Idiots guide to water cooled mid motor making

L

larsb

1 MW
Joined
Dec 10, 2014
Messages
2,843
Location
Gothenburg, Sweden
I will try to improve a revolt RV120 rotor as the original one suffers from losses and heating, suspected to be from eddy currents in the magnets. My plan right now is to do a ironless halbach array rotor just for the fun of it!

fixture and plug stls for 3D-printing:View attachment 2
Rotor plug with shim rings and magnets.jpg


Rotor shell will be made from glass fibre / carbon fibre formed on the outside of the plug,
magnets will be 84pcs of 50x4x4, they should fit the airgap diameter well.

I have abandoned the plans for 50x5x5 magnets, it will be too risky as they only barely fit with an even factor of fourteen at the airgap diameter.

I have tried to find some higher temp magnets but without luck so far..
 
Very exciting stuff. Hugely impressed with your first thread, look forward to seeing how much of a sheen you can get on this turd :lol:

Definitely an old broom situation - one new head and two new handles, still the same broom!
 
Ok i understand your plans.
You plan to build a new rotor without back iron, and instead the shell will be made of composite material.
Lets assume you going to use the stock magnets for that, you would loose about half the air gap field strength. or with other words: the back IRON does the same as if there would be magnets with about twice the tickness (roughly with such thin magnets it can be assumed) as without any steel return. it simply makes them stronger. keep this in mind.
My guess would be that is not what you want.
So either build a steel rotor, or look for magnets with about double the strength.

btw: isn't halbach related to the stator which is ironless (the litz wire winding), and not so the rotor? I'm not sure about the rotor, but referring to the stator there are only the copper windings which create the magnetic field wihout any iron..
 
@madin:
I have thought about the field strength, more on that after vacations, i'll have to dig up my FEMM calculations and try some new ones..
Think revolt magnets are N35, i plan to go to N52, magnets will go from 3mm thickness to 4mm (acc to plans now). No matter what, i won't change the airgap so plug will not be different as to what magnet type i choose.

Halbach is not something specific to ironless stator / litz wire rotors, you can even use backiron with the halbach rotor to increase fields, although it does not make such a big difference as in a normal N-S rotor.

My reasoning for the halbach is:
Halbach is better for losses in general with a more even sinus field.
I will use many segmentations of the magnets for each pole (6-12 segments, depending on magnet i choose) which will cut losses in magnets a lot
If there are losses in the stainless backiron then these are completely avoided.
I'll simulate the difference in fields before i choose final design.

But it is a gamble for sure.. The rewind effort was a near certain win as original winding was so poor, in this attempt i am in deep waters since good motor design is optimisation and balancing of so many parameters.
 
larsb said:
Halbach is not something specific to ironless stator / litz wire rotors, you can even use backiron with the halbach rotor to increase fields, although it does not make such a big difference as in a normal N-S rotor.
Click to expand...

thanks for the explanation. i did not know about the fact that the magnets are twisted 90° for improving field strength, and this is the so called Halbach array. before i was the meaning those are the coreless PM BLDC motors.
well, i should have done this before: http://www.giyf.com/ :mrgreen:

now this means double strength or twice as thick magnets are not needed for same density.
thats all very interesting, and also the part about the closer sine wave form.

larsb, keep on with the great work. It is good to have such smart guys like you here in ES board :)
 
If I understood your plan correct you are skipping the back iron and going for light weight composite? Man you are taking this motor in the direction we all where hoping for when we first learned about the Revolt line. Relatively low Kv motors for single reduction, with light weight, small form factor and impressive power output. The ability to mount such a motor on most swing arms without too complicated modifications or massive cutting and welding was what got me all fired up about Revolt wbw. And here you are making it happen. I tip my hat to you sir. Big kudos for putting lots of thoughts into this, massive hours yet still managing to push forward and keep improving.

When you are finished working with this mod you need to rebuild yet another Revolt and ship it to a chinese motor manufacturer company so they can clone your revolted revolt. Then you got your very own line of electric motors to sell. And the best thing is I am sure it would be possible to scale your finished motor both up and down so motor would fit most every e-bike needs. And if end price is lower then Joby but higher then Revolt I think you got a great product many will take to heat.

I know I would love to see and buy a powerful motor made, refined, tweaked and put to market by a fellow ES'er.
 
Bravo for attempting this :) It's beyond my abilities but great to spectate on. I think building a halbach array would be a first for a DIYer on EndlessSphere?
 
I think you are both too enthusiastic - but hey, thanks!

To get a business together in low volumes is not easy, there'll never be a larsb motor out there for sale :D
The labour cost alone for a slow winder like me, wouldn't want to pay it.. :D
Quality control on the other hand, and a good base design is everything, Revolt dropped the ball on that.

Seems most threads on Halbach diy just died, wonder why? Shouldn't be that hard. I'll certainly give it a shot!
 
larsb said:
Seems most threads on Halbach diy just died, wonder why? Shouldn't be that hard. I'll certainly give it a shot!
Click to expand...

That has been my observation too, hence I'm excited to see your project. You made a great job on the rewind that I think it will be interesting and educational to follow however it works out :D
 
I got very excited about your work yes. Because I know I would not have the stamina to complete such a task myself. I guess half way into winding job project would be abandoned and left, never to be completed with so many great reasons why I could not find the time to complete it. And every time my mind mind would slip and think of the rewind, new arguments for why not to do it would immediately manifest in my brain and the abandonment would see and feel like a very sane and only choice. :D

I know small batches are hard to get manufactured within reasonable price point hence the comment about get it produced in China. But yeah you are right, getting motors manufactured and assembled in China is only part of the equation, CQ would be just as important and probably hard to control without a trusted source in the factory to overlook the QC and making sure care is taken to meet specifications. Otherwise it would just be another revolt.

The other part of course would be to make it economically sound as a business case. Probably most people are not willing to lay down the money for a true high power, light weight, single reduction motor that has proven to work. As most people would be very happy with less then kw peak power.

It is a shame tough that those few motors that meets these criteria has little to no information about e-bike use, setup, matching controllers and in general how to make them perform their best. And there is where I think the niche market might be. Look at Joby, Hacker etc. They got lightweight powerful motors. Not the right Kv for single reduction, but if there was a few good threads here how people have overcome problems, matched correct controller or how they made their reduction gears. Some manufacturers can't even be bothered to answer their emails, or websites contact form if you are not a registered business.

If people could buy a motor that was designed specifically to meed e-biker needs and the motor is represented with various controller choices at various voltages and people could see how to go about making themselves a high power mid drive I think many would consider digging deeper into the wallet or savings to get such a motor. The fact that motor could fit most ready made frame kits, and even regular DH bikes with little or no modifications would the catalyst. With enough information about the motor, what controllers work well and what voltage yields best outcome for certain goals it would be easier for people to decide to take on such a motor.
 
Super! Post a photo of the results!

I'm reading up on halbach in the meantime. Can hardly wait to order magnets since shipping will take a long time.. Either 4x4 or 5x5..but to be somewhat professional i should do some simulations..too much vacation right now
 
Yeah I wouldn't base my design on any of the things Mr Laimer did to his motor. I've seen a video where the motor explode at full rpm.
 
macribs said:
Yeah I wouldn't base my design on any of the things Mr Laimer did to his motor. I've seen a video where the motor explode at full rpm.
Click to expand...

That was an earlier designed and intentionally exploded to see how fast it would spin before it broke.
From what I have seen of Laimers work, he seem knowledgeable. He also have made the best 3dprinted bearings I have seen so far with basic fdm.
 
Hm ok, I guess I got to watch it then. Didn't know he blew up the motor on purpose, I figured it was a design flaw.
 
Will i be going down in the halbach swamp as so many others? I've been reading some old threads here and on other sites and they have been discouraging 8)

If i'm not heard from again then.. But i think the main thing killing good efforts are overtheorizing..
I'll just order the magnets when back from vacation and start building. Worst that could happen is that i scrap the first rotor attempt and start over :) :D
 
subscribed
 
Magnets are now ordered. 120pcs n42M (70pcs needed). Temp rating up to 100degC. Let's see how the shipping goes!
n42M.JPG
This is the seller on aliexpress:
https://www.aliexpress.com/item/120-pcs-N42M-Block-50-x-5-x-5mm-Generator-Magnet-NdFeB-Motor-Magnet-Neodymium-Rare/32564761596.html
 
I was in a hurry and printed those on another printer with not the most level platform, they were reasonably ok, but you might need to grind som imperfections off as the settings were not perfect. Been on vacation for two weeks now and will be home tomorrow. I hope the parts can be used. Otherwise the solution for a more accurate center plug would be lathing.
 
You must log in or register to reply here.
Back
Top