Neumotor 8057/75 middrive project, belt drive, adding halls, 10kW+

icherouveim said:
madin88 said:
douglashart said:
yes! the Sur-ron setup could well work under the qublix (esp with the neumotor) but im afraid it will only work with the 8038. Even at that, i think it could work.

I wonder why you like to use the Neumotor if you probably could fit the QS2000W or an LMX midmotor below the swingarm.
Mentioned motors have 2-3times the weight, but they have all the benefits of an Inrunner design, and can take way more peak power due to the larger mass.
The QS2000 should work well with my 1:6 belt drive as it has above 45Nm peak torque.

If your pulling 10kw from the 8057, would i be correct in thinking the 8038 good for 6 kw - 3kw continuous?

As for the continuous power it is very hard to tell so far, because the Adaptto controller doesn't work properly, or better say not that efficient sometimes -> which i now believe is a GENERAL ISSUE and not just related to this motor.
One day i was riding up that hill, pulling 8-10kw for some minutes and the Controller was below 45°C and Motor less than 100°
We are talking about similar power to QSV3.
Then the other day the motor did overheat at half of the distance when riding up that hill, and the controller was close before shutdown too :x

I am really looking forward to get the Vasili Controller, but there are many other options too if you don't looking for a unit with display, BMS, charge function etc as i would like to have.

Does Vasili's controller have a good auto fine tune function? Does it adapt ind timing setting according to battery's drop voltage? Im asking because with adaptto controller you said you have to stop once a while to change timing setting and the other one.

ASI controllers I think do this but they don't have on the go options at the moment

If I remember correctly the visili controller and ASI are true FOC, and they, at least visilis controller have amazingly easy autotune, and the best part that it adjust all the timings continuously during your ride to get optimal results. Unlike the old adaptto that is not a true FOC.
 
Allex said:
If I remember correctly the visili controller and ASI are true FOC, and they, at least visilis controller have amazingly easy autotune, and the best part that it adjust all the timings continuously during your ride to get optimal results. Unlike the old adaptto that is not a true FOC.

If no FOC what is it then? A sine-wave modulated controller (as infineon) just with bad ass FET's?
I always thought it is FOC as it was advertised to be the "true sine wave vectored control" IIRC..
 
Hey guys,

over the winter i did a few things on another 8057 motor.
The new motor is handwound and it is 20P (instead of 16P) for higher torque. I also was using stronger and segmented magnets to further improve things. The only concern i have is that it won't have the same sine-wave shape BEMF as the 16P has, but the future will tell..
On the stock motor i separated the windings so that i can hook it up in star too for higher torque per amp (about 45kV).

I also swapped out the Midi-E for a Nucular 12F (which works very well) and added a second battery to the downtube.
The 20s12p 25R battery became support from 5p LG HG2 which gives me 30% more range, or around 3kWh of usable energy :D

I wanted to make the bike look more like a streetbike (for what it was built), and so i lowered the seat height by shortening the damper from 50mm travel to 40mm, and using stronger coils (on the fork too). The battery case on the downtube and cover also makes it look much more like a streetbike now.

btw:
Neumotor added two new, larger outrunners to theire product range, a 120xx and 180xx series.
https://neumotors.com/neumotors-120xx-series-bldc-motors-up-to-40000-watts/
 
larsb said:
Where do you buy the magnets? I got screwed over when i bought magnets for my motor build, dimensions of the magnet were useless. Need to find a reliable seller.

I buy them from here:

Huatai Xinding (Beijing) Metal Materials Co., Ltd.
Add:Dongmafang No.415, ShangZhuang Town, Haidian District, Beijing 100094. China.
Tel: +86-10-53345989; Fax: +86-10-61934084;
WhatsApp / Mob: +86-137 1654 7656 ; Skype / E-mail: max@atechmagnet.com

the contact persons name was Zaria

The motor is not completely done yet as i am still waiting for the stator holder from Neu, but i will share some pics the coming days.
 
larsb said:
Now at least two days have passed :D
Pictures?

:) hey, are you trying to put pressure on me?

The past month i had not much time for hobby because i was not at home for some weeks and then someone from the familiy s building a house. Now the actual state of things is the battery is finished, mounted and wired up. Also the conversion of the bike (to make it look more like a streetbike) is almost finished.
There are a few small things i want to change, like the seat, which is from a Qulbix Q76 and it really sucks!
I thought that i will give it to a saddler and let him add more foam, or a different kind of foam (which is harder).
The other thing is the number plate sometimes is touching the rear fender so i will mount it further back (by using a fender and new holder).

With the construction of the motor i have not made any progress since then, only the rotor, which was already quite challenging becauase of the segmented magnets is basically done:

uc


uc


Next thing will be the winding. I have tried to wind a few coils to see how much room it takes up. This is dual 1.5mm wire for the 5T YY which would increase copper fill by around 20%:

uc


uc


I think it's doable but i never wound a motor before. Whats your thoughts?
What i really like about it compared to the multi strand winding is that there is some room for an airflow which should improve continuous power.
 
On the stock motor i have separated the winding so that it can be wired in star or delta.
The original winding is 5T delta and has a cross section of around 2,8mm²
I did run both winding versions with Nucular 12F controller (first version 200A) in FOC mode. It works very well and efficient, but only thing was that i had to turn down the phase amps a bit to not get any overcurrent cut outs at higher RPM (200A works just until 20kmh). I also did some measurements between the two windings:

delta:
75-80kV, 30-50µH, 22mOhm
it takes around 120W at 3400-3500RPM (with speed limit set)

star:
45kV, 100-130µH, 60mOhm
here it took a little bit less, 115W at 3400-3500RPM (WOT at 78V battery)

Both not bad for a 80x57mm size motor, but i have to say that Neumotor designed the magnetic circuit for up to 8000RPM which has influence on max torque or when the core starts to go into saturation.

uc



Thats how the bike does look now. It feels more stable because of the lower center of gravity and higher mass which probably makes the suspension to work better.
Battery is now 20s 12p 25R + 5p HG2, or with other words it has over 3kWh of juice now :)

uc


uc
 
Bike looks really light since there is so much open area. Do your knee’s fit behind the battery cases or do you have to spread them to ride?

Looks excellent! Great attention to detail.

Tom
 
larsb said:
Winding looks doable, nice with such small slots, there isn't a million turns to do on each!

How many turns you did on the Revolt and which wire did you use?

I would prefer to work with one single rectangular wire instead of using dual wires, but the thing that bothers me is that one guy who has wound hundreds of RC motors was mentioning that when he was using wires larger than 1.8mm in diameter, the motor did heat up more even if it has higher copper fill. Probably because of the skin effect and larger eddy currents within the wire itself.
Thats when it comes to round wires, but i don't know how rectangular wires do compare. In terms of skin effect they could be better i think. Any thoughts?

original fill of my motor is 5T x 2.8mm² = 14mm²

dual 1.5mm wire would be 17,5mm² -> 25% more
dual 1.6mm wire would be 20mm² -> 40% more

here a try with 1.6mm, but if i really go this route with the dual wires i think i will stay with 1.5mm because they more easy to bend and to route between the coils.

uc


litespeed said:
Bike looks really light since there is so much open area. Do your knee’s fit behind the battery cases or do you have to spread them to ride?

Looks excellent! Great attention to detail.

Thanks man!
The scale says 27kg on the front and 24,5kg on the rear so less than 52kg, which is indeed very light for a moped considering that it has a 3kWh battery.
No way a grown up person can ride it with the knees behinde the battery cases, but it is not uncomfortable, yes even more the opposite because it feels similar as if you would sit on a motorcycle having the fuel tank between the knees.
 
Think it was 24T 2.5x1.5 wire from here:
https://www.wires.co.uk/acatalog/rt_ec_wire.html

I checked skin effect at controller switching frequency vs wire resistance increase. Something like this, cannot find the right graph now:
image.png
People here might argue that it's only the erpm frequency that matters but i've read several science papers that showed that high switching frequency creates eddy losses and skin effect in motors so i think this frequency should be used for selecting the wire size.

EDIT:
checking this calculator shows that 2x1 wire is a lot better than 2.5x1.25 (4% higher ac resistance at 20kHz vs 20% for the larger wire)
Capture.JPG
round 1.6mm wire has 22% higher AC resistance than DC resistance so it's too large at 20kHz.
https://www.emisoftware.com/calculator/resistance-rectangular-wire/

-->2x(2x1) rectangular wire would give you 40% extra copper fill with acceptable AC losses if you can fit it
I think i'm buying some myself :D
 
Is this a legitimate 10kW motor in an ebike application? I'm looking for a replacement 100mm motor for my race mini moto.
 
larsb said:
Think it was 24T 2.5x1.5 wire from here:
https://www.wires.co.uk/acatalog/rt_ec_wire.html
If i would use this wire i could do 5T side by side (without doing the last turn on top), and copper fill would be also better comapred to dual 1.5mm
Nice find!
I checked skin effect at controller switching frequency vs wire resistance increase. Something like this, cannot find the right graph now:
image.png
People here might argue that it's only the erpm frequency that matters but i've read several science papers that showed that high switching frequency creates eddy losses and skin effect in motors so i think this frequency should be used for selecting the wire size.

Thanks for the link.
When i think about it then PWM frequency should not make things as worse as if it would be true AC where the entire amount of current is changing it's direction of flow. Because due to the inductance of the motor, the AC will get smoothed out into DC for the most part and only a small portion remains left as ripple current.
So we probably should see it like 90% DC and 10% AC (the exact amount of ripple current could be calculated)
ERPM can be seen as true AC, but at max RPM of this motor the frequency would be less than 1kHz.

I definitely need to do more research before i decide on which size of wire, also in terms of eddy losses.
-->2x(2x1) rectangular wire would give you 40% extra copper fill with acceptable AC losses if you can fit it
I think i'm buying some myself :D

Sounds like a good idea. Need to check if it would fit.

coleasterling said:
Is this a legitimate 10kW motor in an ebike application? I'm looking for a replacement 100mm motor for my race mini moto.

Neu says the 8057 75kV can do 422A max, 22,5kW CONTINUOS and 45kW peak :mrgreen:

Ok some real world experience you can trust in:
First i have put a temp sensor on the windings and set my controller to turn off at 130°C

When i ride on a flat road for a longer time at 80-85kmh, (around 4500RPM / 4kW input / 80A RMS phase) -> temps stay below 80°C
When i climb hills or longer grades WOT (around 4000RPM / 7-8kW input / 150A RMS phase) -> it goes above 130°C after a few minutes and controller is reducing power.
The motor is mounted to an aluminum plate with about 10x15cm which is attached to another one which has about same size.

I don't know what the real potential of the motor is in terms of continuous power since first i don't want to go above 130°C, and second thing is that Nucular controller can only do 150A phase on this motor without cut outs.
I am on the list for a 24F which can be set to 500A and hopefully can do something like 300A on this RC motor then.

district9prawn said:
Wow fancy magnets! Is it very difficult installing them into the rotor?

The three single magnets i used for one pole reject from each other if you want to put them sidy by side. So, to answer your qestion: yes, unfortunately it is.
To get an idea how much work it is to glue one single N-pole, and what i did for easier installation:

1) glue magnet with N facing the stator
2) put magnet with S facing the stator beside, but don't glue it
3) glue magnet with N facing the stator beside
4) remove the one from the middle, turn it around so that N is facing the sator and glue it in

Thats how i proceeded with all poles. The magnets are all rectangluar so they can be used as N or S.
 
Impressive amount of labor you put into the Neumotor. Look forward to hear what you think when motor rebuild is complete and you got it all paired up with the vasili 24 f. Amazing power for the weight and size. That 24f controller might get you into high 20's or even into the 30 kw. That is something. Better ride with your helmet on. You got a moped license for that bike, correct? I think you need to visit TÜV and get it bumped up to A2 ol' 250 cc to be true to the license plate :D
 
Hey macribs, thanks for your reminder that i finally should start to finish this project :mrgreen:

macribs said:
Impressive amount of labor you put into the Neumotor. Look forward to hear what you think when motor rebuild is complete and you got it all paired up with the vasili 24 f. Amazing power for the weight and size.

I got 24F Nucular a few weeks ago and already did a test ride, which was promising on the one hand (controller wise), but a bit disappointing on the other hand (motor wise).
The controller can do around 400A phase stable, but it makes the Motor (@stock) to overheat in no time, and the most disappointing issue is that it doesn't turn it into a wheelie machine as i thought it would. QSV3 or MXUS3k with similar kV would definitely throw me off at 400A, but unfortunately the Neumotor must be in saturation so heavy that it cannot turn such high currents 1:1 into torque anymore. In terms of continuos power it is awesome, but in terms of (peak)torque it is crutch.

I hope that this will change with the 20p rotor with the higher strength magnets. In theory it should be able to do around 30% more torque at same amps which should improve things quite a bit, but without a bench test it's hard to say how high you can go until kT starts to drop.
That 24f controller might get you into high 20's or even into the 30 kw. That is something. Better ride with your helmet on. You got a moped license for that bike, correct? I think you need to visit TÜV and get it bumped up to A2 ol' 250 cc to be true to the license plate :D

I would say that i am quite a safe rider and always wear a Helmet and gloves, but it's the unexpected things which scare me a bit, like when there is dirt or gravel on the street in corners, or when riding virtually silent past parked cars or people. I have not count the number of emergency braking i had because of people crossing the street in town not recognizing me, but it were quite a few.

Yeah, 20kW should be easy possible with 20s17p battery as it would take less than 20A per cell and so even be within specs (with OK voltage sag).
Regarding A2 licencse i did talk with them and they said that any vehicle which has bicylce parts (brake system, frame etc.) will never ever be a motorcyle, which was not what i wanted to hear, but it makes sense.
 
I started to wind the stator. So far three teeth, or 1/6 of the job is done and it took me about two hours.
As wire i have used the one with 1,5mm dia because it is much easier to work with.

The stock motor has 5T with 2,8² so 14² in the slot, and my winding is also 5T with dual 1,75² so 17,5² in the slot.
Thats 25% more copper and still a large gap for air cooling.

kVkHjhLKYadIADAKDvDdVgHKeS3IwGODRNZ_GIxJKJ7oAzuIBi-T24hHlBEylZsD3u73J8tZb50roCWJeYBQ_GGULgOFnxQ7wFFKJCJD-5CyI9G6VL6B8CoS2h2cV9uqF84cWi891lNsOiTJIH4uWacuPDms_ecB-yD3Ttj5RKdemnbjUq0O7IhB-uffA6bRy_XvJftzEnKkxuqKPNDDGVqfkEClyMn93IxnOKZ0zaUOD--dKIYktd-5_lJOpugHElAAElLClR3_HTCYNpgim277d7IZMlxfptr51llAdtOWJrP_NS5aRc5tyDJX4iMX5pBapk6nxPjcI6N0M6ORWjruZNRY4dHSbs_f4cbYWvJbE1hg4kEFoiSemEkjTNW7ONvi0DwG83bHQ5_Q5BTlUWkLw82CsMcQTObDn6bV3cliuZzufwOUQG1OBK9xT-QAlD2M8bs5l_bXkfNOy3-e4zD5c0xEXTaHWBE7KKa7zTmjs3hUImkfF0KLo-ZhpVaMGjuwbrELpN71dVxFaQwLuCDNR-yPoH1y8G8XhNRQFBTrQTTVPynS4VDkgNKp1euJ-UUEmLFoWT23iALlT18mmI-FNVbg0CJD-4pE5mdyRsGnQDtpk2fUHw-vEmRQIG4y2JTvygSSV0cglrjyhU7TZea3H14UZ4UeWr0t1aHiC3XPp6aNuNC9zdI=w800-h526-no


The winding type and location of the hall sensors i have copied from a Bafang geared hub motor, which was also 18N20P and termintated in star.

mKhVVQnuqYPll6eUJKu1jkNodJuKa9cLI0jQ0ZqWi0XOCGeyvf_0gqflhRZPRXMWFZvaw2Xv4CETgp9LsT4arAzPDs-ma7QcN9X4cLJHU6B3gTZ1IxmkLpwqTxlM01m1mMSq4UyVYHOxAdDs2bwkzAqPL28_EqPUQY91NnbVAXgM1PScAiKK0OPbrxfVU_XuNhQH1l0yXYZEM1uUnrAcors5gaJmllwMnX9KHcvMe4TUVgfQU-BODJgfGUzvddWo5hu6aFICe4IdosnP3yep5UEQ3Ftv2MnJSKlcJy_UAinOceksa32PtO8oPfg_ijDEosO16rdXrq1SGLaA8ukkbGnLCl7ky0b_EKx7oJfHXVy7ZLwfBxBKHP_CRaLbi5t_6pU4SeaQXlwBNvy8Kxdfrhxv3QHbnhTL-9aKc5WFFSEQUNTSMKgnfa8KJxuRobOsjV1Lqjh47tzdV-3tPd0Sti6zma4yDrizxt-lNPCmVb266v6InnyWGBV5yoReMSzEwgmUdgPaTB0tQ9hRxP2s8bCvhQpEWZl4D5ckw6vH-3U1QjjBWoBRAW4f3QxeuNIgOT-1lYQtNbLE3YSwkHDBTdGz0x4EBN7hMEPMmE0zg9RqcYcvwToJYQfr-QAVczYs1YO5ze86vcQUBkiJOV7khGMjGnc8-QRu-5_Uhft6ZWsP4XwigFLSK7h7nnP7Kd2RF-iCsXvF0AGyiGOggK2Ks4HNWA1oKXmJHk2aTBFQtAuwKkbQ=w600-h312-no
 
madin88 said:
Here a pic of the resolver. It is 3D printed with ABS so that it doesn't melt at higher temperatures. Don't think that this was that easy.. It took 4-5 redesigns in CAD and prints to make everything fit togehter (especially to get 3x4mm magnets into the slots).
Luckily the printing time was not long due to the small size (took about 15 minutes). For the hall carrier it took only one print.
I soldered a pull up resistor to each sensor as stated in the datasheet, and a small cap to stabilize the 5V supply voltage. This cap may not have been necessary, but i have the meaning you never can have enough caps in the supply lines of any electronics -> the more the better :)

gZFRTP.jpg

This addition of hall sensors looks deceptively simple. I'm looking to add halls to a Neumotors 8038/100. Would the same dimensions work? Is the design optimal? Any chance I could use your CAD?
 
nowholeself said:
This addition of hall sensors looks deceptively simple. I'm looking to add halls to a Neumotors 8038/100. Would the same dimensions work? Is the design optimal? Any chance I could use your CAD?


The motors seem to have identical shaft and mounting dimensions so it should work.
If you give me your email via PM i can send you the two files. Do you need .stl format?
 
coleasterling said:
Timely, I just did this today.

I have my doubts that your solution will work well as the sensors are beside the magnets and facing in axial direction.
But i hope you can prove me wrong since this would be a good idea and your implemention looks great :)
 
Back
Top