Diy outrunner brushless BMX

I ordered my magnets and have about 5 spools of different size magnet wire. I am building a in wheel brushless setup for my bmx. I am building it as a learning project and because I want to try a idea or two of my own out. My plan is to put 48 3/4 " magents on the inside of the rim on each side for a total of 96 magnets which will give a coverage of 76% I am in the design stages of the stator but my plan is to only have a stator that is about 1/8 of the circumfrance and only works (pulls on) the magnets in the 1/8th of the rim. Making it light and simple and giving a good center of gravity. Because of the design high torque numbers should be realy easy to achive. Hi rpm is not what is desired top speed of 15-30 mph would be max I know 30 is rediculos lol.
Question #1 what controller to use? I havent purchased my batteries yeat (I will verify it works with a 12v battery before ordering lipos) but because I dont know what controler to use for sure I cant wind my stator yeat. I am realy considering the mamba monster controler from castlecreations seen here http://www.castlecreations.com/products/mamba_monster.html it is good for 25.2v at 120 amp continous that meens 3024 watts which is 4.053 hp (they sent me a message back saying it can surge 300 amps) SO I know watts is watts which is in the end what eats up the batteries and produces heat. Running a engine at 25.2v&120 amps is the same as 50.4v and 60 amps both = 3024 watts of heat and energy being produced. Am I off on all of this? I am not to worried about the money for the controler but I do want to have a good system.
Question #2 What bike controlers would I want to consider for ~3000 watts? Any advice is welcome as I said it is for a fun learning project that I can hopefully ride around on and have fun with this summer.
Any info on building a brushless in wheel engine is welcome I have been studying the brushless outrunners form various places like lrk for over 1.5 years now.

Oh and what I ment buy this "SO I know watts is watts which is in the end what eats up the batteries and produces heat. Running a engine at 25.2v&120 amps is the same as 50.4v and 60 amps both = 3024 watts of heat and energy being produced" Is for 2 different engines. I do understand and engine designed for 100 rpm at 12volts no load will need to be re-wound or re-designed if it is desired to run 100rpm at 6 volts or at 24 volts (all a no load number of course.)

Yes, watts are watts. Experience indicates you get better performance at a higher voltage and lower current since the resistance losses in the copper are a function of current squared.

We've had fairly good luck with the Castle Creations Phoenix 110HV, but you need something like my throttle interface circuit, or at least a servo tester and BEC.

The other option is a controller that uses hall sensors, like the 'Infineon' controllers from ecrazyman. Hall sensors are a bit of a pain, but the controllers are less expensive and will give good torque from a dead stop.

How do you plan to keep magnetic chunks of crap from sticking to the wheel? This could be a problem. Another possible problem is the force between the magnets could cause the rim to deflect under load. You'll need really stiff spokes.

fechter said:

Yes, watts are watts. Experience indicates you get better performance at a higher voltage and lower current since the resistance losses in the copper are a function of current squared.

We've had fairly good luck with the Castle Creations Phoenix 110HV, but you need something like my throttle interface circuit, or at least a servo tester and BEC.

The other option is a controller that uses hall sensors, like the 'Infineon' controllers from ecrazyman. Hall sensors are a bit of a pain, but the controllers are less expensive and will give good torque from a dead stop.

How do you plan to keep magnetic chunks of crap from sticking to the wheel? This could be a problem. Another possible problem is the force between the magnets could cause the rim to deflect under load. You'll need really stiff spokes.

Click to expand...

Wow it looks like the Phoenix 110hv is good for alot of power it says 50v and 110 amp max. But it doesnt say and continious amp rating. Non the less it is good for 5500 watts which meens I could use it and build my motor to a peak amp of ~100-110 so the controler would only see peak amps once in a while 5500=7.37 hp and thats alot for a 20 lb bmx + rider. Even with a end eficentcy of 50% thats 3.685 peak hp (still alot) But I do plan to design this as eficent as possible and may do alot of leaning and redesigning along the way but thats why I like the challenge!
As for magnetic debris hum....? Well I will think about that and it may be something to overcome but good point!
And the rim is a double wall its pretty tough I am more worried about the glue to hold the magnets on it.
The info you give me makes sense. So......
#1 Do you have the info up on how to build your throlte interface?
Hall sensors wont likely be a problem but if it was sensorless it would be eiser for sure.
I am fine with giving a quick peddle to get the motor into the usable torque
#2 what about heat in the 110 hv (i belive the mamba monster uses a cooling fan mounted on it)

Wow, sounds like a really innovative idea.

Let me know if what I'm imagining is correct (a sketch or two would really help )

Stator secured to frame of bike, two sets of windings, one on either side of the spokes, say within the rear triangle section?

Sounds like it may have a lot in common with a linear motor.

One issue would be getting a nice small air gap, because that relies on having the wheel very true.

Torque should be huge, and cooling won't be an issue.

Good luck!

Grinhill said:

Wow, sounds like a really innovative idea.

Let me know if what I'm imagining is correct (a sketch or two would really help )

Stator secured to frame of bike, two sets of windings, one on either side of the spokes, say within the rear triangle section?

Sounds like it may have a lot in common with a linear motor.

One issue would be getting a nice small air gap, because that relies on having the wheel very true.

Torque should be huge, and cooling won't be an issue.

Good luck!

Click to expand...

Hey thanks I am planing to build it yes in the triangle section. And yeh figured it should have some huge upsides. As for the windings I am working on that but yes one set on either side I am thinking 3 winds per side but I may go with 6. I am in the middle of all my stator research and I have a couple motorcycle stators on my bench I can hack up if I decide. I can true the hack out of my wheel lol I am realy good at that. And the air gap I will make adjustable to do tests and I might make it on a slide so it cant be moved on the go. (I have seen some hi end in wheel brushless system that gets up to top speed on what looks like a dyno then it moves out the stator to increase the air gap and it revs higher on youtube.) But linear motor?..... Hu time to fire up the google machine. I will get picturs up soon as i start the build.

Cool!


I had same idea past year but I never ended with a prototype or test build.
this kind of large & light brushless may be the way to go, hight torque at low rpm,
and if you add delta/wye, could be a fast bike also.

I'm looking forward, and if you get results I may try this on a bmx build.


get fun

So as I am waiting for my magnets to come in I keep searching for a calculator or a scientific way of determining the number of turns per pole or at leaste a place to start! So if anyone has good info or links feel free to link me up Otherwise i am thinking between 10-15 with a delta wind.

http://www.bavaria-direct.co.za/models/motor_info.htm
http://translate.google.com/translate?u=http://www.powercroco.de/&langpair=de|en&hl=en&ie=UTF-8

Hey thanks on of those sights was down last time I looked at it. Some very good reading there!

I have the magnets glued into the wheel and now I will start experamenting with windings to figure out a rpm/v range that will work for me. But I need some info on some more controlers I have seen a few that are alot more powerfull then what is normaly out there. Like the "Instant Start 18 fet Infineon" I don't realy seem to be able to find were to buy or order them. I am still interested in the Castle creations stuff but something designed for what I am using it for might work better and I would like the regen option if possible. Any links?

And I guess I could use a list of what I will ned to compleat this. Someone should make a sticky on converting your bike (if they already haven't)
So...... This is what I can think of let me know if I am missing anything.
#1 motor
#2 controler
#3 Batteries
#4 Display
#5 Throtle

Now the best places to get these things? I have built most of my own motor but may buy one in the future anyway.
Thanks.
Arlin.

Some pictures of where I'm at.....

So I did some math and my goal rpm/volt is 8.5945 which with the 20 inch wheel will give me 50mph at 100v (free spinning) Which yes is alot more then a bmx can/should? travel but if you times that by 77% it = 38.5mph which is still faster then a bmx should ever go but I want to build a killer bmx! as well these are goal #s which I have fould will be very hard to achive. For my tests I used 3 different motorcycle stators and would them my self as well as tested on with the 50 turn delta wind it had. I found it was almost the same resuslts it is increadable how clost all the stators are. I also tested a plastic pen wraped with 20 turns to see what it would give for kv. My resuslts were to the point I figure I need 1000-1600 winds per phase. Now if I wire it WYE (star) then I will need 500-800 winds per phase and this can be cut in half to 250-400 per side because I will have a stator on each side of the wheel Wired in series to each other. But 250-400 turns is alot and when I want to be able to run 100v 150+ amps I NEED very thick wire or thin with multiple strands per tooth. Now does any one know what gauge of wire can be used for this? I found the traditial gauge/amp chart but I have read the amps it can take are higher when in a electric motor. I have found the Number of turns is better to be less with more stator teath Eg. one stator tooth would at 50 turns is less eficent then 2 at 25 turns in series. So I figure I will run as many teath on each side as I can ~12 (I know it need to be multiple of 3) so 12 on each side is 24 teath in total. and That meens 63-100 turns per tooth will work. One more thing with my tests is that I found I can only fit 50-65 turns of 20awg per tooth with the stators I have tested and I found keeping coils closer to the magnet were more efeciant. Now where do I get a custom 1/4 of the circumfrence 15 dia set of stator plate made?

Somewhere along the line you made a critical mistake in your number of turns calculations. I think you are off by an order of magnitude.

This idea is very slick, and I look forward to seeing how it works out, and how the stator design ends up looking. Do you have connections to have your laminations custom stamped?

liveforphysics said:

Somewhere along the line you made a critical mistake in your number of turns calculations. I think you are off by an order of magnitude.

This idea is very slick, and I look forward to seeing how it works out, and how the stator design ends up looking. Do you have connections to have your laminations custom stamped?

Click to expand...

I would love to find out I made a mistake in the calculations and find I need less turns. But I am pretty sure I got it right. So I will explain my test I used a unwraped motorcycle stator and wound it with magnet wire the striped the ends and hooked up my oscilloscope and turned it rather fast with the peddles you can see a screen shot of one test.
My notes from that test are 8ms + peak to + peak which I mesured multiple peaks to average the distance for one to the next. And it was 4.363 peak to peak voltage so we need to know max voltage which is 4.363/2=2.1815v.
Then for my wheel I have 48 magnets on the side and that meens 24 of them are giving + peaks so 8ms x 24 = 192 ms for one revalution and to convert that to rpm. 1/.192 = 5.2083333 rpm per second and times that by 60 to get rpm per minute 5.2083333 x 60 = 312.499999 rpm and for the kv rating 312.4999rpm/2.1815v=143.25 rpm/v
That perticular test was on a single tooth wraped with 100 turns spesificaly to get data from. I do understand if I run 3 stator teath on each side I will wind one side to the other in series which would meen 71.6 rpm per volt in delta and in wye it would be 71.6 / 1.8 = 39.777 rpm/volt which devided by my desired rpm/v of 8.5945 = 4.628 which meens I either need 4.628 times as many winds or 4.628 as many stator teath all would in series. and just to go back to the start 4.628x100 = 462.8 x 4 = 1851 total winds for that phase ( I times it by 4 because in a wye wound 6 tooth stater it would energise 4 winds at once) and I under stand it would be 1815 for the total for the two phases that are energised in the wye wound stator. but in a delta it would be 1028 total winds but for only one phase.

And no I have no conections for custom stamped laminations! I am totaly up in the air on how to build the laminations one way I keep coming back to is cutting up a motorcycle stator!

So do to the high cost of building my own stator and time running out for now I bought an x5304 and I am using 24s 5ah of lipo. I have 4 more 6 s packs coming to finish it. But for now it is perfact to get me around on this end of town. I can topout at 70km/h on flat ground and it has alot of wheely power if you are not ready and give it to much gas at anything under 35km/h it will flip you over! I will get more video and picks up and data up soon.
I do plan to finish this diy wheel one day but not right now.

-thread bump-
wow I been surfing the old posts today & found this.
Arlo1,
you are remarkably close to duplicating a wheel motor (axial flux) simular to a solar powerd car drives the university's have been building.

Here are a bunch of visual samples if you ever think to revisit the design. Time to re-think the stator
http://www.youtube.com/results?search_query=axial+flux+motors&search_type=&aq=f

I am going to build it one day..... but for now I started working full time again so time is hard to find for big stuff that doesn't make me money as well I got to pay off some debt. It won't be axial flux but a partial stator is what I planned as well I will need to find a better glue for the magets. I started to price out stator magnet wire and it was going to cost me more for that alone then a x5 brand new does :? Any ways I will put pics and video of what I built on here. and do a write up soon.
So far it is a methods kit 18 fet with regen and a x5304 with 24s 5000mah lipo and another 24s 5000mah lipo on the way. I have a CA and a thumb throttle.
And I can almost bet people if they ride it it will flip them off. I think the free movement at the start of the throttle is part of the reason and the insane power is the other. I am pulling over 150amps continious and seaing peaks of 260+ with the pack sagging to 88-90v at those times. Once my next pack shows up look out! I drag raced my buddies truck and beat him to 60km/h and I have already blown one tire from a big burn out! I am now shopping for a motorcycle tire to fit. Looks like a pirelii 2.5x16 moped tire will work. Once I have that sitting here I will make a big burnout video!

So yup just ordered a set of these. http://www.motorcycle-superstore.com/5/32/404/5780/ITEM/Pirelli-ML-75-Rear-Moped-Tire.aspx I figure one for the rear and one for the front and they should resist flats instead of these thin bicycle tires. And having one on the front wont hurt to help keep the front on the ground a little better!

Your bike sounds like fun
Where did you find a cross referacne to the motorcycle tires?
What is the rim size you are fitting them on?
I have looked a couple of times but have not found any solution as of yet.
my goal is to get a decent 4ply tire onto a 24" (504mm bsd) rim.
thanks for any help-T

I would suggest putting a layer of clear adhesive packing tape over your magnets before you put your bike on the road, when magnetic debris sticks to them (which it will for sure) all you have to do is peel off the tape to get rid of it and then just put on another layer of tape.

Getting ferrous dust off high energy magnets is a major chore since they are so powerful.

Jonathan in Hiram said:

I would suggest putting a layer of clear adhesive packing tape over your magnets before you put your bike on the road, when magnetic debris sticks to them (which it will for sure) all you have to do is peel off the tape to get rid of it and then just put on another layer of tape.

Getting ferrous dust off high energy magnets is a major chore since they are so powerful.

Click to expand...

I actualy road it for 3 months with all the magnets on the rim and suprisingly enough I didn't get any debris on them.

Thud said:

Your bike sounds like fun
Where did you find a cross referacne to the motorcycle tires?
What is the rim size you are fitting them on?
I have looked a couple of times but have not found any solution as of yet.
my goal is to get a decent 4ply tire onto a 24" (504mm bsd) rim.
thanks for any help-T

Click to expand...

uhm whell I mesured the rim and it seems to be a 16 inch rim. then I just did as much reserch as I could to find sizes I need. I work part time at a motorcycle wrecker so I looked at alot of mopeds and mesured 2.0x17 and 2.25x17 and a 2.5x17 tire on a rim to see how much it will add to it. I wanted a touch more diameter so I can get my top speed up. I want to speed at 80km/h or 50 mph on the road.