Rewind a cyclone 3kw motor for more power ?

Sparfuchs

1 kW
Joined
Nov 22, 2020
Messages
305
Hello EV friends,
what i want to build is a mid drive DH bike with a motor as strong as possible for fast uphill, awesome acceleration and wheelies but limited to the size of the bestseller cyclone motor. As i build the battery pack by myself i'm not really limited with the power supply. I just need to use the right cells and a strong controller if needed.

I've got a burned cyclone 3kw motor and i thought if i rewire it as powerful as possible it would be great for my DH project.
But where is the limit and how can i get there ?

I counted the wires and turns in my burned motor and it had 12 turns of 12 wires with 0.56mm diameter (0.5mm without coating) each. That's the winding for the bestseller 3kw version.If it works the same with mid motors than it does with hubmotors, i think i know how the turn count influences the performance. But does it influence the behavior of the motor if i use for example instead of 12 wires with 0.5mm per turn, just 1 wire with the same amount of copper (besides the fact that it would take more space and you wouldn't be able to do the 12 turns). The maximum wire diameter i can get through the tight "entry" of the slot is 1,5mm. So i could for example take 4 or 5 of those wires and do maybe 6 or 7 turns i guess. But does it make sense and would i get more torque of it that way ?

Edit: or what about taking the "original" 0,5 wire and instead of 12 of them maybe 20 and then just 7 turns ?

Thanks a lot for your help and i'd also appreciate other solutions of how to get what i want :)

Best regards
Sparfuchs
 
The turn count / Kv ratio does not affect motor performance (much) outside a specific voltage context and use case (desired speed, slope, terrain etc)

Key is to make sure the maximum copper fill is achieved.

Within that context, wind to achieve 15-25% above your desired top speed

at your desired voltage.

No higher speed than that.

If you have no voltage preference, 14S LI is good, nominal 52V, LFP would be 16S

Going to 72V nominal 20S LI, 23S for LFP

would be lower amps, but lose you energy efficiency at low speeds.

Any feedback from the experts here on these gross generalities would be greatly appreciated.

Chalo and spinningmagnets, especially, if you disagree strongly, do please school us in detail if you have time? Or link to past posts, since these paths are so well worn.
 
john61ct said:
The turn count / Kv ratio does not affect motor performance (much) outside a specific voltage context and use case (desired speed, slope, terrain etc)

Key is to make sure the maximum copper fill is achieved.

Within that context, wind to achieve 15-25% above your desired top speed

at your desired voltage.

No higher speed than that.

If you have no voltage preference, 14S LI is good, nominal 52V, LFP would be 16S

Going to 72V nominal 20S LI, 23S for LFP

would be lower amps, but lose you energy efficiency at low speeds.

Any feedback from the experts here on these gross generalities would be greatly appreciated.

Chalo and spinningmagnets, especially, if you disagree strongly, do please school us in detail if you have time? Or link to past posts, since these paths are so well worn.

Thanks a lot for your reply john61ct,

for the maximum copper fill i'd have to buy the smallest wire available. So below 0.1 mm. But if that would be the best way, why do big companies like cyclone decide to use 0.5 or thicker wires ?
 
What do companies make different when they build a 3kw motor and a 6 kw motor if they have the same size ?
Or for example the cyc x1pro gen 2 uses a similar construction to the 3kw cyclone motor but is way smaller and takes up to 5kw. Whats their "secret" ?

As far as i know, if you use a hubmotor with a low turn count you have the advantage that you can reach higher top speed than a 5T could ever do, and can have the same or even more torque than with a 5T motor (assuming you are able to provide more/enough current). So the only disadvantage is, that you HAVE TO provide way more power to get the same torque than the 5T. Right so far ? And a 3T can also handle more kw than a 5T without overheating because it uses thicker wires with lower resistance ? (Not sure about that). Is this all the same with mid drive motors ?

Cyclone managed to get 12x12 = 144 wires with 0.5mm around one core. So i could for example do a winding with the same wires but 24 x 6 turns. Also makes 144 wires. But does that mean the motor could take way more kw than before ? And more important - would that mean more torque if i can provide the needed current ?
 
Sparfuchs said:
Cyclone managed to get 12x12 = 144 wires with 0.5mm around one core. So i could for example do a winding with the same wires but 24 x 6 turns. Also makes 144 wires. But does that mean the motor could take way more kw than before ? And more important - would that mean more torque if i can provide the needed current ?

If the copper fill is the same (volume of wire, basically) then the power capabilities will be the same weather you have 12x12 or 24 x 6. The difference is that the 6 turn will require twice the phase (motor) current to produce the same torque as the 12 turn for the same input (battery) power. More heat is produced by the 12 turn because of the thinner wire used however. The 12 turn will have half the rpm/volt of the 6 turn too.

Someone can present the proper formula to elaborate; something to do with heat = resistance * current
 
You might like to look at buying a new one direct from cyclone-tw.com and using that as a test bed- they are cheap and rugged. From the looks of it there’s a pretty decent copper fill not sure you can improve too much?

The 2-6kw higher power version has a longer body / stator and also steel gears in the gearbox so if that fits could be an option.

Swapping the basic grade sealed bearings throughout for high grade shielded could get you a few percent more efficient if you don’t submerge it, and the stock controller has pretty uneven power delivery so swapping for a more powerful customisable controller could be an option to make the powerband more accessible.
 
Sparfuchs said:
for the maximum copper fill i'd have to buy the smallest wire available
My understanding from reading hundreds of threads on the topic, is that is not true, or perhaps you are taking my stated guideline too literally. Look at the cross section area times count.

You do need to work within a reasonable range of Kv and turn counts.

Only some of those turn count options result in less copper fill, so should be avoided.

The number of turns doesn't affect the motor efficiency. The amount of heat generated for a given level of torque remains the same.

When you change the winding, you change what **ratio of voltage to amps** are needed to make a certain amount of power.

 
pwd said:
Sparfuchs said:
Cyclone managed to get 12x12 = 144 wires with 0.5mm around one core. So i could for example do a winding with the same wires but 24 x 6 turns. Also makes 144 wires. But does that mean the motor could take way more kw than before ? And more important - would that mean more torque if i can provide the needed current ?

If the copper fill is the same (volume of wire, basically) then the power capabilities will be the same weather you have 12x12 or 24 x 6. The difference is that the 6 turn will require twice the phase (motor) current to produce the same torque as the 12 turn for the same input (battery) power. More heat is produced by the 12 turn because of the thinner wire used however. The 12 turn will have half the rpm/volt of the 6 turn too.

Someone can present the proper formula to elaborate; something to do with heat = resistance * current
Thanks a lot for your reply pwd,

So it is exactly the same way i described it with the hubmotor above ? Or would you not agree to some of my thoughts ?

For me it sounds like i would have lots of advantages with a 6T one: more top speed/rpm, less heating, same torque possible (by providing more current). Would you agree with that ?
If so, what would be a good turn count ? 6 was just an example.
 
john61ct said:
Sparfuchs said:
for the maximum copper fill i'd have to buy the smallest wire available
The number of turns doesn't affect the motor efficiency. The amount of heat generated for a given level of torque remains the same.

When you change the winding, you change what **ratio of voltage to amps** are needed to make a certain amount of power.
Thats interesting and i think i've heared it once but forgot. So the same torque means the same heat however the windings are and however the voltage x amps ratio is used to get to the torque ?

So the only advantage of the low turn count is that it can reach more top speed than a hight turn count motor could. And the disadvantage is that you need a stronger battery and controller to provide enough current to get to the same torque of the high turn count motor ?
 
Its called marketing, they do the same thing with 1kw hub motors advertised as 250W to get all the scared UK sales. In your 3kw vs 6kw it could be just the controller, if the motor is literally the same.
But how can you really tell the motors are exactly the same from a website with little info to go on, or vague info in some cases.
Probably not the same 3kw vs 6kw, as its been mentioned, "2-6kw has longer body, steel gears." I know I thought the same thing with the 3kw cyclone vs Axial 4kw cyclone, but the 4kw is indeed different not the same.

For hub motors its the height of the stator/magnets so mxus its 45H for 3kw, Leaf 1500W is 35H.
You want to get maximum copper fill in motors for maximum power. As its been mentioned some winding counts have max fill others do not. I remember reading a post where someone mentioned wrapping square solid copper, if it can be done and wrapped snuggly. Might be worthwhile to find that post, could very well be in the very large MYTH thread talking about turn counts.

What do companies make different when they build a 3kw motor and a 6 kw motor if they have the same size ?
 
markz said:
Its called marketing, they do the same thing with 1kw hub motors advertised as 250W to get all the scared UK sales. In your 3kw vs 6kw it could be just the controller, if the motor is literally the same.
But how can you really tell the motors are exactly the same from a website with little info to go on, or vague info in some cases.
Probably not the same 3kw vs 6kw, as its been mentioned, "2-6kw has longer body, steel gears." I know I thought the same thing with the 3kw cyclone vs Axial 4kw cyclone, but the 4kw is indeed different not the same.

For hub motors its the height of the stator/magnets so mxus its 45H for 3kw, Leaf 1500W is 35H.
You want to get maximum copper fill in motors for maximum power. As its been mentioned some winding counts have max fill others do not. I remember reading a post where someone mentioned wrapping square solid copper, if it can be done and wrapped snuggly. Might be worthwhile to find that post, could very well be in the very large MYTH thread talking about turn counts.

What do companies make different when they build a 3kw motor and a 6 kw motor if they have the same size ?
Thanks a lot for your reply markz,

i know that most motors are labeled with "250w" to be sold as legal, although they can handle way more current. I'm also quite sure that if sellers on aliexpress ect. sell 1000w and 1500w conversion kits they use same motors but different controllers.

And i also understand that cyclone offers different conversion kits with different controllers. But you can buy two versions of only the motor with the exact same size:

Version 1) 1800-3000w/Torque: max.100Nm/Efficiency: 89,6%/Weight: 4,1kg https://www.cyclone-tw.com/product/1/data/11
Version 2) 2000-4000w/Torque: max.130Nm/Efficience: 97,6%/Weight: 5,2kg https://www.cyclone-tw.com/product/1/data/12

So besides all the performance specs.. if the stronger motor has 1,1kg more than the other one without being bigger, it has to be different inside...or a typing mistake. But what could they have possible done to make the inside that heavy ?! I mean, the weight of all the copper wires i pulled out of my burned 3kw cyclone motor were 340 gramm. So with 1.1kg more copper, you could rewind 3 !! more of those motors.

And i've also got one of those 2-6 kw cyclone motors but they are bigger than the compared versions (199.5mm and not 171,5mm)
 
...and just that you understand why i wanna know all those things in detail: After i bought all the insulation material and high grade copper wires and finally got the job done i don't wanna realize that i've missed the chance to make anything better for my needs.
 
Sparfuchs said:
So the only advantage of the low turn count is that it can reach more top speed than a high turn count motor could?

And the disadvantage is that you need a stronger battery and controller to provide enough current to get to the same torque of the high turn count motor

Yes, fewer windings means less torque per amp.

The only reason to use higher turn count is to match the power available to your desired speed range.


 
I would really appreciate some feedback, from anyone that feels pretty certain,

on my "summaries from taking notes" both above and below.

Not just validation to the extent I got it right, but (much more valuable) details on where I'm wrong!

john61ct said:
The turn count / Kv ratio does not affect motor performance (much) outside a specific voltage context and use case (desired speed, slope, terrain etc)

Key is to make sure the maximum copper fill is achieved.

Within that context, wind to achieve 15-25% above your desired top speed

at your desired voltage.

No higher speed than that.

If you have no voltage preference, 14S LI is good, nominal 52V, LFP would be 16S

Going to 72V nominal 20S LI, 23S for LFP
mean would lower amps (a good thing, cheaper too), but also might lose you a lot of energy efficiency if you are usually running low speeds.

 
Oh I dont know but those motors are 2 pieces, the outer is the gear reduction so that might be different. It was so long ago now, 2 yrs before covid. Funny I bought the axial 4kw cyclone and hardly ever used it, did some steep hill testing from a standstill on 20% grade already in 22x36 gear and it climbed like a goat but it was just to loud. If my memory serves me right, which is a gamble to begin with since its been so long ago, the fins were different, and maybe a bit longer. But I'm throwing darts.

Actually was pretty easy to find with 3 pages on Cyclone 4kw
https://endless-sphere.com/forums/viewtopic.php?f=28&t=69867&p=1316505&hilit=cyclone+4kw#p1316505
robocam wrote:
The motor has a 6:1 reduction just like the Cyclone 3000, but it does not have the secondary chain reduction of 3.38:1, so the final reduction of the Cyclone 3000 is 20.3:1 vs the coaxial's 6:1. It is unknown how this will affect things in real life. It may be no issue for commuters while off-roaders might have an issue with the reduced torque.

Also keep in mind that the coaxial is 4000 watts simply because it is run with a 60A controller. If the Cyclone 3000 is connected to a 60A controller, it too would be 4000 watts plus. You can also easily shunt-mod the stock controller to 90A (at 72V it's around 6500W).

The main advantages of the coaxial motor are the claimed reduction in noise and the in-frame motor design. Not having a secondary reduction chain might also result in an increased reliability because there is no chain to drop. Many have ordered this bike, so I'm eager to hear about their experiences.
 
john61ct said:
Sparfuchs said:
So the only advantage of the low turn count is that it can reach more top speed than a high turn count motor could?

And the disadvantage is that you need a stronger battery and controller to provide enough current to get to the same torque of the high turn count motor

Yes, fewer windings means less torque per amp.

The only reason to use higher turn count is to match the power available to your desired speed range.

I don't need to go faster than 40mph so i guess it would make no sense at all to make less turns when rewinding it right ?
I've plans for a hopefully quite effective active heatsink, and usually wanted to use a 20s battery and at least 60 amps controller. As i guess this would make me go way faster than 40mph, would it make sense to overthink my volt/amps balance ?
Would 20s just produce unwanted heat if i'll just use the motor below 40mph or is there still any benefit ?
 
If you rewind the motor to a different turn count, then a different balance of amperage and voltage will be needed.

more turns = you need more volts to make speed X
less turns = you need more amps to make speed X
 
markz said:
Oh I dont know but those motors are 2 pieces, the outer is the gear reduction so that might be different. It was so long ago now, 2 yrs before covid. Funny I bought the axial 4kw cyclone and hardly ever used it, did some steep hill testing from a standstill on 20% grade already in 22x36 gear and it climbed like a goat but it was just to loud. If my memory serves me right, which is a gamble to begin with since its been so long ago, the fins were different, and maybe a bit longer. But I'm throwing darts.

Actually was pretty easy to find with 3 pages on Cyclone 4kw
https://endless-sphere.com/forums/viewtopic.php?f=28&t=69867&p=1316505&hilit=cyclone+4kw#p1316505
robocam wrote:
The motor has a 6:1 reduction just like the Cyclone 3000, but it does not have the secondary chain reduction of 3.38:1, so the final reduction of the Cyclone 3000 is 20.3:1 vs the coaxial's 6:1. It is unknown how this will affect things in real life. It may be no issue for commuters while off-roaders might have an issue with the reduced torque.

Also keep in mind that the coaxial is 4000 watts simply because it is run with a 60A controller. If the Cyclone 3000 is connected to a 60A controller, it too would be 4000 watts plus. You can also easily shunt-mod the stock controller to 90A (at 72V it's around 6500W).

The main advantages of the coaxial motor are the claimed reduction in noise and the in-frame motor design. Not having a secondary reduction chain might also result in an increased reliability because there is no chain to drop. Many have ordered this bike, so I'm eager to hear about their experiences.

Thanls for looking for this information.

Meanwhile i'm quite sure the two mentioned kits are the exact same motor because cyclone offers the two different kits and lunacycle sells just the 3kw motor with 40 or 60 amp controller (the controller that makes the kits different).

The coaxial motor looks a bit different because it has a different gearbox than the 3kw with the planetary gearbox.
 
neptronix said:
If you rewind the motor to a different turn count, then a different balance of amperage and voltage will be needed.

more turns = you need more volts to make speed X
less turns = you need more amps to make speed X

In practice doesn't this mean, the lower turn count has a higher top speed?

And for torque and efficiency considerations, it is not advisable to design for a top speed higher than you plan to use?
 
neptronix said:
If you rewind the motor to a different turn count, then a different balance of amperage and voltage will be needed.

more turns = you need more volts to make speed X
less turns = you need more amps to make speed X

Thanks a lot for your reply neptonix,

so that means that it doesn't matter at all how i would rewind the motor, or how i balance volts and amps, there is no way (exept better cooling) to get more torque out of it without burning it ? So all i can do is decide HOW i want to reach the max. possible torque ?
When i started the thread i thought of using thicker wires/ less turns to be able to push more current into it to get more torque. But in fact i would just have to get stronger,bigger and more expensive components and wouldn't even get more torque but just more top speed than i want or need. Are my thoughts right so far ?
And does the balance (volt/amps) i use to get a certain amount of Watts influence heating or does 5000W always mean same heat no matter how i get to it ?

Thanks in advance
 
More turns means a lower kv rating and will decrease motor speed while increasing torque. It also means slightly more resistance in the windings which lowers the maximum wattage that the motor can handle.

I'm not an expert and dont rewind my own motors, but I've had a few done with a single strand wiring and lower kv/higher voltage and was satisfied with the results. HV setups are typically more efficent. Double the voltage and half the amps produces less heat from the motor/batteries and esc.

If you really want to do a deep dive into understanding brushless motors this is a good source of information.

https://www.rcgroups.com/electric-motor-design-and-construction-361/
 
john61ct said:
In practice doesn't this mean, the lower turn count has a higher top speed?

And for torque and efficiency considerations, it is not advisable to design for a top speed higher than you plan to use?

Of course, on the same voltage with lower amps you will see a higher RPM unloaded but it will also want to eat more amps, so you may not reach the potential top speed unless your battery can feed it the amps.

For torque and efficiency considerations, it's a tough call since we can't rely on the ebikes.ca simulator to find the sweet spot in a dyno graph.. you would have to assess this yourself using some before and after data.

Some motors get more efficient at higher RPM, some get less vs stock. Depends very much on the pole count and laminations.

it's a lot easier to change the voltage than change the winding to find things like this out ;)
 
Sparfuchs said:
...For me it sounds like i would have lots of advantages with a 6T one: more top speed/rpm, less heating, same torque possible (by providing more current). Would you agree with that ?
If so, what would be a good turn count ? 6 was just an example.

The lower turn count would provide more speed/rpm but not less heat for the same torque. The heat would be roughly the same as a higher turn count version (same copper fill) when producing the same torque. The only time you would get less heat is with more copper. I need to find that formula... ok got it. Heat = I^2 X R or Heat = motor current squared multiplied by the motor's phase to phase resistance.
 
Sparfuchs said:
so that means that it doesn't matter at all how i would rewind the motor, or how i balance volts and amps, there is no way (exept better cooling) to get more torque out of it without burning it ? So all i can do is decide HOW i want to reach the max. possible torque ?


The only way to get more torque (at the wheel) without generating more heat is to have a larger reduction ratio. Increasing reduction will require less motor torque to generate the same wheel torque and the reduction in motor torque will generate less heat.
For a deep dive into these subjects read this thread:

endless-sphere.com/forums/viewtopic.php?f=2&t=64907

Avner.
 
There's no real free lunch in switching windings. Especially when you have a motor with gear reductions inside.

Spin at a higher rpm and you get more gear friction. Go far enough into this and you'll actually lose power capability and efficiency to friction becoming a progressively larger % of the load.

Go in the opposite direction and you don't spin the motor fast enough, so it never can reach peak power or peak efficiency.

The window of an ideal balance between all motor design considerations, where you maximize peak power and continuous efficiency is extremely narrow.

The headline motors that come in the cyclones have very high stated efficiency. I believe they are optimized designs, other than the room for improvement in bearings, gears, and lubrication, which yields you a few % extra power and efficiency.

My best guess is that a faster winding will result in less efficiency here. That is purely a guess given that i don't have dyno sheets for various voltages on these headline motors.
 
Back
Top