More than3 phases

Electric Motors and Controllers
Steph   10 µW

10 µW
Posts: 5
Joined: Nov 13 2018 5:50pm

More than3 phases

Post by Steph » Nov 14 2019 7:57am

Hello,

To reduce the need for reduction, I tend to look for low kv motors (as a way to get more torque, less speed)
So far, my perception (possibly a collection of misconceptions) is that this can be achieved by two kind of design tweaks:
- playing with copper winding (balancing same mass of copper as many turns vs less turns bigger wires). Overall the power output is the same but yielding either more speed or more torque. There are some discussions on this forum on this trade off.
- playing with the number of poles: again, same power envelop but using a cycle to have more or less angular displacement. Again, some discussions (which I didn’t all read in details yet)

Now, I’m curious about another cursor to play with.
My insight is that winding a motor with a higher number of phases (like 5, 7... Is an even count >2 a working option?) would result in more torque and less speed for the same power output.
I tried my luck with google but it seems the topic is rather confidential (mainly one 2014 slide deck but quite high level). I’m interested in any thoughts, discussion or pointers to relevant knowledge.
Obviously, multiplying the phase would result in more commutation complexity. This complexity itself is not really my focus here except if in term of practical design it induce negative effects on overall efficiency.

Cheers,
Stéphane

User avatar
stan.distortion   10 W

10 W
Posts: 94
Joined: Jun 12 2019 12:53pm

Re: More than3 phases

Post by stan.distortion » Nov 14 2019 9:08am

Any number of phases is possible, stepper motors commonly use 2 but 3 is about the minimum for smooth torque. It should be possible to wind many 3 phase motors to use multiple ESCs and it should also be possible to rearrange slot/poll combinations for 6 or 9 phases (with 2 or 3 ESCs), possibly useful when hitting ESC frequency limits with high poll count motors.

User avatar
neptronix   100 GW

100 GW
Posts: 14130
Joined: Jun 15 2010 5:56pm
Location: California refugee living in Utah, USA
Contact:

Re: More than3 phases

Post by neptronix » Nov 14 2019 10:31am

Dyno graphs show that you don't get a benefit from more phases. The real downside is that you have to design a custom controller around 4,5,6,7 phases..

All the highest efficiency, highest power density, and most expensive motor designs use 3 phases. Take note!

More or less poles can be used to optimize initial torque or iron losses; it's a variable you wanna play with depending on what RPM the motor needs to operate at. Really high RPM typically means less poles are ideal.

Winding turn count determines mostly RPM per Volt, aka KV. Sometimes a higher turn count can translate into a little more efficiency/power output, but not always.

Copper fill determines the amount of torque a given size stator can produce. Too much, and you lose efficiency at low to mid loads to copper losses, but you can get higher efficiency at heavy loads. Too little, and you lose peak power, but have excellent efficiency at low to mid loads. What's ideal depends on your application.. IE, does your load have a lot of variability in it ( DD hub ), or relatively steady ( mid drive ) ?
"Love and compassion are necessities, not luxuries. Without them, humanity cannot survive." - Dalai Lama

My first major build: 1.6kW 8T MAC motor on a Trek 4500.
The new all-arounder: Leafmotor 1500w @ 4kW on a Turner O2 full suspension.
The monster scooter: 20" eZee on a Cannondale Semi Recumbent.
Whipper-snapper: ? on a lightweight BikeE Semi Recumbent

User avatar
Toorbough ULL-Zeveigh   100 MW

100 MW
Posts: 2528
Joined: Feb 09 2007 3:02am
Location: Marlboro

Re: More than3 phases

Post by Toorbough ULL-Zeveigh » Nov 14 2019 7:13pm

yeah, you want odd number of phases, can't remember why exactly; mutual saturation i think.
so the next odd multiple of 3 would be 9 phase to possibly use multiple existing 3 phase inverters.
https://www.danatm4.com/blog/5-benefits ... -inverters
5 Benefits of Multi Phase Motors Inverters

1. Increased power
2. Reduced component costs and DC bus ripple
3. More efficient use of cable cross-sectional area
4. Easier integration
5. Cost
https://patents.google.com/patent/US6791226B1/en
https://patents.google.com/patent/US9774290B2/en
neptronix wrote:
Nov 14 2019 10:31am
The real downside is that you have to design a custom controller around 4,5,6,7 phases..
they said the same thing when 16 bit processors hit the market, have to be 8 bit backward compatible, installed s/w base, blah blah...
now they can't obsolete previous gen fast enough.
Kick down the barricades Listen what the kids say.
From time to time people change their minds But the Frock is here to stay.
I've seen it all from the bottom to the top Everywhere I go the kids wanna Frock.
Around the world or around the block Everywhere I go the kids wanna Frock.

User avatar
neptronix   100 GW

100 GW
Posts: 14130
Joined: Jun 15 2010 5:56pm
Location: California refugee living in Utah, USA
Contact:

Re: More than3 phases

Post by neptronix » Nov 14 2019 7:48pm

Yeah people have claimed that 5 phases is better for eons but i haven't ever seen a dyno graph that proves it. I've seen some that prove otherwise though.
I remember falco did this with a hub motor. And the thing performed less well than a crappy crystalyte of the same size.
Toorbough ULL-Zeveigh wrote:
Nov 14 2019 7:13pm
The real downside is that you have to design a custom controller around 4,5,6,7 phases..
they said the same thing when 16 bit processors hit the market, have to be 8 bit backward compatible, installed s/w base, blah blah...[/quote]

It's a pain in the ass to adapt to something that isn't a multiple of 2. Computers had to go 16 bit, then 32, then 64 for a lot of good reasons.

In this case, installed software base is the fact that 99.9% of BLDC motor controllers operate on 3 phase. It's breaking compatibility for no good reason. 6 phase hasn't turned out to be better than 3 in any way. But it's easier to adapt to existing controller designs..
"Love and compassion are necessities, not luxuries. Without them, humanity cannot survive." - Dalai Lama

My first major build: 1.6kW 8T MAC motor on a Trek 4500.
The new all-arounder: Leafmotor 1500w @ 4kW on a Turner O2 full suspension.
The monster scooter: 20" eZee on a Cannondale Semi Recumbent.
Whipper-snapper: ? on a lightweight BikeE Semi Recumbent

dustNbone   10 kW

10 kW
Posts: 946
Joined: Mar 13 2016 9:59pm
Location: Vancouver, BC

Re: More than3 phases

Post by dustNbone » Nov 14 2019 8:06pm

And on the analogy of going with "wider" CPU architectures, mass adoption of 64 bit software didn't happen until there was a good reason for it, namely the desire to access more than 4GB or memory. The performance increase either didn't exist or was so minimal that it wasn't worth the bother.

It wasn't until the old tech hit a hard limit that prevented it moving forward that we decided it was worth the pain of starting to develop 64 bit software. In many cases that meant giving up compatibility with a huge library of excellent 32 bit code that over it's long lifespan had been optimized to a degree that would take many years and millions of coding hours to replicate.

It's actually a pretty good analogy now that I think about it.

User avatar
Toorbough ULL-Zeveigh   100 MW

100 MW
Posts: 2528
Joined: Feb 09 2007 3:02am
Location: Marlboro

Re: More than3 phases

Post by Toorbough ULL-Zeveigh » Nov 15 2019 7:08pm

neptronix wrote:
Nov 14 2019 7:48pm
Yeah people have claimed that 5 phases is better for eons but i haven't ever seen a dyno graph that proves it. I've seen some that prove otherwise though.
I remember falco did this with a hub motor. And the thing performed less well than a crappy crystalyte of the same size.
where is this imaginary graaf that only you can see Oh Grate Gazoo??
i've searched b4 & again just now did an exhaustive sphere-search for 'dyno, falco & 5-phase' with no pix.
also come clean with full disclosure that you've never ridden much less owned a multiphase.
dustNbone wrote:
Nov 14 2019 8:06pm
It's actually a pretty good analogy now that I think about it.
hey, i appreciate the validation; truly.

lot of my time on daSphere feels like i'm just pissing into the wind.
getting nothing in return except a lot of blow-back.
Kick down the barricades Listen what the kids say.
From time to time people change their minds But the Frock is here to stay.
I've seen it all from the bottom to the top Everywhere I go the kids wanna Frock.
Around the world or around the block Everywhere I go the kids wanna Frock.

User avatar
neptronix   100 GW

100 GW
Posts: 14130
Joined: Jun 15 2010 5:56pm
Location: California refugee living in Utah, USA
Contact:

Re: More than3 phases

Post by neptronix » Nov 16 2019 4:39pm

Toorbough ULL-Zeveigh wrote:
Nov 15 2019 7:08pm
where is this imaginary graaf that only you can see Oh Grate Gazoo??
i've searched b4 & again just now did an exhaustive sphere-search for 'dyno, falco & 5-phase' with no pix.
also come clean with full disclosure that you've never ridden much less owned a multiphase.
Falco wouldn't provide me with graphs, just a few peak figures. They were less impressive than a Crystalyte HSxx, which you can easily find a dyno sheet for. Yet they refused to back off the idea that they had some kind of superior motor.

This was in the era of 0.5mm laminations though, where 84% peak efficiency was super impressive.

I've read hundreds of motor graphs by now, looking for the holy grail hub and/or single stage chain drive motor. Show me what you've got for 5 phase motors and let's have data prove the point, instead of opinions.
Toorbough ULL-Zeveigh wrote:
Nov 15 2019 7:08pm
lot of my time on daSphere feels like i'm just pissing into the wind.
getting nothing in return except a lot of blow-back.
You're not alone, i get it all the time
"Love and compassion are necessities, not luxuries. Without them, humanity cannot survive." - Dalai Lama

My first major build: 1.6kW 8T MAC motor on a Trek 4500.
The new all-arounder: Leafmotor 1500w @ 4kW on a Turner O2 full suspension.
The monster scooter: 20" eZee on a Cannondale Semi Recumbent.
Whipper-snapper: ? on a lightweight BikeE Semi Recumbent

User avatar
spinningmagnets   100 GW

100 GW
Posts: 11330
Joined: Dec 21 2007 10:27pm
Location: Ft Riley, NE Kansas

Re: More than3 phases

Post by spinningmagnets » Nov 16 2019 7:02pm

Fisher & Paykel makes 7-phase motors. Falco makes 5-phase motors.

Somewhere around here there is a thread that has large six-phase motors, which allows the option of running it on two common 3-phase motors.

kiwifiat   100 W

100 W
Posts: 159
Joined: Mar 16 2013 5:49pm
Location: Vancouver, Canada

Re: More than3 phases

Post by kiwifiat » Nov 17 2019 3:23am

Phase_7_motor_specs.PNG
Phase_7_motor_specs.PNG (12.72 KiB) Viewed 1373 times
spinningmagnets wrote:
Nov 16 2019 7:02pm
Fisher & Paykel makes 7-phase motors. Falco makes 5-phase motors.

Somewhere around here there is a thread that has large six-phase motors, which allows the option of running it on two common 3-phase motors.
F&P phase 7 is a marketing term, motor specs from the phase 7 manual:

https://www.manualslib.com/manual/78784 ... e=6#manual
Steph wrote:
Nov 14 2019 7:57am

Obviously, multiplying the phase would result in more commutation complexity. This complexity itself is not really my focus here except if in term of practical design it induce negative effects on overall efficiency.

Cheers,
Stéphane
I agree, additional complexity equals added cost and losses. Efficiency gains are unlikely to be found by adding more phases and associated silicon.
“You never know if quotes on the internet are genuine.” - Abraham Lincoln

Steph   10 µW

10 µW
Posts: 5
Joined: Nov 13 2018 5:50pm

Re: More than3 phases

Post by Steph » Dec 26 2019 5:58am

Hello Everyone,

Thank you for your replies. I opened this topic based on a lot of gut feeling, and reading through the contributions, I realised that more than a few proved wrong. I initially though to take part to the discussion, but as it went, I got worried that it would just amount to shame myself :p
My actual starting point was to visualise the phases graph of 3 phases and considering that with more phases, I would get more 'peaks' in the same period. Then, I conceptually mapped each peak with a pole position and related that for a given period, the pole would need to hop between more poles; the relation is like RPM = 1 / (pole_count * phase_count * hop_duration) so has the number of phases increases, the RPM decreases. (And as consequence, for same power input, I would get more torque, which was my goal).
So far, there may be approximations, but I'm still feeling that my way of modelling is not incorrect. However, what I overlooked is that RPM is rather influenced by other dimensions such as the spacing between poles and field power, windings... than by the period (I didn't consider the feedback loop). So, the (phase_count * hop_duration) quantity would probably remain somewhat constant and it would rather be the period that would increase. In that case, one benefit of multiple phases would probably be to reduce eRPM (and therefore switching losses, maybe eddy losses - gut feeling again).

One trigger for me to start this discussion was finding rather cheap used motors on AliExpress (salvaged from crashed agricultural drones, with damaged windings) that could be used as an experimentation platform. (My trade is in embedded software so I was not much worried about designing the controlling software and even the electronics, but at-home mechanical engineering needs better tools than I have and using a prototyping service too expensive to motivate from a gut feeling.) I some here are interested in experimenting, I think they can provide an interesting platform.

From my software background, I'd like to tune in on the proposed analogy (I don't agree with all I read).
Architecture generation jumps have been, in my opinion, firstly made possible by technology advance (we started with 4-bits processors whose mask was fully designed by hand, probably a few hundreds to a couple of thousands of transistors, to reach 64 -and probably even up to 512, if we consider specialised architectures- processors with billions of transistors... designed by software running on increasingly powerful computers), secondly by increasing capacity needs (8 to 16, 16 to 32 bits, to manage bigger data, more directly-addressable memory), and finally by marketing needs (in my opinion, 64bits on an iPhone is not justified, neither it is on most personal computers running office software around, but 3GHz proved kind of a limit for frequency, and we still "had to" get bigger in some way).
Here, we are still in the first segment, where technology drives the evolution (BLDC replacing Brushed DC is a rather recent thing, still seeing improvement, from trapezoidal to sin wave, FOC... maybe more yet to come to better manage the switching losses). The question is whether we will get in the need-driven segment, in the sense that will evolution of electric motorisation bring applications/efficiency approaches that will require more than 3 phases.

I think the complexity of a more than 3 phases controller is irrelevant. We know how to make one, we just lack cheap on the shelves component to build one... because there is no market for such components now, classic chicken-egg story. But this lack can be addressed in various ways: FPGA/Asics, software driven PWM, coupling of MCUs (I believe this is the typical approach to make 6 phases controller. In my view, a cheap McGiver'ing of controllers togethers instead of designing a real one).

I'm still interested in this topic, but my confidence into my gut feeling has been somewhat humbled through this discussion. As a result, I don't think I will pursue the experimentations I initially intended. With a family, time is a too expensive commodity to spend on random experimentations.

User avatar
macribs   1.21 GW

1.21 GW
Posts: 3565
Joined: Jul 22 2014 5:59pm

Re: More than3 phases

Post by macribs » Dec 26 2019 12:58pm

neptronix wrote:
Nov 14 2019 10:31am
Dyno graphs show that you don't get a benefit from more phases. The real downside is that you have to design a custom controller around 4,5,6,7 phases..

All the highest efficiency, highest power density, and most expensive motor designs use 3 phases. Take note!

There have been a few threads about 6 phase motors here in the past. Where one of the benefits or drawbacks are that you can use 2 standard controllers. Using a 9 phase motor you should be able to use 3 standard controllers. Hubmonster motor was 6 phase? Or was it in reality 2 x 3 phase motors in same casing? Not sure. But people have made great claims about those motors. And many of those first hubmonster builds continue to evolve and be rebuild but the motor stays the same. Got to mean something, right?

There was a youtube video (removed due to closed account it says) a while back that claimed doubling up on phases on an electric motor would up the efficiencies and power output. Supposedly there should be a part 2 of that video that explained it all in details but now that account is closed. I am no engineer so I don't understand the working of it all. But the fact that 3 phase became the standard, must that mean 3 phase is superior or could it be that 3 phase motors are superior for the manufacturers in terms of man hours to assemble electric motors?

Also there was a 5 phase and a 7 phase thread here few years back. I even think there was a 5 phase build thread.

I know that those who have build bikes using the hubmonster 6 phase motor seems very pleased with the end result. I also seems to remember claims of those motors having brutal start up torque and hill climbing abilities. Maybe use the search function of the forum to search up those threads and see what people that have done their own testing of more the 3 phase motors says about it.

I think all threads I've read here on the forum regarding more then 3 phases motors have been kind of like this, some people embrace them motors and other loath them. Lots of opinions and people sharing their own results but I can't remember seeing a dyno or some hard data. It would be nice to see a dyno sheets of same size motors, 1 with 3 phases, one with 6 phases etc.

User avatar
amberwolf   100 GW

100 GW
Posts: 28299
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: More than3 phases

Post by amberwolf » Dec 26 2019 2:44pm

macribs wrote:
Dec 26 2019 12:58pm
? But the fact that 3 phase became the standard, must that mean 3 phase is superior or could it be that 3 phase motors are superior for the manufacturers in terms of man hours to assemble electric motors?
it's more like 3 phase is cheapest--it's the simplest version of the motor and controller that can be built, especially given that so many drive systems for that number already exist, so it's the one that gets used most of the time. some applications need or want something that doesn't give, so they use other versions.

just like with brushed motors, two brush (two pole) versions are simplest and cheapest and work for most applications. some need higher torque / lower rpm in the same package, so they use four brushes (four pole).

Punx0r   100 GW

100 GW
Posts: 5350
Joined: May 03 2012 8:16am
Location: England

Re: More than3 phases

Post by Punx0r » Dec 27 2019 5:34am

3 phase is a great compromise for transmitting as much power as you can over as few conductors as possible and delivers roughly constant power to the load over the AC cycle. Power transfer can be improved with more phases but the benefits are relatively small and diminishing.

Higher phase counts (like the 6 phase hubmonster) do have an important advantage though: they let you spread the power they require across two motor controllers, which are easier to acquire than one large one.

Similarly, many envision high current, low voltage systems for powerful EVs with slow turning, direct drive motors. These would benefit from many-phase systems to phase current to manageable levels.

You don't need more phases to make a motor turn slower/produce more torque though: just increase the pole count. Or are you specficially looking to modify an existing motor with a fixed pole count?

Steph   10 µW

10 µW
Posts: 5
Joined: Nov 13 2018 5:50pm

Re: More than3 phases

Post by Steph » Dec 27 2019 7:06am

Punx0r wrote:
Dec 27 2019 5:34am
Or are you specficially looking to modify an existing motor with a fixed pole count?
That was indeed the idea, salvaging the frame of motors with already an high pole count: https://h5.aliexpress.com/item/32288259948.html
Fully designing a motor is not in my reach.

User avatar
ElectricGod   100 MW

100 MW
Posts: 2931
Joined: Nov 01 2015 5:24pm
Location: 3 OH 3

Re: More than3 phases

Post by ElectricGod » Dec 28 2019 12:56am

I want to experiment with rewinding a typical 3 phase BLDC outrunner for 6 phases.
I have an Alien Power 12090 that has a shorted phase. With 24 stator teeth, splitting it into 2 motors on the same stator would be pretty easy to do...12 teeth per motor or 4 teeth per phase.

Image

The Kv is going to double. I don't know the turn count so lets pretend its 18 turns for 50 kv from the factory. At 18 turns with half as many stator teeth, the new KV ought to be 2X higher or 100 kv. At least I think that's right, but I could be wrong. :)

I have a couple of ideas to deal with this.
1. The motor was never wound for decent copper fill from the factory so there is room for more turns per tooth.
2. It is currently wired for delta. Wire it for WYE and that will reduce the Kv by 1.7X. This is almost enough by itself to stay at the same Kv.

I have just rewound an outrunner that was originally wired delta. I took advantage of the lowered Kv for WYE.
For delta to get 80 kv, I need 10 turns per tooth.
For WYE to get 80Kv, I need 6 turns per tooth.
Since I now had 4 less turns per tooth, I was able to add about 33% more copper cross section to each turn.
With the maximum copper fill I had and wired delta, I was measuring about 13 mOHms for each separate phase.
I used this measured resistance to estimate what I should get with adding 33% more copper cross section and the lesser length of 6 turns vs 10 turns.
My estimate of 2 phases in series like is in WYE, calculated out to almost identical resistance to the single phase in delta.
I like to try stuff and just becasue my estimate didn't look compelling, I was not deterred.
I went ahead with the rewind anyway just to see if it worked out any better or worse. As it turns out, it was a lot better.
I always measure each phase resistance 3 times so a poor connections or other oddities are unlikely.
In the new wind, with 33% more copper cross section and the shorter length per phase, I am measuring 4.17 mOhms per individual phase.
Two in series is about 8.3 mOhms. This worked out pretty good compared to the previous delta wind.

Between rewinding my AP 12090 for better copper fill and then wiring it WYE on each of the two separate sets of winds, I think I will be able to get 80 Kv or so and still have lots more copper cross section than the factory wind.

What I don't know...and this is the whole point...is will the motor be better at 6 phases than it was at 3 phases?
Will it be worth the trouble or not?
Are there other implications that will make this worse than 3 phases?
XB-502: viewtopic.php?f=10&t=83302&p=1222730#p1222730

Currie scooter: viewtopic.php?f=35&t=83830&p=1227407#p1227407

Benjamin Franklin - "Any fool can criticize, condemn and complain and most fools do."

Grantmac   10 kW

10 kW
Posts: 680
Joined: Oct 22 2018 12:43pm
Location: Victoria, BC

Re: More than3 phases

Post by Grantmac » Dec 28 2019 1:40am

Would a 6 phase have less ERPM and hence be easier to drive?
Also could you opt to "sync" the controllers to in essence change the Kv as needed? I'm not confident on motor design so that's a pretty wild set of guesses and probably not beneficial for outrunners.

User avatar
ElectricGod   100 MW

100 MW
Posts: 2931
Joined: Nov 01 2015 5:24pm
Location: 3 OH 3

Re: More than3 phases

Post by ElectricGod » Dec 28 2019 4:03am

Grantmac wrote:
Dec 28 2019 1:40am
Would a 6 phase have less ERPM and hence be easier to drive?
Also could you opt to "sync" the controllers to in essence change the Kv as needed? I'm not confident on motor design so that's a pretty wild set of guesses and probably not beneficial for outrunners.
eRPM = Battery voltage X magnet poles X Kv

If I assume the motor has 6 phases and both sets of phases are wound identically, then the eRPM for each set of phases is also the same. The same stator wound for 60 Kv and 3 phases will have the same eRPM as it would wound for 6 phases at 60 Kv. The magnet count does not change. Same Kv, same stator, same battery voltage...same eRPM. That may be overly simplistic...lol.

I'm not sure what you mean by "sync the controllers to change the Kv". Please explain further. This may be something I have never heard of before.

In any 6 phase or 9 phase or 12 phase configuration, each BLDC motor (set of 3 phases) has its own set of halls. Since all the motors are locked onto the same stator and armature, they are for all intents and purposes always "in sync". They can't operate any other way. You don't need to worry about the 2 or 3 or 4 controllers being able to run their individual motors correctly either. They get their individual position data from the halls associated with the phases that specific controller runs. The controllers have no choice but to run "in sync". About the only thing that really matters is that one controller isn't driving its motor harder than the other controllers are driving their motors. This is easily adjusted with throttle or maybe adjusting individual phase currents a little so the controllers all drive close to the same.
XB-502: viewtopic.php?f=10&t=83302&p=1222730#p1222730

Currie scooter: viewtopic.php?f=35&t=83830&p=1227407#p1227407

Benjamin Franklin - "Any fool can criticize, condemn and complain and most fools do."

User avatar
macribs   1.21 GW

1.21 GW
Posts: 3565
Joined: Jul 22 2014 5:59pm

Re: More than3 phases

Post by macribs » Dec 28 2019 10:54am

What I don't know...and this is the whole point...is will the motor be better at 6 phases than it was at 3 phases?
Will it be worth the trouble or not?
Are there other implications that will make this worse than 3 phases?
You would run that motor using 2 motor controllers, right? I mean just because more is...well more. :D And you will likely get an insane take off torque and you can get more power from your new motor using 2 controllers without pushing the controller to the absolute limits in regards to max amps.

I think I saw one thread here years back where someone actually did the math and it did not result in twice the power but was like 1.8 or 1.9 x the 3 phase power.

That could lead you to run cheaper controllers. Or just to explore how much power you can squeeze out and use the same controller you did when it was a 3 phase motor. Or even go balls deep and get two higher power ones for maximum fun.

I haven't done a rewound so I am not speaking of experience, but I do know you like to thinker and you have build some cool stuff in the past and even have done a rewound of a motor so I think for you this will be smooth sailing and well worth it. And for the rest of us here, it will for sure be legendary to see your result.

Worse the a 3 phase? More power more torque how can that be worse :twisted:

Grantmac   10 kW

10 kW
Posts: 680
Joined: Oct 22 2018 12:43pm
Location: Victoria, BC

Re: More than3 phases

Post by Grantmac » Dec 28 2019 1:26pm

I'm imagining that if you ran just one set of 3 phases there would be a gap between the poles. Wouldn't that lead to a faster "jump" towards the next magnet? Perhaps I'm seeing it wrong.

The phases should have lower resistance though because of shorter windings so KV will be affected somewhat.

Being able to run two small, cheap controllers is appealing. Or even one of those Eskate double VESCs.

My syncing comment was about firing adjacent poles in unison. Basically making for a lower pole count on the fly.

User avatar
ElectricGod   100 MW

100 MW
Posts: 2931
Joined: Nov 01 2015 5:24pm
Location: 3 OH 3

Re: More than3 phases

Post by ElectricGod » Dec 28 2019 5:18pm

Grantmac wrote:
Dec 28 2019 1:26pm
I'm imagining that if you ran just one set of 3 phases there would be a gap between the poles. Wouldn't that lead to a faster "jump" towards the next magnet?

My syncing comment was about firing adjacent poles in unison. Basically making for a lower pole count on the fly.
From the perspective of a single motor...wether that is on a 3 phase motor or a stator wound for 6 phases or 12 phases or whatever...the single motor/controller only sees what it sees. That specific motor and controller combination don't need to access the other motors, halls or controllers to run. The halls for any specific motor/controller show position for that specific motor. Commutation is identical in any specific motor regardless of how many other motors may also be on the same stator. From playing around with a hubmonster, it runs on a single motor or both of them at the same time. There is no lack of smoothness of rotation running just a single motor vs both of them once the armature gets moving. You have more magnetic cogging to overcome. It takes a bit more throttle to get a single motor turning and to overcome the cogging for the other motor than it takes to run both motors at the same time. Once that is overcome and the armature is spinning, the entire mechanism spins smoothly running a single motor or both of them.

Poles? I think you mean stator teeth? You can't really do anything about the magnets which are what are typically referred to as "poles".

So assuming stator teeth:
I know of several motors wound across 2 or more teeth. This is an AstroFlight 3220. You can see that winds cross 3 stator teeth and overlap with the other 2 phases. I've wondered what benefit or effect this might have, but I'm no motor expert so I don't know much more than observation. I'd love to know why AstroFlight winds like this on a stator with so many teeth. Anybody know? It clearly leaves large empty spaces in the stator teeth that could have copper in them.

Image

If I understand you correctly...
1. The magnets are fixed quantities. Changing the magnets on the fly is not really possible. in my 12090, the magnets are 12mm wide and have a gap between each magnet. I could replace them with much wider magnets so that the distance across 2 stator faces is spanned by a single magnet. I could then wind 2 adjacent stator teeth as if they were a single tooth. This would lower the motor pole count by half.

2. I could also wind the stator for 6 phases and use those wider magnets as well. I'm completely uncertain about how well or badly this would work. The same magnet would get attracted or repelled by 2 motors at the same time. One would need to be careful to get it right so that adjacent stator teeth are not working against each other. I have no idea, or even a guess, if this is better than or worse than how motors typically work with a single stator tooth and a single magnet.

3. How you wind the motor won't change the pole count in any way. Pole count is the number of magnets/2. That's a fixed thing. Whatever magnet count is in the motor is what it has.
Last edited by ElectricGod on Dec 28 2019 6:54pm, edited 1 time in total.
XB-502: viewtopic.php?f=10&t=83302&p=1222730#p1222730

Currie scooter: viewtopic.php?f=35&t=83830&p=1227407#p1227407

Benjamin Franklin - "Any fool can criticize, condemn and complain and most fools do."

User avatar
ElectricGod   100 MW

100 MW
Posts: 2931
Joined: Nov 01 2015 5:24pm
Location: 3 OH 3

Re: More than3 phases

Post by ElectricGod » Dec 28 2019 6:40pm

macribs wrote:
Dec 28 2019 10:54am
What I don't know...and this is the whole point...is will the motor be better at 6 phases than it was at 3 phases?
Will it be worth the trouble or not?
Are there other implications that will make this worse than 3 phases?
You would run that motor using 2 motor controllers, right? I mean just because more is...well more. :D And you will likely get an insane take off torque and you can get more power from your new motor using 2 controllers without pushing the controller to the absolute limits in regards to max amps.

I think I saw one thread here years back where someone actually did the math and it did not result in twice the power but was like 1.8 or 1.9 x the 3 phase power.

That could lead you to run cheaper controllers. Or just to explore how much power you can squeeze out and use the same controller you did when it was a 3 phase motor. Or even go balls deep and get two higher power ones for maximum fun.

I haven't done a rewound so I am not speaking of experience, but I do know you like to thinker and you have build some cool stuff in the past and even have done a rewound of a motor so I think for you this will be smooth sailing and well worth it. And for the rest of us here, it will for sure be legendary to see your result.

Worse the a 3 phase? More power more torque how can that be worse :twisted:
2 smaller controllers is definitely cheaper than 1 big controller. 2 10kw controllers is about 75% the cost of a single 20kw controller. From that perspective, 6 phases wins.

Please see if you can find that thread about power gains from 6 phases vs 3 phases. If there is something to this from the perspective of the motor, I'd like to know about it.

It seems logical to me that the same total copper cross section ought to work equally well in 3 phases or 6 phases. That's assuming a single controller capable of running the motor harder than its capabilities at 3 phases or 2 of those same controllers and 6 phases...aka keeping as much equal as possible. The motor ought to deliver the same torque across 3 phases as it does across 6 phases with the same amount of total phase amps. In my mind this is a lot like a 3 piston gas engine vs a 6 piston with the same total displacement and compression ratio. More pistons ought to make the power curve smoother, but that's about it. I could be wrong...especially so since I really don't know.

I think this is the important details...lol...but I could be wrong...what else would make 6 phases better than 3 phases in the motor?

1. Iron losses ought to be the same.
2. Magnetic current losses ought to be the same.
3. Assuming winding for the same Kv 3 phase or 6 phase and both using the same wiring scheme, phase wire lengths and resistance would be pretty close to the same.
4. bEMF losses ought to be the same too.
5. Magnet cogging would be the same.

Having said all of that...
I still want to try a 6 phase wind on that 12090!
Worst case it needs to be rewound no matter what so why not 6 phases?
I don't really care that I don't know of a reason why beyond the controller cost for doing 6 phases.
If you count how many times I've rewound motors for any reason, I'm at something like 20 motor winds now.
If you go on what actually saw the light of day and got ran at least on the bench, then 5 motors.
I've never done 6 phases in all of those motor winds.
I'd do it just for the curiosity factor...which is good enough reason for me.
XB-502: viewtopic.php?f=10&t=83302&p=1222730#p1222730

Currie scooter: viewtopic.php?f=35&t=83830&p=1227407#p1227407

Benjamin Franklin - "Any fool can criticize, condemn and complain and most fools do."

John in CR   100 GW

100 GW
Posts: 13964
Joined: May 20 2008 12:58am
Location: Paradise

Re: More than3 phases

Post by John in CR » Dec 28 2019 7:45pm

Steph wrote:
Nov 14 2019 7:57am
Hello,

To reduce the need for reduction, I tend to look for low kv motors (as a way to get more torque, less speed)
More torque per amp, yes. More torque period, no, because current handling decreases proportionately.

Since a motor with the same copper fill % is capable of exactly the same torque regardless of how it is wound, your approach is exactly the opposite of the correct answer. Greater gear reduction will in fact increase torque potential at the wheel.

John in CR   100 GW

100 GW
Posts: 13964
Joined: May 20 2008 12:58am
Location: Paradise

Re: More than3 phases

Post by John in CR » Dec 28 2019 8:00pm

Other than controller cost savings and what seems to be smoother operation with 6 phase motor vs 3 phase, the only benefit I know of is that say you take a stator and wind it 3 phase for a given Kv with that pole count. If you wind it as 6 phase instead for the same Kv, the inductance of the phases will be greater making the motor an easier load for the controller to drive.

Torque has a direct inverse relationship with Kv, and the number of phases has no relationship with Kv, so it doesn't impact torque. Over a decade ago a forum member built sizable outrunner from scratch. If I remember correctly it had 24 coils. He terminated the motor as a 3 phase motor and also ran it as a 24 phase motor. While the controllers were quite different, a normal 3 phase ebike controller vs 8 tiny RC controllers, he did report much better results with the 8 controllers in the 24 phase alignment. Unfortunately he tried to sell an online book about how to make your own motor, and we never heard from him again, so we couldn't discuss his results in more detail.
Last edited by John in CR on Dec 29 2019 3:19am, edited 1 time in total.

Grantmac   10 kW

10 kW
Posts: 680
Joined: Oct 22 2018 12:43pm
Location: Victoria, BC

Re: More than3 phases

Post by Grantmac » Dec 28 2019 8:26pm

I had my pole vs stator terminology mixed up but it seems you understood what I was grasping at.

That Astro is a very interesting wind. Overlapping fields but looks like a lot of unused space in the stator.

Post Reply