Converting a hubmotor to a middrivemotor

[mr.electric]

Nice work! Heat sink looks awesome.

[kfong]

Nice conversion, I might be doing something similar this winter.

[mr.electric]

My heat bridge is coming along. I pulled the axle out completely. It rides in a bearing pressed into the rotor. Does the shaft need to be locked to the rotor or planetary carrier. As it sits now it seem that the shaft would just spin freely disconnected from everything. The sun gear is part of the rotor on my BMC.

[bØb]

My heat bridge is coming along. I pulled the axle out completely. It rides in a bearing pressed into the rotor. Does the shaft need to be locked to the rotor or planetary carrier. As it sits now it seem that the shaft would just spin freely disconnected from everything. The sun gear is part of the rotor on my BMC.

mr.electric,
yes, the shaft must be locked to the the planet carrier. In a MAC, and I would assume in a BMC, just put the key back in shaft for the clutch. The clutch will then be a freewheel for the shaft. The clutch will engage and turn the shaft. The shaft will turn in the opposite direction from the way the housing turned. You must mount the housing in the frame in the opposite direction from the way it was used as a hub. Your output shaft, what use to be the axle, will turn in the proper direction for chain or belt drives.

I hope I have explained this so somebody can understand it. Planetary gearing can be very hard to get your head wrapped around. These little motors use the simplest of planetary gear drives. There should be some some examples of planetary or epicyclic gearing on the web or you could borrow a Machinery's Handbook. There are examples that will blow your mind for hours analyzing the outputs.

Just remember: Lock the stator to the case and the planet carrier to the axle (new output shaft), turn the motor around in the bicycle frame.

[Miles]

http://science.howstuffworks.com/transp ... /gear7.htm

[crossbreak]

yes, the shaft must be locked to the the planet carrier. In a MAC, and I would assume in a BMC, just put the key back in shaft for the clutch. The clutch will then be a freewheel for the shaft. The clutch will engage and turn the shaft. The shaft will turn in the opposite direction from the way the housing turned. You must mount the housing in the frame in the opposite direction from the way it was used as a hub. Your output shaft, what use to be the axle, will turn in the proper direction for chain or belt drives.
...
Just remember: Lock the stator to the case and the planet carrier to the axle (new output shaft), turn the motor around in the bicycle frame.

That's a good abstract of what I wrote the past 10 pages thanks for that, maybe you should write a wiki page or HowTo

[bØb]

crossbreak,

Thanks for the complement. I should have mentioned to make holes in cover to get wires out though.
Yesterday I order some 2.75"Ø aluminum rod to make the header/heatbridge. It will probably come in next week. That will be the last big operation to complete the motor. After that it should just be wiring and assembly. I have been saving pictures and will probably try to start a new MAC specific to How To soon. I didn't really want to start it until I had the motor completed. You were really the inspiration that got me started on this. I'm not really a bicycle person I bought the bicycle that the motor will go in >30 years ago. It was used then and I still have not worn out the half used up tires that were on it. I did buy the bike new tires, rear wheel, freewheel cluster, pedal cranks and chain rings for the project tough

Sorry about the thread jacking,
bØb

[spinningmagnets]

I think I now understand what you are doing. As I now understand it, I am very impressed and I now feel this is a very important development. I have been pondering the various options for the "ES BB-drive" thread http://endless-sphere.com/forums/viewto ... 28&t=45313 (similar to the GNG, but capable of double the continuous power). My concern over using the MAC 12T was how wide the axle had to be, even after trimming a small amount to make it slightly narrower.

In this configuration, The BB bracket will attach to the fixed shell, and the hub axle can be trimmed even narrower. The fact that the planetary ratio is improved to 6:1 from the stock 5:1 is an additional benefit. I believe the BMC and Bafang geared hubs both have two winding choices available, but the five windings from cell-man (including the very hard to find elsewhere 12T) would make the MAC my first choice to try out.

[crossbreak]

I think I now understand what you are doing.

ok... now you blame me was it really so confusing what I told all the time? Maybe you should have told me. I thought that most people here know how a planetary works, mabye I was wrong I just wanted to motivate other people to do the same as i did, cause I see some benefit, and wanna get the necessary feedback to get on further. Wanted to give back what I got from the sphere, the half bakery, just as half baked as I am. I spend almost spend $200 for this project (of course this was my fault), just to have a look if it might work. I really should have split both topics in the first place.

I think this topic is done... it's time for two new ones, the first one called "converting a MAC to middrive - -making the axle rotate" If this is said first, most people will get it in the first place I hope.

And another one called the "through the jackshaft drive" which is about having the crank sprocket on the left and only one final output drive sprocket on the motor (on the right). It could be called the "ANTI freewheel crank" topic - cause I'm a resistance fighter here really hope I wont stay the outcast, seeing so many drives going to trash after weeks of use, can you image where I get my €30 250W bafangs?

Seeing the GNG Gen2 which still runs through the cranks using a tiny 9T drive sprocket shows that this topic is completely off any sence at all. Seeing all those guys here buying something which lasts a few weeks like 1200W cyclone and 1000W GNG just makes me sad

No excuses plz about thread jacking, it was meant to be an enlightening in the first place. I have no time to answer all questions, neither for a good Middrive Wiki, that I started no one will ever further develope, I guess. There is no need to chew the hole story to a perfect HowTo, getting through the mess is something educating for noobs, and it's always fun and encouraging like bob can tell now i hope Bob, just start a new thread with the above topic. This thread is really done now. People may read this one if they really wanna know how this all came up, if they dont (like most), they can just stick to the two new ones.

[bØb]

I was kind of thing along the lines of "Problems in Paradise,How To Solve Them. Get the Hub Motor Between the Pedals"

bØb

[crossbreak]

mine fit just perfect using my octalink cranks... do you use a front or a rear one?

ok guessing from your pics you use the rear one.. this will be a hard time for you though

[Green Machine]

Crossbreak.

Thanks a lot for starting this thread...it is super innovative and informative.

Is there a separate thread where we can see pictures and performance specs of your bike?

[crossbreak]

thx greenmachine, actually there is none yet. I'll start one if i can go on with my build (I will summarize the story for the new subscribers then), ATM i am moving and dont have time. i'll start a new one when it goes on . i'll make sure all three new topics that will continue this one, will be linked here.

As for the specs, it will only do 1000W input, but that is still a straight number for a 250W bafang. It does wheelie yet but it's not up to being reliable yet, so i've still got some work to do. ATM it draws about 1150Watts. I had a look at the gears and the are fine yet, even they are peanut-butter-style polyacetal gears. The performance measurement will have to wait, till I've got someone at university who will test the hole system on the bench. There is a bench at my university, but it does not fit my bike, so there is no simple way to go. Sadly I dont know any private person here in germany owning a test bench. Hope this will change soon, since there is a lot of interest about these topics in my country. It's always hard to convince people about the reliable and cost effective drive systems, cause they always have lack of power and power density in comparison . BUT I have got one advantage: Everyone here compares with the Bosch middrive, which has obvious drawbacks both in power and power-density compared with my proposed system

[fechter]

That's pretty cool. It took me a while staring at the pictures to see how the parts get rearragned. If the housing is stationary, you don't have to run the wires through the axle anymore, which is a huge advantage.

The heat bridge, while it will certainly help, won't be all that effective in transferring the heat from the windings due to the long thermal path. Forced-air cooling would cool better but create the possiblility of dirt entering the motor. Water cooling would be feasible since the water lines don't need to go through the axle. What you really need is the shortest possible heat conducting path from the copper to the outside case. Other than the hall sensors are in the way, it might be possible to make an aluminum ring that fits between the stator windings and the case cover so the path is only 10mm or so long. Silicone between the copper and heat sink will be pretty effective for heat transfer.

It seems the attachment of the sprocket to the axle is sort of a weak spot. A keyway of some sort is probably needed. If you cross drilled the axle and pressed in a pin that only sticks out on one side, it could fit the keyway on a freewheel sprocket.

[mr.electric]

Got it just put the key back for the planet carrier. Sorry mental laziness has prevented me to think through how this motor will work conceptually.

[Kepler]

Great thread guys. Taken a bit to get my head around the conversion but its coming together now. I can now see a mid-mount in my future

[StudEbiker]

I found some internal pictures of the BPM motor on another thread.

http://endless-sphere.com/forums/viewto ... =3&t=12582

Does it look like this modification would work with it?

[spinningmagnets]

If the DIY mounting bracket for a BB-drive clamps to the non-spinning shell, it "looks like" the bracket ID space only needs to be about 4-inches (100mm). The shaft could be cut off completely flush on one side, and only needs to be long enough on the other to hold the freewheel. The motor can be flipped, so FW can be on the left or right without needing to reverse motor.

The bracket plate on the right could be fairly flat and thick, the plate on the left would probably need to be staggered a little. (or one plate only like the AFT kit?) http://endless-sphere.com/forums/viewto ... 28&t=44572

[StudEbiker]

Part of the idea of this system though spinningmagnets is to not use freewheel cranks. To do that, you will need to have a gear on both ends of the motor. Putting this on a recumbent like the Bike E like I would plan on doing eliminates the need to have the motor fit inside the cranks. There's LOTS of places to put it so that the cranks won't come anywhere near motor.

[mr.electric]

Part of the idea of this system though spinningmagnets is to not use freewheel cranks. To do that, you will need to have a gear on both ends of the motor. Putting this on a recumbent like the Bike E like I would plan on doing eliminates the need to have the motor fit inside the cranks. There's LOTS of places to put it so that the cranks won't come anywhere near motor.

As you point out, If you could find a frame where the motor fits properly you don't need freewheeling cranks. The motor shaft free wheels in the motor because of the one way on the planetary carrier. You do not even need two free wheels to set up a mid drive. One sprocket and one freewheel would do it. Both chains could be on the same motor axle , cut off the extra axle or you could do jack shaft style it would work either way.
I can't wait to ride one if these.

[mr.electric]

ImageUploadedByTapatalk1355203146.060577.jpg (82.31 KiB) Viewed 1460 times

Ready to spin my BMC mid drive motor. Controller comes tomorrow. After years of threading wires through impossibly thin axle passages I made a huge hole in the cover plate to allow assembly / disassembly with absolutely no interference with the wires.

[bØb]

It has been a few days since I have had any thing else to report on the MAC conversion. Been waiting on a piece of aluminum round stock to come in. I got here today, so I made the header / heat bridge to anchor the motor stator to the motor cover and heat sink.

After a few calculations of dimensions....

PC110459.JPG (79.44 KiB) Viewed 1437 times

and a lot of work....

PC110455.JPG (78.62 KiB) Viewed 1437 times

I ended up with this little piece. It is the small piece laying on top of the steady rest.
It's only 0.60" long but had to be cut from a 12' long piece.
After this I drilled the six holes to mount it and cut a small recess in the opposite side to clear a bearing in the cover. I assembled the motor and everything turns freely. I will check the mesh alignment of the sun gear, planet gears & ring gear with Prussian blue or some layout fluid to make sure.

Soon the time will come to put the motor in a frame. I was going to put it in my old vintage mountain bike, but I couldn't work around the diamond frame without too many compromises. I have a beach cruiser I restored for my wife many years ago that she hasn't ridden in many years, so I think I am going to build something like an Xtracycle EDGERUNNER. There should be plenty of room to work with there

bØb

[mr.electric]

Spun the motor after some frustration. I used a bmc with roasted wires as the donor motor. Turns out the hall sensors are fried on this motor. One of the hall power wires must have shorted out when the phase wires melted into a clump in the axle tube. I think I have repaired four bmc motors with melted wires and never had the halls blown before this time. Guess it was unlucky. We really wanted to spin this motor so we hooked it to a low speed sensorless controller. As expected the controller losses sync easily but at least we got to give it a run. It is cool to see the shaft turn nice and slow looks 100 rpm-ish. It makes a vague gear noise but not nearly as loud as an rc motor. I am anxious to replace the hall sensors.

[fechter]

After years of threading wires through impossibly thin axle passages I made a huge hole in the cover plate to allow assembly / disassembly with absolutely no interference with the wires.

That's one of the best parts of the conversion. There would be room for cooling lines as well, though forced air would be much simpler.

[Ariane]

....