I received my XM-3500li back in the beginning of December 2009. I rode it for a week before disassembling it to modify it with an extra 4 lithium cells, a BMS, a 4.5KW motor and a controller with regen.
It got to be too cold during the winter to really work on it, but now that things are warming up I'm excited to finish this conversion.
Yesterday I removed the rear hub motor wheel and took it down to the local motorcycle shop with my new 4.5kw Kelly hub motor to have them swap the tire and disk brake. Today I tried to mount the new motor on the moped, but I came across an unexpected problem: The original motor's axle has cut-outs on both sides to hold it securely into the frame. The new Kelly motor only has the cut-out on the side where the electrical cables come out, but not on the side with the disk brake.
I've attached pictures of the problem. So what do I do? I have an angle grinder that I could use to cut down the axle, but I'm afraid I might damage the thing.
I've heard of others creating XM-4500's, so how'd you make it work?
Thanks,
Alan
Alan;
Mark the area with a felt pen first and then carefully scribe as you have hi-lighted in the photo. Put the nut, or nuts on the shaft and lock them together to protect the treads if you slip with the grinder. Use the angle grinder to carefully remove the bulk of the material. Leave @ 1/8" and use a file to get the exact profile you need. Get a cheap caliper from Harbor Freight to check your measurements as you go.
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=93293
The alternative is to use a die grinder on the swingarm. How is the 4.5Kw axle stopped from rotating in comparison to the 3.5Kw? With more torque, you must be sure the axle can not rotate and destroy the wires going into the motor.
I have sent you a PM
I've always found that on precision machining on a valuable part someone who really knows what they are doing is worth the money. I'd say find a good machine shop.
IF you do modify the axle, be sure that you get it accurately lined up with the other end! -- Also, be sure that the frame/axel slot on the vehicle is heavy enough to retain the axle without rotation, as this more powerful motor will put much more stress on its mounting. is there any additional bracing, or torque bracket supplied?--If not, use of a quality machine shop would probably be your safest and best bet, for any alterations.-Good luck!-Bob
Robert M. Curry
Hi Alan,
I don't remember encountering this problem. It's possible that the motor axel configurations changed in the interim. It's been a long time since we did our "XM-4500" - I'll check it out at the workshop (likely not until Wednesday though).
John H. Founder of Current Motor Company - opinions on this site belong to me; not to my employer
Remember: " 'lectric for local. diesel for distance" - JTH, Amp Bros || "No Gas.
Mr Blurrycam says sorry, but here is another picture of the Kelly axle.
As you can see, the threaded part of the axle is already keyed to fit into the frame perfectly. It's the smooth, larger-diameter part that extends too far and won't fit between the frame. I could follow the existing shape, so alignment would be fairly easy. What concerns me most is that where the threaded part meets the smooth part, it appears to be two separate pieces.
Another option I suppose is I could bend the frame somehow, but the thing's steel and I don't know how I'd go about doing that (especially keeping the wheel in alignment). I do have some clamps that I can set to expansion mode...
The original motor and the Kelly motor appear practically identical on the outside, except that the original is silver and the Kelly is black. It seems like this should be a drop-in replacement, so I don't see why they changed their axle!
The "smooth part" on the 4500 appears to be a sleeve. Do not bend the frame. I agree with the advice of taking both motors to a machine shop.
Maybe try a sympathetic instructor at a high school or vocational school machine shop? It is a very simple job for a machinist.
As far as alignment goes, all you need is a straight 2 X 4 long enough to reach the front wheel. Set the front wheel straight ahead and put the 2 x 4 against the side of the wheel, as high up as you can, before you hit the bottom of the fender or battery box. Align the rear wheel to the 2 x 4. If you have different sized tires, you need to notch the 2 x 4 one half the difference in size.
I've inspected the wheel again in detail and have gotten closer to a hopefully simple solution. The axle on the side where the wires come out are identical on the stock motor as well as the Kelly motor. No problem there. And looking at the disk-brake side I can see that there is, indeed, a sleeve over the axle. The sleeve turns with the tire, not with the axle and so it really must be removed, and not notched to match the axle.
I've sent an email to Kelly about this, but I thought I also ask here: Is there a way to remove this sleeve?
This is the original wheel:
Notice the shape of the axle below the threads. The cut-out ends about 1/2 inch from the black ring at the base of the wheel.
This is the Kelly motor:
On this, the brass-colored sleeve extends an entire inch from the rest of the wheel.
On a side-note, what is an axle sleeve for?? Because it turns with the tire this would mean that if it were pressed up against the frame it would cause friction!
Good news, everything has been solved. The wheel is now mounted on my bike.
I put a clamp on the sleeve and was able to just pull it off. Easy.
I then measured and marked how much of a cut I had to make in the axle. I just followed the existing flat surface, removing most of the material with an angle grinder and then finishing it off with a metal file.
After some fine tuning it fit in the bike frame!
The existing mounting hardware fit perfectly.
I hear the disk rotor scraping against the pads a little bit, so I'll have to learn how to adjust them. In the meantime I'm relieved everything went so well. It took about an hour to cut, file, and mount.
I am guessing, but I don't think that sleeve should turn with the wheel. The friction from the bearing seal is probably dragging it with the wheel. Without a detailed understanding of how the axle is assembled in the wheel, you are just shooting in the dark and may end up destroying your hubmotor. Ask Kelly for details on the construction. An exploded diagram? Good Luck.
I believe that you will need to shorten and re-install the sleeve! The sleeve fills the gap between the axle and the wheel bearing oil seal, and is likely to also provide security for the wheel bearing to axle positioning. Operating without it may cause bearing damage and/or wheel alingnment and stability problems. The sleeve should fit securely against the inside of the bike frame, and thusly be prevented from rotating, by that friction fit, when the axle is tightened. (Yes, it will turn with the wheel, due to the bearing seal friction, but if you HOLD it, to overcome the seal friction,the wheel will still turn.) The inside of the seal, and the outside of the sleeve, should be lubricated with light grease. Do not damage the surface of the sleeve that goes inside the seal! This area must remain very smooth, or the seal will be damaged, allowing later bearing damage and failure.--Bob
Robert M. Curry
By all means, reinstall the sleeve after modification per Mr. Curry.
JTH/Amp Brothers Electric Cycles/MI
Hopefully, you protected the bearing area from the grinding/filing debris! After the spacer was removed, you need to be certain there is no debris inside the bearing seal area, if there is, use a "Q tip" or similar, coated with oil/grease, to clean and remove this contamination, and re-insert the sleeve, after shortening it to provide proper fit. This sleeve is critical to maintaining proper bearing spacing and wheel position on the axle--that may resolve your "Dragging brake" problem. (Do this spacer shortening carefully!-You may be best to seek machine shop assistance)--Good luck, Bob Curry
Robert M. Curry
I thought I saw dust from axle filings on the wheel in the picture, but wondered if it was just "Mr. Blurycam?"
As far as I know, Kelly doesn't produce the motor.
I made the modifications to the axle and everything has been working great for the past few months. I think if I were to do it again, though, I would not choose the high-speed motor. I can get it to 55MPH with room left in the throttle, but when carrying two people (270 lbs combined) acceleration from a stop on level ground is slow. This makes it scary when you're at a stop sign and you have to turn onto a high traffic road and accelerate quickly to not be hit from behind.
Does anyone know of a place in Oregon that might be able to rewind my hub motor for more torque?
edit-
I should also mention that my Cycle Analyst is showing that I get around 45 to 50 watt-hours per mile at 25MPH with a 150lb load (me).
Wow great job on the conversion. I am wondering about this last statement. How did you connect the magnet to the wheel to get the speedometer to work? I just ordered the Cycle Analyst. I check my front wheel and I don't see anyplace to connect the magnet. I have the XM-4000, so should be the same as your XM-3500.
You can carefully remove a little of the coating on the wheel and use epoxy to mount the magnet. Even silicon will work as and adhesive if you leave it for a few days before riding.
Sorry, but how did you mount the CA console? I assume that the XM-3500 has the same handle bar cover, plastic housing.
Get a bigger controller instead.
it will cost about the same, but give you way better performance than a rewind.
I would suggest a 600A controller.
Matt
Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km
I had some neodymium magnets that I got from DealExtreme. I believe they're 10mm by 1mm disks. My wife and I had used them to make magnets for our fridge. We got this great epoxy-like stuff called Hodge Podge from the local craft store. It's what I used to attach the magnet.
Here's the kicker, though. When you use the Cycle Analyst with a BLDC motor like we have in our scooters, you don't really need to go the magnet route. You can hook up one of the hall sensor wires coming from the motor directly to the CA. Because there's no gears, you can calculate speed by the number of times the sensors in the motor fire. I should have done this (although I've had zero problems with the magnet route), but just hook up the CA sense wire to one of the *thin* yellow, green, or blue wires between the controller and motor.
I got the large-format display Cycle Analyst that you have to enclose yourself. I decided that the CA is much more accurate and entirely replaces the functionality of the analog gauges in the dashboard, so I remade the dash.
I do 3D animation, so I used calipers to measure the acrylic housing for the dashboard, and then I designed a replacement insert. I submitted my design to a website called Big Blue Saw and they were able to cut my template out of aluminum. If you're curious about the structure I can take pictures the next time I have the dash apart.
600A?! I'm not looking to do wheelies here, and it would probably instantly fry my motor. I was thinking a 50% increase in torque by going from 200 to 300A would be good.
The whole problem is that Kelly will not tell me what the motor is rated for, besides "4500 Watts". It is supposedly a 72V motor, so does that make it 62.5A continuous? But it's not watts or volts that kill a motor, it's the amps. Amps cause heating which can melt the windings' lamination and cause an internal short.
I didn't realize a few things when I got into electric vehicles about a year ago, so I'll repeat it here for the sake of other people like me.
yes it would mean the rated current of the motor is likely to be 62.5A continuous.
it will be higher for 30 mins.
the tricky thing is, controllers don't have alot of thermal mass, the way motors do.
so if you buy a 300A controller, beyond 2 minutes, its only good for ~100A.
at that point your controller is in thermal limit, so it becomes a 100A peak controller until it cools.
thats no good if your motor can take 100A for 15 minutes.
a 400A controller would be a closer match, still giving you more peak power.
you can always program a big controller to back off, but you can't make a small controller act like a big one.
Matt
Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km
It looks like I have two main options for upgrading my KBL72201:
The KBL series has a can-bus port that I'm not using in my current setup, while the cheaper KEB series does not. I really only need the power while accelerating as I don't really ride on hills. Therefore, even with its shorter duration maximum output, I'm leaning towards the KEB controller.
Because the Kelly Controller bolt pattern is different than the stock controller I made an adapter plate out of stainless steel. About a month ago I read that you should mount the controller onto a heat sink. Stainless steel, especially when it's painted, is a horrible conductor of heat. So I'm going to redo my adapter plate out of maybe quarter-inch aluminum and use some kind of thermal paste between the two.