Miles' 90mm inrunner build thread

[Miles]

Ok.

Outer endturn is 12.0mm long

Inner endturn is 8.6mm long

6 x 60 = 360mm

6 x 12 = 72mm

6 x 8.6 = 52mm

Total length for each coil 484mm

8 x 484 = 3872mm

Round up to 4000mm for interconnections.

Cross sectional area 2.285 sq. mm

I make that 0.031 Ohms....

Phase resistance reported in Emetor is 0.03246 Ohms

 

[F4P]

Looks beautyful! You are going to make 8 coil in parallel then the winding be easy with these wires.

 

[Miles]

Thanks F4P!

There are 2 opposing groups of 4 coils for each phase.

I had to reduce the coil section to 2mm x 1.125mm. Probably just as well for manipulation......

How are you getting on?

 

[F4P]

I made a construction decision. I will use a laminated rotor and skip an additional outer shell. My dad, who is working more or less in metall industry, told me that an 150mm tube can have 1mm eliptical missmatch. This is a lot when working with small air gaps.

More about playing around with emetor comes within my thread.

But there is still the question about controller eff open, which will finally define the number of poles for my motor.

 

[Miles]

I made a construction decision. I will use a laminated rotor and skip an additional outer shell.

That has the advantage of being able to machine in the magnet spacing and also, if you are using rectangular magnets, the ability to create facets to get good contact. You're dependent on the interlaminar bond, though.... better bolt through as well.

 

[Miles]

.

Managed to wind all the coils in the correct orientation.... Just need to connect them up correctly, now

.

 

[F4P]

I take one just because of the cool design!

PS: would it make sense to flip the open ends of your turns to the outside. the connection wire wouldn't disturb the flow of the air from the fan so much.

 

[Miles]

Well, I've just realised there's a more serious flaw........... I don't think it can be wound as 2 coil layers, the way it is.......

 

[Miles]

It's AaAabBbBCcCc etc Crossing over from one coil to another the slot opening is blocked.

 

[Miles]

...

 

[Miles]

.
Here I've inverted the coils so that the interconnect no longer blocks the slot opening but now it's impossible to wind the outer layer of the coil because you're winding past the bottleneck at the beginning of the slot... This was why I, naively, started with the coils configured as they were...... Outrunners are easier to wind, for lots of reasons
.

 

[F4P]

Single Coil winding and then welding the copper?

 

[Lebowski]

The clou (<- french word) is to get a certain current density through the slots between the teeth
For instance, with tooth 1 and 2, putting 10A through the space between them can be done by winding
each tooth with 5 windings and putting 1A through the wire. A second option is to wind 10 windings
around tooth 1 and leave tooth 2 empty (and again, 1A through the wire).

Maybe winding only every other tooth with double the wire is easier ? you'll get more end-turn losses though
(but maybe you can fit 13 or even 14 instead of only 12 turns, looking as there's still quite some empty room)...

 

[Miles]

Single Coil winding and then welding the copper?

The problem is that I need 6 turns.. A 9 turn coil with 3 layers would be wind-able, I think.

Single layer coil would need to use two strands. I wouldn't be able to use the rectangular wire.

 

[Miles]

Maybe winding only every other tooth with double the wire is easier ? you'll get more end-turn losses though...

Hi Bas,

That's what is known as a single layer winding. The other drawbacks to it are that the back-emf is less sinusoidal, there is more torque ripple, more harmonic content and greater magnet losses....

 

[Miles]

I'm quite tempted by the idea of making the tooth heads removable..... That way you could pre-wind each phase sub-group off the motor and slip the coils on 8)

 

[madin88]

I do not know much about how to build a BLDC motor, but i like the density of the windings. This looks great.
Maybe winding every single coil and then solder the wire is a good idea. I think this makes winding more easy. Does it?

I have a question Miles. I know this is a bit off topic but many motor experts are represent in this thread so i hope it is ok to ask.

Would it make sense to rewind a Cromotor with that good copper density like in the pic above?
I mainly mean in relation to efficiency at high torque, because lots of energy seems to get lost there. (heat rise up quick during stop and go or during climbing steep hills).
I also would change kV from 9,3 to about 12 for a bit more top speed..

thanks in advance

 

[speedmd]

Hi Miles

Great work on this. Why not twist the outer layer 90 degrees on the end closer to the case. You would only need to do this on one or two wraps to easily capture the area you need. Maybe just on one side. It will help even current flow in the skin also. Dead soft Wire this profile with the generously rounded corners will easily take it. Time to play with some and see what is possible.

 

[Miles]

Would it make sense to rewind a Cromotor with that good copper density like in the pic above?
I mainly mean in relation to efficiency at high torque, because lots of energy seems to get lost there. (heat rise up quick during stop and go or during climbing steep hills).
I also would change kV from 9,3 to about 12 for a bit more top speed..

Yes, if you can increase the slot fill factor significantly, it would be worth doing.

 

[Miles]

Single Coil winding and then welding the copper?

Maybe winding every single coil and then solder the wire is a good idea. I think this makes winding more easy. Does it?

It makes the winding possible but I'm not crazy about the idea of having 37 solder joints....

 

[Miles]

Hi Miles
Great work on this. Why not twist the outer layer 90 degrees on the end closer to the case. You would only need to do this on one or two wraps to easily capture the area you need. Maybe just on one side. It will help even current flow in the skin also. Dead soft Wire this profile with the generously rounded corners will easily take it. Time to play with some and see what is possible.

Hi speed,

I'll model that and see how it works out - it's a pretty tight radius......It would certainly be better than reducing the fill-factor even further.

Yes. Playing with the stator core design isn't going to solve this problem without decreasing the fill-factor, so I might as well go ahead and order the core and start experimenting.

 

[speedmd]

In real terms you will likely end up with a less than true 90 degree twists and a bit of slack at the end turns in spots but you will have the fill where you need it. With this gauge wire resistance/loss from a few extra mm is negligible.

 

[Miles]

.
This is how the connections need to run.
.

 

[Arlo1]

Miles why not come out on the bottom of the tooth? You can wind from top to bottom or bottom to top as long as it goes the same direction clockwise or counter clockwise....

 

[Miles]

Miles why not come out on the bottom of the tooth? You can wind from top to bottom or bottom to top as long as it goes the same direction clockwise or counter clockwise....

Hi Arlo,

With two coil layers, you have to come out the same end of the tooth that you go in. With three coil layers, you would go in and come out on opposite ends.

So, both ends of the coil would have to be on the head end of the tooth, which is where I started....