6 phase hub motor?

[PhoenixOSU]

I did some searching and found that RMartin makes a moped with a "6 phase" motor, normally BLDC motors are 3-phase, 6-phase would greatly increase efficiency, though most controllers only do 3, so I'd make my own! Anyway, if anyone knows where buy just the motor or who makes it, or if it even is really a 6-phase motor or just a "6 commutation phase" motor, which is really just a silly name for 3-phase motor.

For that matter, if anyone has any links to BLDC motors with more than 3 phases, I'd be interested, preferably hub motors.

 

[Miles]

There was some discussion on this, here: http://endless-sphere.com/forums/viewtopic.php?f=28&t=9061&p=142145#p142145

 

[John in CR]

Sounds good. I just wouldn't want to have to figure out the correct wiring combo. It can be frustrating enough with 3, but if it's 6 phase and 6 halls....ouch!

John

 

[Toorbough ULL-Zeveigh]

ahem, yes.
i've been riding an optically encoded 7-phase hub motor for the past 4 years.
restricted as it is to 19.5 mph, from what i can tell only two winding combinations are implemented.
there's an electronic 'gear down' anytime the speed drops to below 10 mph while the power draw exceeds the controller limit for (guessing) 5 sec when climbing a steep grade.
after switching to the higher wind combo the current drain is about half, only switching back once the load drops off for ~5 sec.
overall it doesn't seem to be any more efficient than it's 3-phase successor.
if u could make a replacement controller for the Wavecrest motor to fully utilize all of the possible winding configurations u would be hailed as a god by a legion of orphaned owners.

 

[PhoenixOSU]

do I hear a challenge?

Send me a list of inputs and outputs and some basic functionality description info. At 7 phases it would be expensive due to the number of FETs, and a lot would be in software too.

 

[Toorbough ULL-Zeveigh]

yes, it's almost entirely software.
it's a closed proprietary box, the Mac equivalent to the Xlyte PC.
other than what's in the filed patents which only describes the system in a general way, getting any kind of technical info is tough to come by.
u wouldn't happen to know where to get a copy of JTAG Explorer, that would help?

 

[PhoenixOSU]

I think there are some boards from microchip, and I know there are some from xilinx that should do the trick as far as USB - JTAG or Serial to - JTAG

 

[TimWms]

I have a 2008 EVD scooter from RMartin, which is made by Huari in China. The motor is described as a 6 phase 3000W brushless DC hub motor. I added a 6th battery to give me 72V, and modified the controller to accept the higher voltage, so I am pretty familiar with this controller. The controller tends to overheat, but when the motor cut out, I could not tell for sure why, because the diagnostic LED was not properly connected. While tracking down the source of this problem, I moved some wire bundles around, which I believe allowed a couple of poorly insulated phase wires to short out to each other, which blew up the controller. For all the gory details, see:
http://visforvoltage.org/forum/8880-60v-72v-upgrade-vrla-evd

I have been searching for a controller, and have not found any that are designed for a 6-phase motor. I thought the Kelly controllers might work, since they say they are configurable for any number of poles, and compatible with 60-degree and 120-degree hall sensors, but there are only 3 hall sensor inputs and 3 phase winding leads coming out of their controllers, so they definitely are 3-phase controllers. Motor control is new for me, so I've been trying to learn the terminology as I go.

The controller that came with the bike contained two identical PCBs, each driving three phase wires, and being fed three hall sensor signals. At first, I thought maybe these two boards were hooked up in parallel to drive double the current to each of three phase wires, but there are definitely 6 separate phases, and 6 separate hall sensor leads from the motor. I have checked the resistance on these leads when not connected to the controller, and there are two groups of 3 phase wires each. Measuring the resistance of any of the leads in a group to any of the other two leads in the same group shows 0.6 Ohms. There is no connectivity between the two groups. There is also no connectivity between the hall sensor leads. So basically, it looks like two separate motors working together. Right now, I think I will need to have two KEB72331 controllers.

One note about the Kelly controllers - On their website, they show the KBS72121 controller as capable of handling 50 A continuous, 120 A for 1 minute. I ordered one, but when i opened the package, it was obvious it is designed for a much smaller drive. It is 1/4 the size of the KEB series controllers, and the phase leads are made up of 2 wires in parallel, each with a diameter including insulation of 3 mm. If you put 50 Amps through those wires they would melt like a fuse. I am convinced the spec should read this controller as rated for 120 Watts, not 120 Amps, but so far I have not convinced Kelly support.

 

[only1jake]

I'm sure they will be able to build you a 6 phase controller if you emailed them. With your own specs too! Kelly controllers that is.

 

[RoughRider]

so, what is the real advantage of a 6-phase system?

 

[liveforphysics]

so, what is the real advantage of a 6-phase system?

You get the advantages of delta and wye at the same time (high torque per phase amp, and high speed range).
You also get more controller channels to distribute current loads, which is good for performance and stability, bad for price and package footprint.

Also, as far as im aware, every 3phase motor can be ran from a 6 phase controller, just undo the legs at the delta or wye connection it had before and run all 6 out.

 

[RoughRider]

do i understand you right?

you would make 2x windings with 3xphase each...

1x winding with Y
1x winding with delta

is that right?

 

[liveforphysics]

do i understand you right?

you would make 2x windings with 3xphase each...

1x winding with Y
1x winding with delta

is that right?

No.

You don't terminate any coils. You just group the coils that happen to be in phased the same.

So, if you have 3 teeth on your stator, you just collect the ends of each coil for your 6 wires. If you have 6 teeth, you parallel the groups of 2 coils that are in phase. If you have 36 teeth for example, then you would have 6 groups with 6 coils paralleled together, 72 teeth would mean 6 groups of 12 like-phased paralleled coils in each group.

At least this was how the diagram was for the 6-phase scooter motor I was checking out worked.

 

[John in CR]

For the 6 phase controller replacement couldn't you just run a pair of typical 3 phase controllers? Use one with halls to one set of phase wire and the hall sensors, and a sensorless controller on the other set of phase wires. The hall sensor placement might be an issue though. A pair of sensorless controllers should do the trick though, plus with 6 phases and the higher Kv of scooter hubs due to the smaller wheel than a bike, should make sensorless operation smoother than on Ebikes. Going with 2 more common controllers enables you to go to higher voltages economically, and the programmable ones virtually guarantees greater performance.

 

[crossbreak]

Hi guys,

I read a lot of threats about this topic but i still don't get it. Mechanical Delta-Wye switching is simple but also dangerous, because of high back-EMF in WYE-Mode at high speed, I agree. Thanks for all the info posted.

But why would a 6-phase motor be more efficient?

6-Phase sounds much better. But connecting two FETs to each winding lead would just be the same as using a 12-Fet controller, only the Delta-connections between the phase leads are missing:

This is not true 6-phase, is it?

For true parallel/series connection the WYE leads must be conducted together with a mechanical switch a thyristor for each lead. Like this Delta/Wye switching, which looks much too complicated, note that the 6 Fets on the right won't be used in WYE-Mode:

I see no other possibility. I'd like to see a true 6-phase circuit if someone can post plz! I found nothing on the web.

 

[dcb]

yes, old thread, I know I know.

I *think* 6 phase is the same as 3 phase if you use 6 poles instead of 12. If you were to use all 12 poles in the above picture the resolution change, the potential improvement in smoothness, and power transfer would be more obvious, though I don't really know what the heck I am saying. The difference between poles is 60 electrical degrees instead of 120, so a finer level of control. If there is physical room for more poles and winding then that would have a similar effect, though with more phases to the controller you can transmit more power for a given supply voltage, maybe, more area under the curves.

 

[macribs]

Old thread I know, but as this thread shows up when searching for 6 phase hub I might as well weak it up.
AFAIK one of the best things about 6 phase motors is that you can share the controller load between two 3 phase controllers. That way each controller will be left with half the load, yet together they will give you all the power. So you can run cheaper controllers yet harvest big power.

To read more about 6 phase look into John in CR's hubmonsters. IIRC John harvest peak performance of 27 kw from his hub monster 6 phase motor. 8)