Converting a hubmotor to a middrivemotor

[bØb]

Oh OK I need yo get mine apart, its just too hard to visualize with only pics...

In a nut shell:

1> Run the motor before you take it apart and void the warranty :wink: You can hold the motor vertically or horizontally in a vise by the brake axle flats.
2> Remove nuts and etc from both ends of the axle. Clamp the brake side axle vertically in a vise.
3> Remove the freewheel side cover.
4> Push the hub and brake side cover and hub center section off the motor guts. Do this with your hands. Leave these pieces on the axle on top of the vise.
5> Remove the Case and guts from the vise. Slide the hub and cover down to the end. Don't try to take it off, just lay it on the bench (the one the vise is attached to
6> Put the guts back in the vise, with the brake side axle up.
7> Remove the circlip, clutch, key and the washer. The clutch is not real tight, but is hard to get a hold of. You may need a 3 leg gear puller, but you can turn the gear puller screw with you hand. If you can't, you probably forgot something Slide the clutch and stuff down the wire too.
8> Take the guts out of the vise again and put the brake side axle in the vise again. I know this seems like a lot of flipping, but with my 20 20 hindsight, I can see it is the easiest method.
9> Cut all the wires in the cable now. Be sure to leave a little bit of wire on all the connections so you can identify them later. You are now committed
10> Remove the circlip and six screws that hold the little cover over the wires and key the stator holder to the axle. You may have to remove some excess epoxy and bend some wires a little. The epoxy they used seem to be super quick setting and is quite brittle. When you remove these screws, there is nothing supporting the motor. Drop it down the axle and set it on a rag wrapped around the axle. If it doesn't fall free, bump the stator with you hand or rock it a little. The little plate will be held by the key.
11> Remove the little plate and the key. You might be able to pull it off by hand, if you can get hold of it. If not use the gear puller. If it pulls hard you forgot the circlip in step 10.
12> Cut the end of the cable stub flush to the axle and slid the motor off the axle. It is a good idea to keep the motor in a container to keep it from grabbing up all kinds of steel trash in the shop. I kept mine in a small plastic paint pail with a lid on it except when I was measuring something on it.
13> Clamp the freewheel side of the axle horizontally by the flats and jerk the big end of the cable out of the axle. I'm not going to use the phase wires out of mine. I have 10 ga. silicone wire to use for that, but I will probably use the hall effect sensor wires and temperature wire and put them in a heat shrink tube.

I'm ready to go to bed,
bØb

 

[Whiplash]

Awesome THX!

 

[mr.electric]

Since you have to cut the wires during disassembly I took the opportunity to run thicker phase wires. I used two pieces of the stock phase wire for each phase lead just to avoid the dreaded melted phase wire.
I like the b0b method of drilling the holes. I wonder if a second stator center cover plate could act as a ready made heat bridge.
Gluing the entire stator - side cover assembly would be good. The old Mac scooter motors had the stator permanently attached to the motor face plate with some kind of black epoxy.
Alternatively one could make an entire new side plate that incorporated a heat bridge and a frame mount in one Cnc chunk. It could have a mount for little plastic chain guard too. Unfortunately It would be expensive to make such a piece.

 

[crossbreak]

Happy new year to all our converters

This may be a little OT, but i think this has to be said once here:

Gearing a middrive for good speed and hill climb isn't simple as we all know. The spread is dependent on the motor torque, and thus we have determined a max speed of 35mph we can calculate gear spread for a given motor power (if geared/reduced right). I made some calcs about a bike with a weight of 150kg including rider. What came out is this rough contract:

1kw: 400% spread needed
2kw: 200%
4kw: 100%=single speed.

It's all about torque at the rear wheel. One can not push infinite power into it without flipping over (As can be demonstrated spectacularly on one of LiveForPhysics bikes for example) My next bike will be single speed, too

 

[Whiplash]

Don't forget though for off road purposes gearing very low is for climbing slowly without overheating the motor. My 6t Mac in the wheel would climb very well but if you tried it at a low speed for any length of time it would melt! The gearing is for more than just acceleration... Also, it HUGELY reduces power consumption in off road or even street hill climbing...

 

[crossbreak]

The 6T MAC is only rated for 500Watts. If you only push in the rated 500Watts, it wont overheat, it also wont climb a steep hill as can be seen from my simple look-up table We should not compare a 4KW Turnigy 80-100 with a MAC, they play in complete different leagues. A MAC has 1:5 reduction, an 80-100 needs 1:15 reduction

 

[mr.electric]

The 6T MAC is only rated for 500Watts. If you only push in the rated 500Watts, it wont overheat, it also wont climb a steep hill as can be seen from my simple look-up table We should not compare a 4KW Turnigy 80-100 with a MAC, they play in complete different leagues. A MAC has 1:5 reduction, an 80-100 needs 1:15 reduction

Well said.
I think we should be happy to get away with 1000 watts continuous from these little motors since they are only rated at 500 watts anyways.
I ran my bmc at 1250 watts for many a mile when it was in the wheel and it was still fun. Friends and I experimented extensively with over 1250 watts and invariably something went wrong - melted wires, dead controllers ect.
I think the purpose of running through the gears for most builders is to be be able to do wheelies and spin the back tire with the front brake locked which should work at 1000 watts. Climbing steep long grades slowly is only important when riding off roading technical trails. A Mac/ bmc in the wheel will still climb almost any hill here in San Francisco with limited pedal effort.

 

[Whiplash]

I have used the RC motors as well, and unless you do a lot of special mods they too are limited in what they can handle indefinitely. I had a 63/74 with a BIG reduction and even at 44.4v it was hard pressed to do over 1500 watts continuous even though its rated at 3000. Trust me there just isn't enough mass there to dissipate the heat...

 

[spinningmagnets]

Trust me there just isn't enough [copper] mass there to dissipate the heat...

I agree. My reason for being so interested in this thread, and why I feel it's so important, is that the size and shape of the geared hubs have shown they are the configuration that will support the 2,000 watts continuous (3,000W peak?) that will prove to be popular for off-road, and for high-performance street use. I feel the 1,000W stock GNG kit is adequate for the majority of non-posting E-bike buyers that want a commuter with a broad range of performance characteristics, but the question for ES is what's the upgrade to that?

The 80-85 / 80-100 RC motors are sometimes hard to get, and even then they are best used (IMHO) as a BB-drive with a Thud modification where they are re-wound to a lower Kv and terminated in WYE. That is hard for the average builder to do properly, but this crossbreak conversion is more doable for the average garage builder, and this conversion uses less expensive controllers.

 

[Whiplash]

For me it was the low speed control I was after without having to add halls to the RC motor. Not to mention the NOISE!

 

[crossbreak]

The 63/74 is made for 1500Watts continuous if properly cooled, but cooling it right is almost impossible, since it has a long stator. Still the 80-100 is the smallest one for a single speed drive IMO if one wants to reach 30-35mph since it can take the 4kW, cooling is an issue here, too. it does not fit on a bike well and it's load no doubt. Recently the 80cc rotomax (would be called an 110/70 in old school terms) came to market, which should be more simple to cool. I look for a better power/weight ratio and hope the throttle issues will get solved by software. Just wanted to add this here since this point of view is helpful to design reduction/spread/shifter setup for the MAC,too.

 

[Whiplash]

Actually the 63/74 I have was rated at 3000 watts+ but it could barely take 1200 and even less continuous...

 

[mr.electric]

A converted hub motors just looks cleaner than the exposed belts and chains of GNG or most R/C builds. There are some amazing R/C and GNG builds out there but a converted hub motor is so well contained in a small package.

 

[crossbreak]

A converted hub motors just looks cleaner than the exposed belts and chains of GNG or most R/C builds. There are some amazing R/C and GNG builds out there but a converted hub motor is so well contained in a small package.

A derailleur doesn't look nice also. It's function which makes people use it. There will always be people who search for the most clean solution. At intermot exhibition i even saw single speed middrve bikes

ok this one is three-speed, I sadly took no pic of the single speed Sunstar.

 

[crossbreak]

3-speed could be something suited for your build also, mr. electric

 

[crossbreak]

I'm a bit disappointed from my converted bafang. I run 25amps through it, but it does not have much better acceleration than my hub mounted bafangs. The only big difference is it doesn't overheat with 25amps and can climb every hill i met. Power can only be substituted by power.

Edit: Thing is: There was a GREAT difference in offroad performance between converted and unconverted tongxin. With the small bafang the difference isn't that great. It is already almost as good as my converted tongxin was. I just had no real comparison at that time (sitll bothered about the 250W legal issue at that time :lol, since I didn't use a MAC ever in my life. I can imagine that the difference will be even less with a MAC. I only do light offroad, so this might differ for you whiplash.

 

[mr.electric]

Yes a 3 speed would probably be a nice set up. Would a shimano 3 speed hold up to 1000-1250 watts?

 

[crossbreak]

It will if it is as tough as the SRAM 3-speed, if you use good lubricant, which one I can not tell exactly. Should be relatively thick/viscous. I guess you will have to swap the whole unit after 5000 miles :?

 

[StudEbiker]

The Nuvinci developer's kits are back on sale for $150 at Fallbrook, they would be an excellent option for this type of conversion.

 

[MitchJi]

Hi,

Yes a 3 speed would probably be a nice set up. Would a shimano 3 speed hold up to 1000-1250 watts?

http://endless-sphere.com/forums/viewtopic.php?f=28&t=5297&p=79741#p79741

I'm sure the Nexus hubs can handle the power. There's a guy with a bike shop down in San Diego (Rusty Spokes, in Pacific Beach...) who custom builds chopper-style motor bikes, using Etek and Mars motors. He uses these same Nexus 3-speed hubs with these beasts, and they have worked flawlessly, even pumping 15 hp through them. Instead of chain, however, his bikes are belt driven, using regular toothless v-belts.

There's a little bit of "slippage" when starting out, which is like a clutch. Works great, he says, and these things can hit close to 60 in just 2nd gear. Anyway, he's never had a single problem with any of the Nexus hubs.

Even without the slippage 1000-1250 watts sounds safe.

 

[amberwolf]

If you don't shock-load it, I bet a lot of 3speed hubs can handle plenty of power...but if they get any shock-load, sudden applications of torque, and they have any gear-lash in them, they'll probably eventually rip teeth out of the gears at the roots just like in the (I think) Sturmey Archer 3pseed that was posted in disassembly after this happened. (by 1000w? I forget who).

 

[fechter]

Yes, shock loading creates many times the normal force on the gear teeth. It would be nice if there was an easy way to add a damper of sorts. Like the springs inside the clutch disk on a car or elastomeric coupling.

Most of the 3 speeds I've seen have a somewhat limited ratio change from 1st to 3rd but probably good enough. Most of them are also 1:1 in second and have the highest efficiency there since the gears aren't loaded. If you spent most of your time in second, I think it would last a lot longer.

 

[Whiplash]

A cush drive as its called is used on a lot of motorcycles and would be fairly easy to make if someone wanted. You just need to look into pouring two part polyeureas..

 

[bØb]

I present you the LOVEJOY sprocket :lol: The front half is made with the sprocket free to rotate on the hub and the back half is keyed to the shaft with an elastic spider between the two halves Not available anywhere, because they don't work

bØb

 

[rodgah]

torque limiting clutch??? they seem to work very effectively for me if set up correctly. Interesting enough if you have it set fairly loose it gives a CVT type of feel to accelerating, though the friction material wouldnt last forever.

by 1000w? I forget who

I think it was Kim, I seem to remember a video of his friend trying a burnout on his bike and the hub gave up.

Rodger