PDA

View Full Version : Hub motor repair



teddillard
16 October 2010, 1248
I just started tearing down the hub motor on the X-Treme scooter I salvaged, here's the first phase of the repair. I say repair with hope, and pure thoughts. Unicorns too, I have unicorns.

I'm lazy right now, so here are a few teaser photos. There are more on the post, and I might update this post with more as they come.

http://evmc2.wordpress.com/2010/10/16/inside-a-hub-motor/

http://evmc2.files.wordpress.com/2010/10/picture-51.png?w=628&h=416

http://evmc2.files.wordpress.com/2010/10/picture-11.png?w=631&h=415

http://evmc2.files.wordpress.com/2010/10/picture-22.png?w=634&h=416

larryrose11
16 October 2010, 1357
Is that a PM Motor? What is wrong with it? Can you borrow an inverter to test it out?

teddillard
16 October 2010, 1401
Yes, the wires feeding into it are frayed, and he tried to solder them but they were exposed. I'm going to start by remaking the wires, I'm not sure how to test the Hall sensors, but there's an Instructable post that talks about it.

http://www.instructables.com/id/Electric-Bike-Hub-Motor---How-to-Replace-a-Hall-ef/

Oh. I see there's no real way to test them... I guess it's easier to replace all three.

What's an inverter?

magicsmoke
16 October 2010, 1625
Oh. I see there's no real way to test them... I guess it's easier to replace all three.


Dead easy to test them Ted. Wire a PP3 9V battery across the Red/Blk (pos/neg) and with your DC volt meter measure across Blu/Blk, Yel/Blk and Grn/Blk when you waggle a fridge magnet or whatever over the relevant sensor. The reading should go from high to low or low to high depending on the polarity of the sensor. If you don't see a change, they may be 'current output' devices. You can check these by putting a resistor (1K to 47K range) in series with your meter probes. Set your meter to DC milliamps and probe between Blu/Red, Yel/Red and Grn/Red when waggling.
Caveat, some sensors only good for 5V so PP3 (9V) might be too much, but unlikely.
Also, can you check with a meter the value of those resistors. I see they say 470, but in normal surface mount world that means 47 ohms i.e. 47 and no zeros. Seems very low to me, I'd expect 471 at least i.e. 47 and one zero .. 470 ohms.

Rob

teddillard
16 October 2010, 1644
how do you know this ****? lol!

teddillard
16 October 2010, 1701
wait, I should ask, how the **** do you know this ****?

harlan
16 October 2010, 1716
Hacked my magura throttle with one of those hall effect sensors. Throttle response is much smoother than with the POS trimmer that it had before.

484

485

486

BaldBruce
16 October 2010, 1723
wait, I should ask, how the **** do you know this ****?

Because he's Rob and you're not......:D

teddillard
17 October 2010, 0426
Because he's Rob and you're not......:D

I wanna be Rob when I grow up.

magicsmoke
17 October 2010, 0509
how do you know this ****? lol!

It's my job Ted, it's how I eat.
If I was paid for my photography Knowledge I'd be on elsoupkitchen.net :)

Harlan, that looks like a cool mod. Any chance you could detail it in a thread?

Rob

teddillard
17 October 2010, 0601
OK, fixed the wires.

Before:
http://evmc2.files.wordpress.com/2010/10/picture-12.png?w=634&h=420

After:
http://evmc2.files.wordpress.com/2010/10/picture-23.png?w=633&h=418

http://evmc2.files.wordpress.com/2010/10/picture-32.png?w=628&h=417

Now to decide if I want to test the Hall sensors or just slap it together again and see if it works. But how would I grow up to be like Rob, if I don't try to tamper with the elemental forces of the Universe?

teddillard
17 October 2010, 0716
Update. Tried to test the Hall sensors, there's no change on any of them using method 1. For method 2, (the 'current output' method suggested by the "no change" result) I don't have the required resistors, and my VOM is acting wonky, so I put it back together and tried it out. Deadness.

Before I first took it apart, I tried it one last time to do a little video for the how-to. It ran once, but when I tried it a second time it didn't move. I'm wondering at this point if the controller fried when the wires shorted out. I had no idea they were actually bare at the time. Crap.

I'm guessing the advice would be to get the resistors, test the sensors properly, then if they test out replace the controller...

DRZ400
17 October 2010, 1531
How's the controller? If you're going to replace it there's plenty available now that are <hall> sensorless. I used a sensorless 48v unit on my ebike. From Ebay, it was a 48v 30a unit thats worked flawlessly....$25.00! You just put it in learn mode and spin the hub...it then knows which phase to fire.

teddillard
17 October 2010, 1726
Seriously? Do you have a link or model?

...wait, this doesn't involve aluminum foil, does it? :D

larryrose11
17 October 2010, 1853
Ted, an inverter converts DC, like your battery, to AC. This power converter is necessary to drive an AC or a brushless motor.

teddillard
18 October 2010, 0305
oHHHH, THAT kind of inverter!

teddillard
23 October 2010, 0704
http://evmc2.files.wordpress.com/2010/10/picture-14.png?w=640&h=421

So, I decided to replace the Hall sensors since I was getting some wonky results of the tests, and don't feel like buying a new VOM in spite of the fact that I've been very very good and deserve it. At least that's my story. The sensors are marked with S41 814. Or maybe 641 814, but I think it's an S. I can't find a reference to it anywhere, can anyone point me to a replacement? And stop laughing at me Rob. :D

DaveAK
23 October 2010, 0956
I'm no expert but maybe one of these would do:

http://www.mouser.com/Search/Refine.aspx?Keyword=S41+hall+effect+sensor

(Just searched on S41 instead of the whole thing).

Here's a Google search on the same thing:

http://www.google.com/custom?hl=en&safe=active&client=pub-3794288947762788&cof=FORID%3A1%3BGL%3A1%3BLBGC%3A336699%3BLC%3A%230 000ff%3BVLC%3A%23663399%3BGFNT%3A%230000ff%3BGIMP% 3A%230000ff%3BDIV%3A%23336699%3B&channel=1975384696&q=S41+hall+effect+sensor&btnG=Search

teddillard
27 October 2010, 0335
Thanks, Dave!
Update- Mr. Wizard (David O'Brien, Master of Space and Time) came over last night with a working VOM, (and more than a clue of how to use it... unlike me. :D) and we ran through the tests that Rob suggested and a few that David wanted to do, and we got no response on any of the sensors whatsoever. What concerns him is that all three look like they failed at once- kind of bizarre, from his opinion. We've narrowed the part down to the Honeywell SS41 sensor at around $5 each so I'm ordering them from Digikey... will update.

In the process of all this (mostly while he was poking and waggling, and I was daydreaming) I came up with the BRILLIANTLY crackpot idea to mount two motors like this on the same axle- making one double-wide hub motor. I know, I know... this scares me as much as I'm sure it does you... Tony, you in?

DRZ400
27 October 2010, 1321
Here is the controller I bought....hall sensor or sensorless works great. Brake, cruise control, rpm out. Several guys on ES use them. Some have swapped out the 12 drive fets to IRF units (low (RDSon) to boost the output.

http://cgi.ebay.com/ebaymotors/500W-E-Bike-Brushless-Hub-Motor-Hall-Sensor-Controller-/120472346149?pt=Motors_Car_Truck_Parts_Accessories&hash=item1c0cb61e25

teddillard
27 October 2010, 1401
Here is the controller I bought....hall sensor or sensorless works great. Brake, cruise control, rpm out. Several guys on ES use them. Some have swapped out the 12 drive fets to IRF units (low (RDSon) to boost the output.

http://cgi.ebay.com/ebaymotors/500W-E-Bike-Brushless-Hub-Motor-Hall-Sensor-Controller-/120472346149?pt=Motors_Car_Truck_Parts_Accessories&hash=item1c0cb61e25

I'm really confused. These will work on a motor that's designed for Hall sensors, but without the Hall sensors? How does it work?

DaveAK
27 October 2010, 1520
$5 a piece? You could get all three from mouser.com for less than that, (see my link). Also from DigiKey I would have thought. Maybe you're going for extra special ones. :D

teddillard
27 October 2010, 1633
yeah David emailed me today with a price around $1.50- more like it. The $5 price was Digikey I think. I'm going to update tomorrow though, he set up a test pad and we tested them all outside of the motor- they tested out fine. Next dissection is the controller...

511

update: Basically David was not confident that we were getting good test results the way we were trying to do it. Sitting on the couch, holding several leads together, resistors in series and all that. He put together the little test rig shown above and tried it out with some sensors he had lying around, and it was really easy to see what was going on. I'm honestly not sure exactly how it is set up, but there are resistors in the circuit and the LED apparently provides some load. If anyone's interested maybe I can get him to shoot me a diagram, but I think you can figure it out from what Rob posted above.

The end result is that they are "latching" sensors- they turn on and stay on until they're turned off - and all three work fine.

BaldBruce
27 October 2010, 2001
I'm really confused. These will work on a motor that's designed for Hall sensors, but without the Hall sensors? How does it work?

Sensorless inverters work by infering the position from the voltage signal directly off the drive wires rather than seperate sensors. They either look at the third leg while driving the other two if it is a two phase at a time type or they can measure the back emf on all three to infer the position of the rotor. This type of controller could care less if you originally had sensors, the more compicated circuit internal to the controller makes the sensor unnecesary.

teddillard
28 October 2010, 0336
Sensorless inverters work by infering the position from the voltage signal directly off the drive wires rather than seperate sensors. They either look at the third leg while driving the other two if it is a two phase at a time type or they can measure the back emf on all three to infer the position of the rotor. This type of controller could care less if you originally had sensors, the more compicated circuit internal to the controller makes the sensor unnecesary.

Wow, cool. I poked around ES a bit and read up on them... seems like a great option. I'm wondering about the China ordering on Ebay, though- last item I ordered from China took 6 months to arrive. Did you have better luck, DRZ400? (Shipping is almost as much as the item, too- $24- at $50 total or so, it's almost the same price as the stock replacement...)

I also noticed on one site it says the sensored version of a controller was "instant start" where the sensorless was "pedal first". Any info on that? Are all sensorless needing to be spun up first, or was that just that particular model?

Sounds like it'd be a fun little twist on the project...

SplinterOz
28 October 2010, 0415
Sensorless controllers have to guess he position of the motor if the motor is stationary. Most sensorless controllers will "shudder" a little at startup as they determine which phase to power to get the motor moving. Also they also they tend to blow the power circuit if the motor does not move immediately.

DRZ400
28 October 2010, 0624
I tried mine both with and without sensors....just as smooth either way. I've used sensorless for over 9 months now....ALL of my 10 R/C brushless planes and helis are sensorless as well....nothing new really as hall sensors have not been used in over 10 years in RC. I got mine in 5 days from China, I think it was fed-x. It works very well and is dirt cheap.....with a simple FET upgrade you could easily double the power handling if needed.

SplinterOz
28 October 2010, 1430
I know RC planes and Heli's use them all the time. They are ideal for that for a few reasons. One of the big ones is that they have virtually no load on the motor at start up. However EV's have most load at startup.

I have heard that the newer sensorless controllers are smother and I am getting one to test for my build as soon as they build one that can do 72volts and 200amps for me. :)

teddillard
28 October 2010, 1621
However EV's have most load at startup.

Especially when you're doing burnouts... :D

Since the sensors seem OK, I'm tearing into the controller. Here's the guts:

http://evmc2.files.wordpress.com/2010/10/picture-43.png?w=640&h=422

I don't see anything melty or anything, but Mr. Wizard might...

spassovm
09 November 2010, 1131
Hi Rob,
I was really impressed by your experience and knowlege. So I am encouraged to ask you about my bike. I bought it from China a week ago. it is 72 V 1500 wats .It worked well... for one day. I tried to drain the batteries to test the limits of the scooter, and after the voltage felt under 62 V the motor stopped, and never started again. I changed all the fuses, I ordered a new contrller, but nothing happens. The voltmeter on the screen reads 79 Volts, the lights, horn are OK, but the motor only slightli ckliks when i twist the throttle. Please advise.

It's my job Ted, it's how I eat.
If I was paid for my photography Knowledge I'd be on elsoupkitchen.net :)

Harlan, that looks like a cool mod. Any chance you could detail it in a thread?

Rob

teddillard
09 November 2010, 1151
I noticed your question on my blog too- I'm going to answer it here. My advice would be to get the manufacturer to repair it under warranty.

teddillard
14 November 2010, 0823
Well, duh. After all that, I realized that I could use the test method Rob described with everything in place. Power the scooter, it will power the sensors, then I just test across the ground and lead to each sensor.

Here you go:

<object width="480" height="385"><param name="movie" value="http://www.youtube.com/v/efYFOHnXh0E?fs=1&amp;hl=en_US"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/efYFOHnXh0E?fs=1&amp;hl=en_US" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="480" height="385"></embed></object>

EarlyEVer
14 November 2016, 1037
I was wondering what XM that this motor is off of? 5000Li, 3500Li, 3100? I own the XM-5000Li and I have over 28,000 miles on the bike. However, at this time the lights do not work and the connection from the controller to the DC has no power coming to it. So I am very interested in your post here. I just found it and I have not read the entire list of comments. However, I was wondering if you ever finished your repair?

Tony.



I just started tearing down the hub motor on the X-Treme scooter I salvaged, here's the first phase of the repair. I say repair with hope, and pure thoughts. Unicorns too, I have unicorns.

I'm lazy right now, so here are a few teaser photos. There are more on the post, and I might update this post with more as they come.

http://evmc2.wordpress.com/2010/10/16/inside-a-hub-motor/

http://evmc2.files.wordpress.com/2010/10/picture-51.png?w=628&h=416

http://evmc2.files.wordpress.com/2010/10/picture-11.png?w=631&h=415

http://evmc2.files.wordpress.com/2010/10/picture-22.png?w=634&h=416