Same debate again, motor windings!

Electric Motors and Controllers
Post Reply
User avatar
sn0wchyld   1 MW

1 MW
Posts: 1668
Joined: Mar 18 2011 8:27pm
Location: South Aus.

Same debate again, motor windings!

Post by sn0wchyld » Oct 22 2019 6:37pm

Brought this here so as not to clog a sales thread any more...
original thread:
viewtopic.php?p=1504502#p1504502
Theres a thread somewhere on here that discusses the below issues at length - ill post myself if I can find it, unless someone else does first?
ElectricGod wrote:
Oct 17 2019 12:43am
sn0wchyld wrote:
Oct 16 2019 9:31pm
ElectricGod wrote:
Oct 16 2019 6:52pm
Clearly there is desire for higher voltage. I want this too. I'd have bought 2 of the 24 fet version that worked at 150v max if they existed.
higher voltage controllers (mosfet based) are really only of benefit if you have a slow wind motor that is significantly underutilized at 80VDC, that you cant rewind (for whatever reason)... at least until you get to really big power levels where conductor sizes and breakers can get impractically huge (we're no where near such sizes here). I have one such motor that I haven't yet got round to rewinding (too many projects!) from its 17kv now to ~50kv - a higher voltage controller would be good and save me time/effort, but still not as good as a rewind for higher kv (80VDC mosfets are the highest power density, for example). Runnning 500A conductors really isn't so bad compared to dealing with 300VDC :P. One is just a bit more copper, the other is properly deadly. One is transient, the other is perpetually hazardous.
Not true...depends on what you are doing and why.

1. The hubmonster, when they were being manufactured were put on scooters and ran at 60v. Despite that, this motor does much better at 130v than it does at 60v and might be even better yet at 200v.

2. "Slow wind" indicates low Kv or lots of turns per stator tooth. This means more resistance in the windings compared to motors with a higher Kv.

3. All motors at higher voltage have a broader power band than they do at lower voltages. This is quite advantageous to have. There's excellent reasons why electric car makers use 380v.

4. Some motors due to the iron losses do poorly at higher voltages. I have a couple of large inrunners. They heat up running above 82v. These motors are a bad option above 82v. Others do great at higher voltage.,

5. Any motor is capable of only so much wattage. Current is what heats up a motor, not voltage. So to maximize motor wattage, run it at higher voltage and less phase current so it heats up less.

6. I've done rewinds on outrunners. I always get more copper on them than they had from the factory.

7. I think you need to read up about electricity. Current kills, not voltage. How many times have you been shocked by static electricity? That's voltages in the 100,000 to millions of volts range, but may be 10mA of current. Compere that to 48v and 400 amps...which will cook you to a crisp in seconds

So yeah...150v Nucular controllers would be way cool! I'd be running them at 131v just as soon as I could build the battery pack!

1. The hubmonster, when ...
This is why i mentioned rewinding motors - did you read my post at all? Yes, wound for 150+V they need... 150+V to realize their full potential. Rewind them for 80V will achieve the same as running them at 150V, only your controller will be better, and voltages (marginally) safer.The motor im talking about above is a mini monster after all...

2. "Slow wind" ...
ture, but irelevant when it comes to torque production. If you halve the Kv, you halve the turn count - so you can double the parallel wires, ie halve the reistance. Then you've only gone half as far for each winding, so you end up with 1/4 the reisistance... so the same copper losses per unit of torque produced. I^2R=P after all.

3. All motors at ...
No, read above. if you change the winding, then you end up at the same place. Cars use higher voltages because they have the funds to pay for safety systems, and at the same time 6+kA is hard to manage vs such safety systems. We're not talking about 500kw systems though - not even 1/10th of that. Im guessing you also skipped over where I said 'MOSFET BASED'? IGBTs are a different story/beast, as are the power levels involved.

4. Some motors due to the iron losses do poorly at higher voltages.....
Again, its not about voltage, but about the rpm they hit for a given voltage AND winding. Rewind those large in-runners with double the turns per tooth and they'll do fine at 82+v (assuming the motors do fine at 41V). Their torque capability wont shift at all though.

5. Any motor is capable of only so much wattage...
You just gave an example where maximizing voltage caused issues... see your own point #4. You can get a motor and match it to a controller with the right voltage. or you can get a controller with the right voltage and match it to a motor. The highest power density controllers are ~80VDC, so its (often, not always) worth getting or rewinding a motor to suit such a voltage, at least until >50-100kw. The relative simplicity of lower voltage is just a added bonus. There's a reason you can easily get a 12vdc switch to break 1000A (12kw) but relatively hard to find a 120Vdc swith to break 100A (still 12kw) and even harder to get a 1200Vdc switch to break 10A (still... 12kw). DC is a bitch to break as voltages rise, it maintains an arc easily and for a long time, particularly when coming from low impedance sources like batteries.

6. I've done rewinds on outrun.....
Not sure your relevance here? more copper fill is almost always good. Not relevant to rewinding to suit different voltages

7. I think you need to read up about electricit....

How do you get current to flow through something, like say... a motor? is it... voltage?
The fact is you need BOTH to be dangerous, because one is dependent on the other - and the quantity needed is dependent on the situation. 400VDC batteries are incredibly dangerous, because 400VDC is enough to get the current flowing through your skin, once the skin barrier breaks down your a dead man, because we are basically sacks of impure (ie highly conductive) water, covered in a thin layer of rather poor insulation (skin). 80VDC is far safer (Though still potentially deadly, ie if your hands are sweaty or you have a cut thats bleeding) because with dry hands you'll probably be fine to touch the terminals, though if you lick them say good by to your tongue. V=IR after all...
Me thinks it is you who need reed up on electricity (particularly low source impedance DC), young grasshopper :P


So yeah...150v Nucular controllers would be way cool!.......
The hubmonster...a 6 phase hub motor does great at double its design voltage (60v). It does fine at more than that. I have a couple of outrunners that do great at better than 40+ over their spec'd voltage. Just depends on the motor.


Again, and hopefully for the last time... only if you cant/dont want to rewind a motor. 150V controllers wont have the same power density as the current batch of ~80V controllers either, they'll be larger per kw (all else being equal). Rewind your motor with ~60-70% of the turns it currently has and run it at 80V with the current batch of controllers, and you'll be in a better place than waiting on a 150V controller - both because you'll have a higher power density controller, an easier/safer voltage to deal with, and you'll be able to get it by next week if you really want/have some time, just buy some wire and your good to go... vs ???? months/years until a 150V controller is developed (if ever).

Not that anybody will ever acknowledge this, but for a while in the 1990's I worked for a company that rebuilt and serviced industrial breakers. The smallest breaker I serviced was rated for 3 phase AC at 120v per phase and 600 amps.
Cool story. I've had electricians working for me that, despite decades of experience in AC and DC systems couldn't hook up 4 lead batteries in 2S2P configuration, calling me in confusion as to why they were getting 0V rather than 24V... despite having instructions, schematics and photographs of built units in hand. Experience is only useful if its relevant and you learn from it... and im unsure what experience with AC mains breakers in particular has with DC batteries, controllers and motors and how they interact WRT the motor windings. They are very different beasts other than the movement of electrons.



Happy to answer if you have further questions, but otherwise ill go the way of sam, and only answer this 'argument' for the sake of other readers. Vasilli, love ya work mate. Hopefully this saves you some work given you're already struggling to keep up with demand!!!

district9prawn   100 W

100 W
Posts: 264
Joined: Jul 20 2013 10:46pm
Location: Sydney

Re: Same debate again, motor windings!

Post by district9prawn » Oct 22 2019 8:04pm

ElectricGod wrote:
Oct 17 2019 12:43am
5. Any motor is capable of only so much wattage. Current is what heats up a motor, not voltage. So to maximize motor wattage, run it at higher voltage and less phase current so it heats up less.
I feel like this interpretation of voltage and current is where a lot of misunderstanding comes from. Motor torque depends only on phase current. To make X nm of torque requires the same phase current regardless of battery voltage. Of course the battery current is less on the higher voltage system. Controller (mosfet) heating is also based on phase current and not battery current.

EG, if you did indeed mean battery current and not phase current, my apologies.
Neu 8057 6kW left hand drive hardtail with 18 fet Vesc: viewtopic.php?f=6&t=96754

User avatar
sn0wchyld   1 MW

1 MW
Posts: 1668
Joined: Mar 18 2011 8:27pm
Location: South Aus.

Re: Same debate again, motor windings!

Post by sn0wchyld » Oct 22 2019 11:11pm

district9prawn wrote:
Oct 22 2019 8:04pm
ElectricGod wrote:
Oct 17 2019 12:43am
5. Any motor is capable of only so much wattage. Current is what heats up a motor, not voltage. So to maximize motor wattage, run it at higher voltage and less phase current so it heats up less.
I feel like this interpretation of voltage and current is where a lot of misunderstanding comes from. Motor torque depends only on phase current. To make X nm of torque requires the same phase current regardless of battery voltage. Of course the battery current is less on the higher voltage system. Controller (mosfet) heating is also based on phase current and not battery current.

EG, if you did indeed mean battery current and not phase current, my apologies.
my presumption was this was the 'volt up, gear down' mantra - ie by gearing down less ph amps are required for the same torque output (at the wheel) - and then volt up to compensate for the 'higher' gearing to achieve the same top speed.

Point being you can achieve the same thing by re-winding the motor, keep the same battery, same controller and end up with the same benefits.

larsb   100 kW

100 kW
Posts: 1291
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Same debate again, motor windings!

Post by larsb » Oct 24 2019 10:21am

Thumbs up to both of you, nicely explained.
Theres a thread somewhere on here that discusses the below issues at length - ill post myself if I can find it, unless someone else does first?
I’ve thought about doing a thread about ”the myth about the myth” for a long time. The myth thread is this one: viewtopic.php?f=2&t=64907#p974291

I agree on all above from sn0wchyld; winding has to be matched to controller and battery, there is a rated rpm for each motor that has to be reached for full performance. Either with high voltage/low kV or the opposite.

But there is more to add that i realised after winding a bit: winding transitions. A low turn motor (even with same copper fill goal) will have relatively less active vs inactive copper.

1) bunching and stacking of wires in the slots
It will not be possible to get the same copper fill on low turn motors since wires bunch up at the entry and exit of the teeth. More parallell wires, more bunching. It won't be possible to fill the slot entirely either due to the large bundles of parallel wires that cannot be stacked in an optimal way.
7CE8F74F-5D44-4C61-8E63-BDE05D9582D4.jpeg
7CE8F74F-5D44-4C61-8E63-BDE05D9582D4.jpeg (111.26 KiB) Viewed 357 times
From left to right slot in picture, absolute max theoretic fill factor figures below.
Round wire dia 1.5mm:
58 turns per tooth, pi*1.5^2/4*58*2/267= 77%

Square wire 2.5x1.25:
32turns per tooth, 32*2.81597*2/267= 67%

bundled wire 7x1mm dia:
28 turns total 28*7*pi*1^2/4= 58%

—-> stacking of large bundles will lower fill factor.

The bunching at entry and exit of the teeth is not so easy to value, it’s substantial for the really low turn counts like 1, 2 and 3 turn windings.
Compare 10parallel*3turns winding to 3parallel*10turns and it gets pretty obvious.

2) Teeth winding and transitions between teeth
Torque vs loss comparison for the teeth:
4 turn*1-wire winding, current I for a given torque T
1 turn*4-wire winding, same copper fill, current needs to be 4*I for same given torque T.
4x parallel wires --> 1/4 of the resistance and 1/4 length copper --> total 1/16 of the resistance for the 1 turn winding makes the losses around the tooth the same as the 4 turn. Nothing new from above posts.
A9751881-CE8A-4818-B6C4-D9C97D6672CF.png
A9751881-CE8A-4818-B6C4-D9C97D6672CF.png (10.12 KiB) Viewed 661 times
Transition wire losses on the other hand.. Length of transitions are the same for both windings so R*I*2 losses for the transitions are 4x higher for the 1T winding than for the 4T winding at the same power output.
(Current 4*I, resistance of transitions R=1/4 --> (1/4)*(4)^2)=4)

It will make a difference in efficiency. For single turn motors a lot, i’d guess several percent at full current. Difference will be decreasing/diminishing with higher turn counts. One might think ”who the hell runs such low turn motors as 1,2,3T?!”
I can name some names here on ES :D
Last edited by larsb on Nov 04 2019 1:03am, edited 5 times in total.
Ride on :D

User avatar
sn0wchyld   1 MW

1 MW
Posts: 1668
Joined: Mar 18 2011 8:27pm
Location: South Aus.

Re: Same debate again, motor windings!

Post by sn0wchyld » Oct 24 2019 4:57pm

larsb wrote:
Oct 24 2019 10:21am
Thumbs up to both of you, nicely explained.
Theres a thread somewhere on here that discusses the below issues at length - ill post myself if I can find it, unless someone else does first?
I’ve thought about doing a thread about ”the myth about the myth” for a long time. The myth thread is this one: viewtopic.php?f=2&t=64907#p974291

I agree on all above from sn0wchyld; winding has to be matched to controller and battery, there is a rated rpm for each motor that has to be reached for full performance. Either with high voltage/low kV or the opposite.

But there is more to add that i realised after winding a bit: winding transitions. A low turn motor (even with same copper fill goal) will have relatively less active vs inactive copper.

1) bunching
It will not be possible to get the same copper fill on low turn motors since wires bunch up at the entry and exit of the teeth. More parallell wires, more bunching. This is not so easy to value, it’s probably substantial for the really low turn counts like 1, 2 and 3 turn windings. Compare 10parallel*3turn winding to 3parallell*10turn and it gets pretty obvious.

2) Teeth winding and transitions between teeth
Torque vs loss comparison for the teeth:
4 turn*1-wire winding, current I for a given torque T
1 turn*4-wire winding, same copper fill, current needs to be 4*I for same given torque T.
4x parallel wires --> 1/4 of the resistance and 1/4 length copper --> total 1/16 of the resistance for the 1 turn winding makes the losses around the tooth the same as the 4 turn. Nothing new from above posts.
A9751881-CE8A-4818-B6C4-D9C97D6672CF.png
Transition wire losses on the other hand.. Length of transitions are the same for both windings so R*I*2 losses for the transitions are 4x higher for the 1T winding than for the 4T winding at the same power output.
(Current 4*I, resistance of transitions R=1/4 --> (1/4)*(4)^2)=4)

It will make a difference in efficiency. For single turn motors a lot, i’d guess several percent at full current. Difference will be decreasing/diminishing with higher turn counts.

One might think ”who the hell runs such low turn motors as 1,2,3T?!”
I can name some names here on ES :D
Completely agree - end turn losses are marginal as end turns are still contributing to usefull flux, just not as much, and transition losses are (usually) marginal as they are a small % of the total (even less so as turn counts go up as you say) and the extra losses in both cases are only increased in proportion to delta I... as you still get double the conductor dia, just no longer half the length for those bits, so (fortunately) still no ^2 relationships which hurt so much... Just didn't bother to put all the caviates in as it detracts from the underlying point even as it somewhat undermines it hahah :P trying to explain the Forrest and ignore the trees for now, lest the trees block the view of the forrest.

Punx0r   100 GW

100 GW
Posts: 5307
Joined: May 03 2012 8:16am
Location: England

Re: Same debate again, motor windings!

Post by Punx0r » Oct 24 2019 5:28pm

sn0wchyld wrote:
Oct 24 2019 4:57pm
end turns are still contributing to usefull flux
Are they?

User avatar
sn0wchyld   1 MW

1 MW
Posts: 1668
Joined: Mar 18 2011 8:27pm
Location: South Aus.

Re: Same debate again, motor windings!

Post by sn0wchyld » Oct 24 2019 7:51pm

Punx0r wrote:
Oct 24 2019 5:28pm
sn0wchyld wrote:
Oct 24 2019 4:57pm
end turns are still contributing to usefull flux
Are they?
Yes (well... afaik/iiuc) ... a magnetic field forms around a current carrying conductor. the fact that this conductor makes a right angle turn, or runs 'beside' rather than 'between' some iron doesn't change this fact. If such a turn results in a air gap, then some of this flux inside the coil is wasted, but none the less some will still interact with the iron in each stator tooth, and increase the flux density within it. It is not captured anywhere near as well as it is between each tooth, but it still has some interaction with the rotor, and thus isn't entirely useless/wasted. Imagine a really long stator tooth, one that is wider (measuring along the circumference) than it is measuring axially. Does the section of winding running along the side of the stator tooth contribute nothing to the flux density captured and directed by the tooth? or just less than the sections between each tooth? Alternatively think of a perfectly round tooth - it'll give you a larger effective 'air gap' but no part of the winding is contributing more, or (most importantly) no part is contributing nothing to the flux within the tooth.

major   10 kW

10 kW
Posts: 788
Joined: Mar 10 2013 10:12am
Location: NW Ohio, USA

Re: Same debate again, motor windings!

Post by major » Oct 24 2019 8:57pm

Punx0r wrote:
Oct 24 2019 5:28pm
sn0wchyld wrote:
Oct 24 2019 4:57pm
end turns are still contributing to usefull flux
Are they?
No.

Been through this on this forum before. I'm not too familiar with this forum search feature. Some recent white papers on the subject show negligible results. Or study Ampere's Law.

Regards,

major

Punx0r   100 GW

100 GW
Posts: 5307
Joined: May 03 2012 8:16am
Location: England

Re: Same debate again, motor windings!

Post by Punx0r » Oct 25 2019 4:12am

Sorry, my post was a bit tongue-in-cheek - I thought I was going to have to find that massive, angry ~40 page thread about it from a year or two ago :wink:

kdog   10 kW

10 kW
Posts: 598
Joined: Feb 02 2014 9:22pm
Location: hobart, tasmania

Re: Same debate again, motor windings!

Post by kdog » Oct 25 2019 5:55am

I learnt a hell of a lot from that thread.... One of the best here. I was glued to it. Probs 3-4 years ago now . Time flies (with the right winding).

larsb   100 kW

100 kW
Posts: 1291
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Same debate again, motor windings!

Post by larsb » Oct 25 2019 1:37pm

I wanted to check what the transitions do for a higher turn motor. Revolt rv120 has about 12cm of transitions per phase. Let’s pretend it’s star connected for simplicity :D then two phases are active at the same time with 24cm of transitions.

Original winding is 5*12T of 0.8mm dia wires. If we compare some windings and match the current to the same torque:

Resistance is 1.6087 mohm for the 5*12t winding transitions—>
RI2 losses @ 200A (which is about max for this motor) is 64W. System voltage 72V to get to the max rpm.
RV120 winding calc.JPG
Readily available controller combinations marked in blue
RV120 winding calc.JPG (40.88 KiB) Viewed 508 times
We always say that the turn count can be matched with the correct controller and battery to get same performance but it has some boundaries, eh?

Already with the 6 turn winding it’s harder to find a suitable >36V 400A controller. The real point is that motor winding must be matched with controller and battery and be able to get to the max rpm to get max power - within reasonable volts and amps :D

One thing i thought about:
A normal power/rpm curve has the peak power before max rpm since BEMF lowers the current that flows in the winding. That should mean that it’s possible to use a higher voltage (or higher amp&kV) system with rpm limited by the controller in order to increase max power and have max torque at all speeds. Wheelies at 100km/h :D
Ride on :D

User avatar
sn0wchyld   1 MW

1 MW
Posts: 1668
Joined: Mar 18 2011 8:27pm
Location: South Aus.

Re: Same debate again, motor windings!

Post by sn0wchyld » Oct 29 2019 6:02pm

larsb wrote:
Oct 25 2019 1:37pm
I wanted to check what the transitions do for a higher turn motor. Revolt rv120 has about 12cm of transitions per phase. Let’s pretend it’s star connected for simplicity :D then two phases are active at the same time with 24cm of transitions.

Original winding is 5*12T of 0.8mm dia wires. If we compare some windings and match the current to the same torque:

Resistance is 1.6087 mohm for the 5*12t winding transitions—>
RI2 losses @ 200A (which is about max for this motor) is 64W. System voltage 72V to get to the max rpm.
RV120 winding calc.JPG
We always say that the turn count can be matched with the correct controller and battery to get same performance but it has some boundaries, eh?

Already with the 6 turn winding it’s harder to find a suitable >36V 400A controller. The real point is that motor winding must be matched with controller and battery and be able to get to the max rpm to get max power - within reasonable volts and amps :D

One thing i thought about:
A normal power/rpm curve has the peak power before max rpm since BEMF lowers the current that flows in the winding. That should mean that it’s possible to use a higher voltage (or higher amp&kV) system with rpm limited by the controller in order to increase max power and have max torque at all speeds. Wheelies at 100km/h :D
great post mate - 5% variance is higher than I'd have thought. That said, its 5% change in efficiency (ignoring transitions and just looking at end turns right? so a underestimation if so) across a kv/voltage/amp range of 6000% - so a relatively marginal change in efficiency for a pretty extreme change in kv etc in anyone's book i think hahah. I was more referring to the benefits of a 80v controller vs a 150v controller, but yes outside of that the ratio can change more dramatically. Interesting too that the benefits from a very high turn count can offset the relatively high switching losses of a IGBT based controller (vs MOSFET). 'tis why I like this forum, even when your wrong (or not quite right) you can learn something.

For interests sake, what is the ratio of transition to 'useful' wire? I'm curious how the net copper losses vary by winding?

I did find a great thread by the way, really great info and changed my thinking on the subject (prompted by major).
viewtopic.php?f=30&t=93145&p=1506819#p1506819

larsb   100 kW

100 kW
Posts: 1291
Joined: Dec 10 2014 5:12am
Location: Gothenburg, Sweden

Re: Same debate again, motor windings!

Post by larsb » Oct 30 2019 12:41am

This calculation is only for the transitions between coils, not endturns. I’ll add in the total copper losses.
For interests sake, what is the ratio of transition to 'useful' wire? I'm curious how the net copper losses vary by winding?
Ride on :D

Post Reply