Converting a hubmotor to a middrivemotor

[crossbreak]

another pic

 

[neptronix]

I like what's going on here.
I've thought of using a hub motor as a cheap and simple mid drive for a long time. The MAC motor got crossed off my list for this purpose long ago though because of it's tallness. On a cargo bike, that's no problem, but on a mountain bike, it takes an awful lot of space.

I always wanted a more square shaped motor, like the MXUS geared 350w.. but more powerful.. sadly, i think the big BPMs come closest to that ideal, but aren't much better than the MACs, if at all.

One thing to note is that you want the fastest MAC motor you can manage to gear down. Continuous power goes up the faster these spin. The faster they spin, when you gear them down, the more torque they make.

8T is good for this purpose as on 36v you make 320rpm unloaded, which is about 250rpm loaded. Not hard to gear that down to 100rpm pedal cadence.

Even better would be a 6T, 400rpm unloaded, 320rpm loaded, you just need a hair over 3:1 to get the right cadence with that setup.

It should produce 750-1000W continuous in that kind of situation. i would limit it to 750W to prevent heatsoak.

The 12T would likely produce under 400W continous without producing a bunch of heat.

These are 36v figures, so higher voltage can change the picture.

 

[crossbreak]

...better would be a 6T, 400rpm unloaded, 320rpm loaded, you just need a hair over 3:1 to get the right cadence with that setup.

This motor would do 267rpm loaded if modified (from stock 5:1 reduction to 6:1 when modified). With a 44/13 you get an 80 cadence. This is @36V still think this motor could get much more than 1500W continuous if treated right. It does not matter if we use a 12T MAC @ 72V or a 6T MAC @36V. The 6T MAC do can much more @72V than a 12T as it can be more efficient and can do more WATTs continuous since it runs faster.

Edit:
The trick about "treatment" is that we have to use a low phase current limit. This way it drives like a gas engine. The power rises when you go up with rpm. Just like the power rating of our gears. That's perfect since it "skills" the driver to really USE the abilities of his gears

133V are a good treatment for the 10T after ebikes.ca

The this pic shows the difference treated (FAT lines) by current limit or not. 200NM (look where the original blue line ends on the left) of motor torque would kill EVERYTHING, not only our plastic gears in the MAC - even the freewheel at the rear wheel

 

[neptronix]

The 6T MAC do can much more @72V than a 12T as it can be more efficient and can do more WATTs continuous since it runs faster.

The trick about "treatment" is that we have to use a low phase current limit. This way it drives like a gas engine. The power rises when you go up with rpm. Just like the power rating of our gears. That's perfect since it "skills" the driver to really USE the abilities of his gears

It could produce more power, theoretically, but there are obstacles in the way.

1) A 16 pole motor ( the MAC hub for example ) has a very high electrical RPM, which is problematic for many ebike oriented controllers; for example, the EB2 controllers had major problems with these motors when ran at >400RPM at the wheel. EB3 is OK at higher speeds on the 8T ( ~40mph on a 26" wheel ), but much beyond that could be problematic ( controller losing sync etc ). I'm not sure where the eRPM limit is on the EB3, but you run a 6T motor on 72v and you'll probably find it

2) even if you're getting 80% efficiency, when you run 1000W continuous, you've got 200W to shed continuously. On 2000W, you've got 400W to shed. Geared motors are awful at shedding heat, in fact i can get my 8T 26" MAC quite hot after 2-3 hours of 500-750W continuous. Pretty sad, no?

3) stator losses are higher the faster you spin and the more poles you have. It's likely that the 80% efficiency you get at stock voltage becomes 70% or 75% efficiency on that 16 pole MAC hub when you are spinning it at 800rpm ( 4000rpm internally ) at the wheel. Cell_man warned me about doing this when i proposed the idea to him.

4) you've got a higher reduction ratio to deal with at the crank, so you gotta deal with that. 700rpm loaded is going to require a double reduction, no way around it. And you've just negated the whole idea of using a hub motor at that point :/

Sorry to take your idea out to the woodshed.. :lol:

 

[crossbreak]

The 6T MAC has ONE real benefit over the 12T: With a 6T and 12s lipo/74V you can still touch both battery poles wihtout dying, with a 12T MAC and 148V this is not the case

 

[crossbreak]

. It's likely that the 80% efficiency you get at stock voltage becomes 70% or 75% efficiency on that 16 pole MAC hub when you are spinning it at 800rpm ( 4000rpm internally ) at the wheel. Cell_man warned me about doing this when i proposed the idea to him.

So the ebikes.ca calc is wrong. I already doubted it. It does not take the eddy current losses into account. But where is the best Q-point we should deal with, using the MAC? There has to be specific voltage and load @which it's driven best. I guess we should ask miles which program we should use to find it, hope we don't have to write our own one

The BEST way to find it would be a bench test of course. e-bikes.ca did some and extrapolated the obvious wrong data. There has to be a more sophisticated motor model than the one they used, we just have to find and calc.

The trick about "treatment" is that we have to use a low phase current limit. This way it drives like a gas engine. The power rises when you go up with rpm. Just like the power rating of our gears. That's perfect since it "skills" the driver to really USE the abilities of his gears

This sentence is still right even if the MAC cannot do so may WATTs as calced.

 

[neptronix]

Well, it's not far above the factory rated specs unfortunately.
12S - 14S on a 6T is probably the breaking point.

 

[neptronix]

By the way - the Bafang BPM2 is supposedly an 8 pole motor - it might spin up a lot better.

 

[Whiplash]

The 6T MAC do can much more @72V than a 12T as it can be more efficient and can do more WATTs continuous since it runs faster.

The trick about "treatment" is that we have to use a low phase current limit. This way it drives like a gas engine. The power rises when you go up with rpm. Just like the power rating of our gears. That's perfect since it "skills" the driver to really USE the abilities of his gears

It could produce more power, theoretically, but there are obstacles in the way.

1) A 16 pole motor ( the MAC hub for example ) has a very high electrical RPM, which is problematic for many ebike oriented controllers; for example, the EB2 controllers had major problems with these motors when ran at >400RPM at the wheel. EB3 is OK at higher speeds on the 8T ( ~40mph on a 26" wheel ), but much beyond that could be problematic ( controller losing sync etc ). I'm not sure where the eRPM limit is on the EB3, but you run a 6T motor on 72v and you'll probably find it

2) even if you're getting 80% efficiency, when you run 1000W continuous, you've got 200W to shed continuously. On 2000W, you've got 400W to shed. Geared motors are awful at shedding heat, in fact i can get my 8T 26" MAC quite hot after 2-3 hours of 500-750W continuous. Pretty sad, no?

3) stator losses are higher the faster you spin and the more poles you have. It's likely that the 80% efficiency you get at stock voltage becomes 70% or 75% efficiency on that 16 pole MAC hub when you are spinning it at 800rpm ( 4000rpm internally ) at the wheel. Cell_man warned me about doing this when i proposed the idea to him.

4) you've got a higher reduction ratio to deal with at the crank, so you gotta deal with that. 700rpm loaded is going to require a double reduction, no way around it. And you've just negated the whole idea of using a hub motor at that point :/

Sorry to take your idea out to the woodshed.. :lol:

You must have some serious battery! 3 hours at 750 watts is like over 2kwh. Far more than anyone would carry on a bike designed for off road use if not for the size than the sheer weight would not make sense. Honestly I would not hesitate to run an EASY 1500 watts continuous on my Mac and would venture to say most normal riders would not have enough battery for thermal overload even with 2000. Obviously if you are talking pikes peak where its loaded down hard the entire time, you may be right, but in the real off road world you simply can't hold a bike wide open that long you have to lift for obstacles, turns etc. My findings with the Mac are just that it'll take the abuse as long as you let it spin and use gearing to get your torque. It takes the load off the planetary gear bearings. They seem to be the new weak link at least mid driven...

 

[neptronix]

Yes i do, i have 2.5kw-hrs in total. I live in a very, very spread out state. I've done >60 mile rides, and the MAC hub motor is frankly unsuitable for that long of operation because it builds up a heat peak over a long period of time. Eventually the halls start getting pissed off and start causing intermittent operation :|

I've done 1500W-2000W continuous on my MAC before, but that's good for about 10 miles of mildly hilly terrain in the winter.

I've got the motor and controller super hot doing super steep offroad riding, to the point of stalling. Takes forever to cool off .. it would be different in a mid-drive system where you can keep the RPMs up, for sure.

Mind you i am talking about a heavy rider on a heavy bike with tons of battery ( 30AH 12S, or more ) in a 26" wheel, not in a mid drive.

For short rides, the situation is different. All i'm trying to point out here is that the MAC hub does a bad job of shedding heat, all geared motors tend to suffer the same problem since the stator heat needs to go through multiple walls of steel and aluminum to get out. It's an ez bake oven in there

You must have some serious battery! 3 hours at 750 watts is like over 2kwh. Far more than anyone would carry on a bike designed for off road use if not for the size than the sheer weight would not make sense. Honestly I would not hesitate to run an EASY 1500 watts continuous on my Mac and would venture to say most normal riders would not have enough battery for thermal overload even with 2000. Obviously if you are talking pikes peak where its loaded down hard the entire time, you may be right, but in the real off road world you simply can't hold a bike wide open that long you have to lift for obstacles, turns etc. My findings with the Mac are just that it'll take the abuse as long as you let it spin and use gearing to get your torque. It takes the load off the planetary gear bearings. They seem to be the new weak link at least mid driven...

 

[crossbreak]

I've got the motor and controller super hot doing super steep offroad riding, to the point of stalling. Takes forever to cool off ...

All geared hubmotors are badly cooled. You obviously can not push 2000W through them with the stock cooling. But you should be able to do that after conversion. Heat dissipation is increased a lot by this conversion, at least 3-4 times I would guess (due that heat bridge). After conversion of my 180W tongxin I could push almost 700W into it, using 450Wh for over half an hour. It got only warm (since perfectly cooled). If I would have done than unconverted it would have been molten Sadly I could not push that one further since it's friction drive failed.

The BPM would indeed be a better candidate, the bafang page tell it's 16pole. Even my SWXB has 20 ok it was cheap, but my next one will definitely be a BPM! Hope they still sell the version with cable through the shaft.

 

[mr.electric]

Once the motor case is stationary wouldn't it be easy to force air cool the motor. With old scooter and bike motors people regularly doubled output with out burning out as compared to sealed case non cooled set ups.

 

[crossbreak]

Once the motor case is stationary wouldn't it be easy to force air cool the motor. With old scooter and bike motors people regularly doubled output with out burning out as compared to sealed case non cooled set ups.

Hey, plz read the FIRST page before posting! We don't want dirt in the gears. This does not make sense. Already killed one by doing that. I will NEVER try again. We dont want to play around here, we want some SERIOUS SOPHISTICATED vehicle.



btw. we want to TRIPPLE output. double is not what we aim for, that would not justify the effort.

This thread got messed up. I'll start a new one, once the prototype is finished an we can have a look at the first test results.

 

[mr.electric]

I am interested to try converting a Mac/ Bmc to stationary case. I have an old bmc 1000 with burnt wires sitting on a shelf and my friend has one too. I think the BMC 1000 is equivalent to a lower turn Mac.
I will post some pics of bmc internals when I get a chance.
Realistically I will not be able to try this out until after the holidays but it seems
Ike the motor mods could be performed in one day using hand tools. A great addition to the endless sphere world. Hats off to Crossbreak

 

[crossbreak]

I will post some pics of bmc internals when I get a chance.

that's highly appreciated!

Ike the motor mods could be performed in one day using hand tools.

You will need a bench vice of course. For mounting sprockets to a shaft you need a spacer part made with a lathe (the brass spacer between shaft and freewheel sprocket). If you really want to try, I can send you this part to you, it fits in an envelope. Already thought of making 10 of them, cause this will be needed a lot by members who wanna try this.

 

[crossbreak]

Some more calcs for the MAC

My proposal the MAC:
With 44/13 crank to MACshaft reduction you get 270rpm@MACshaft@80rpm cadence.
That's equivalent of an unmodified MAC (with only 1:5 internal reduction) of 324rpm, which would be 39kph if mounted in a 26" rim.

That would be 1620 motor rpm or 27 revolutions per second. The MAC has 16 pole pairs. So the field changing frequency would be 432Hz, that's a lot. 39kph should not be too much for a MAC so I wonder how thick are the laminations ? If we had a close up pic of the stator and would know the stator width that could be cleared. Then we could figure out a good gearing for the MAC.

The same for the Bafang would mean only 216Hz which seems acceptable. Still we also need to know the lamination thickness for this motor.

 

[mr.electric]

I don't have a mac or BMC with me today but I do have a few pics on hand. No measurements as of yet.

 

[crossbreak]

Thanks, for the pics. That's not what I was looking for. Could you be so kind to take off the gears, pull off the rotor and take some more pics?

 

[neptronix]

If the MAC hub is anything like my 2kW MAC hub, it should have about 0.33mm laminations.

 

[crossbreak]

Thank you neptronix! Means that 400Hz-450Hz might be ok, running it above will get lossy, so far i learned from miles. SO if we used this hub in a 26"inch rim it will start to get lossy over 45kph / 28mph Edit: This still fits the calcs using the simple 44/13 overdrive to the MAC shaft.

Tried to drive the prototype the first time. Was only good for 1 wheelie. After that, the torque arm (4mm thick steel!) I used to fasten the freewheel on the shaft looked like this Guess that the 30amps limit was too much for it
Seems like I have to make one from stainless steel. I'll take a 5mm sheet this time. And, what might be more important, i'll set the max phase amps to battery amps ratio to 1:2 or 1:2.5 at least.

 

[mr.electric]

Nice torque. I like it.
So turn down the amps for the next test?
I wonder if the chain will snap with all that torque.

 

[mr.electric]

Ok I'm convinced. I think I will start converting my burnt phase wire bmc to a mid drive motor this Friday after work. The finished motor is to be installed on a friends bike, I will just be setting up the motor. I will be sure to post copious pics. The plan is to aug out the hole in the aluminum stator and remove the shaft key to allow the axle to spin. I see that I will need to play with the various clips and thrust washers inside the motor to find a combination that is not grindy. If anyone does the conversion to a Mac/ bmc before me post pictures! I would rather follow a recipe that works.

 

[crossbreak]

Ok I'm convinced. I think I will start converting my burnt phase wire bmc to a mid drive motor this Friday after work. The finished motor is to be installed on a friends bike, I will just be setting up the motor. I will be sure to post copious pics. The plan is to aug out the hole in the aluminum stator and remove the shaft key to allow the axle to spin. I see that I will need to play with the various clips and thrust washers inside the motor to find a combination that is not grindy. If anyone does the conversion to a Mac/ bmc before me post pictures! I would rather follow a recipe that works.

I guess you wont get around to dremel some material off the stator carrier. Had to get rid of almost 2mm axially till the shaft could be turned freely in my bafang (Edit: assembled, with all washers and snap rings). Radially i took less than 0.1mm of the dia of the shaft away (using sandpaper #180), that was enough.

How will you repair the burned wind?

I started a new thrad about measuring heat dissipation here: http://endless-sphere.com/forums/viewtopic.php?f=30&t=45871

It would be nice if you could perform a with heatbridge test and without, to see how much cooling was increased.

btw. this is what my torque arm for the crank-to-hub shaft looks like (I rather use as motor sprocket ATM for testing), using a 5mm (edit: wrong, it's 6mm) thick non-stainless steel bar and a M4 screw. Note the treads in the brass spacer, it fixes the sprocket axially. How to instal: First a washer and then the torque arm plate is pushed on the shaft, then the sprocket is screwed on. Then both get connected by the M4 screw (the torque arm plate has M4 threads). It holds up well even with the 30amps controller. I can wheelie in gears 1,2 and 3

 

[bØb]

Ok I'm convinced. I think I will start converting my burnt phase wire bmc to a mid drive motor this Friday after work. The finished motor is to be installed on a friends bike, I will just be setting up the motor. I will be sure to post copious pics. The plan is to aug out the hole in the aluminum stator and remove the shaft key to allow the axle to spin. I see that I will need to play with the various clips and thrust washers inside the motor to find a combination that is not grindy. If anyone does the conversion to a Mac/ bmc before me post pictures! I would rather follow a recipe that works.

I have been lurking here for 2~3 weeks and have been inspired by crossbreak enough to order a MAC 500/1000 geared hub motor to convert to a shaft motor. I ordered the motor, 12FET controller, CA and some other stuff from Cell Man a week or two ago. As soon as that stuff comes in and I confirm that it runs, I will begin modification. I will try to take pictures and measurements along the way. I have never seen a hub motor in person, only the pictures on this forum, so it will probably get interesting.

 

[mr.electric]

How will you repair the burned wind?

It has burn phase wires in the axle tube a common Mac / BMC problem. The actual windings are ok.
I have a special puller I made just for the purpose of disassembling BMC motors. I have repaired burn wires on 4 or 5 BMC motors for people in the Ebike community here in San Francisco. The skinny phase wires have been an issue for anyone running over 25 amps and climbing mountains or long hills.
A mid drive BMC would have terrifying potential for speed and torque judging by what these motors can do when built into a wheel.