason it shouldn't work):
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=30686
 
You proved it's not the motor phases. It sounds like the controller is still working. That narrows it to a bad connection, bad wiring combo, or a bad hall sensor.

I'd suggest check connections and try the 5 other combos of hall wires. There are 36 total combos of hall/phase wires. With that motor there are 6 valid ones, 3 reverse and 3 forward. For any combo of phase wires, there will be one valid hall combo. When a phase combo gives you a false positive in the forward direction, the good hall combo will be reverse, and swapping 2, only 2 and any two phase wires will result in a good forward combo when you find the correct hall combo to go with it. That's why I said swap 2. You didn't have the right combo to start with, so you haven't proven that you have the right one now. When they put controllers together they can be inconsistent with the wiring, so going by colors doesn't guarantee good results.

It's very easy to just try the other combos by getting the hall spades out of the connector from the motor. There's a little piece of metal that you have to compress to remove them, for which I always use one of the smallest allen keys in my set. I've done it with small alligator clips too, but I find removing the spades easiest.

If you take a few minutes and try the hall combos and still get no good result, then it's pretty likely it's a hall or a hall wire. Gordo came up with an easy to make hall tester from some old stuff laying around. http://endless-sphere.com/forums/viewtopic.php?f=2&t=30686&p=442937&hilit=hall+tester#p442937
 
I just woke up and wenet right to popping open the motor. I remember smelling this very strong dog-bone smell in the insulation around the wire coming out near the motor. It appears liquid orange stuff came through the axle in the the insulation.... Well.. it's not good. the string holding the windings together is burnt.. it's crumbly... :cry:
 
I found a little drip of solder but I'm unsure where it came from... Probably in the white burnt sheaths? I'm not sure if it was facing this way or the other way.. I put it back to show you where I found it. I checked under the sheath to find that there was just a little less solder on one of them.. I now need to figure out how to pop off the other side to check the sensors.

6025250415_74bb739712_b.jpg
 
John in CR said:
Ok, good your motor phases didn't get so hot that they melted insulation and shorted.


Now it comes down to the hardest thing to test without the ebike tester, and that's the motor's hall sensors. I've never burnt a hall, so you'll have to do a search on how to do that, but essentially you provide 5V or so to the red and black to neg and turn the wheel to see if each hall is firing.
Click to expand...

Simple Hall Tester
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=30686

Edit; Sorry, I didn't read all the way through before I posted.

To get the cover off on the other side, just push on the axle with a 3 jaw puller until the cover is up enough to get some wooden wedges under it and then, wedge it off. The uneven gap is because the cover is off.
 
I would have just tested the halls without opening the motor. Then you would have been testing the hall wires at the same time.

Now that you popped the motor open you might as well make it worthwhile and replace the spindly phase wires that 9C uses.

BTW there's no clip to remove. That's only for removing the axle from the stator. The only things holding the magnet backing ring and other cover on are magnetic attraction between the magnets and stator core along with the inner race of the bearing to the axle (which is typically an easy to slip on fit). That magnetic attraction is strong, so be careful.
 
hillzofvalp said:
I am testing halls now.. I put it back together..
Click to expand...

I hope you marked the covers and wheel so you got them clocked as they were assembled. If not, check all the screw holes in the cover line up with those in the wheel.

If you pull the stator out of the wheel, watch the youtube video first so you get an appreciation of the force the stator will snap into the wheel with. VERY HARD ON FINGERS. :shock: :oops: :oops:

Do the LEDs light when you connect them across the battery? Positive to red, negative to black, Hall wires to + side of LED, negative end of LED to negative on battery?
 
I finally got the stator out.. The sensors have a little solder around them. some globs completely loose. One sensor has a glob that shorted out the left two pins (looking down). Is there a way of telling if a sensor is burnt out? Or is it most likely just disconnected?

Would one sensor shorting explain why all three LEDs did not light?

Since I tore it apart, should I install magnet wire? What kind of compound can I get locally to replace the black sealant and the white stuff on the cover edges? Silicone outdoor?
 
hillzofvalp said:
I finally got the stator out.. The sensors have a little solder around them. some globs completely loose. One sensor has a glob that shorted out the left two pins (looking down). Is there a way of telling if a sensor is burnt out? Or is it most likely just disconnected? YES, USE the LED

Would one sensor shorting explain why all three LEDs did not light? YES

Since I tore it apart, should I install magnet wire? What kind of compound can I get locally to replace the black sealant and the white stuff on the cover edges? Silicone outdoor?
Click to expand...
I would not use a type of sealant that hardens unless I used just a tiny amount. It makes it rather hard to open the motor the next time. Sorry on the magnet wire, don't understand the question?
 
You must log in or register to reply here.

Similar threads

ElectricGod
My kick scooter project
15 16 17 18 19
Replies
462
Views
87,291
ElectricGod
ElectricGod
Reid Welch
"Extreme MY1018" archival, closed out
Replies
11
Views
17,912
GIGATT
G
Back
Top