DIY motor is going 6phase with dual R/C esc's the way to go?

[John in CR]

OK, so like others I've got the DIY motor bug thanks to Thud. LFP and I talked briefly about the advantages of going 6 phase on a motor. It would also be nice to eliminate sensors from the items can go wrong, I also want to avoid the complexity of multi-stage gearing, but I'd want better than typical take off torque. I like the size and price of the R/C controllers, even if it does mean lower voltage. Also, if I'm thinking correctly, I could limp home on a single esc if one blew.

Does this potentially solve some of the startup and sync issues of using R/C controllers with larger diameter lower Kv e-bike motors?

Bad idea, or one maybe with some merit?

John

 

[liveforphysics]

I think by adding RC controllers, the best you could do it separated sets of 3-phase windings rather than 6-phase, because I think they require having delta or wye terminations for the BEMF voltage to be phased correctly and gain the ability to sync.
Grouping RC controllers seems like a really good idea for keeping costs low and simplicity in higher power designs.

I think a true 6-phase setup has 3 series or parallel, or series parallel coil sets, with a dedicated set of leads for each of the three coil sets. This gives you the torque potential of a wye terminated 3-phase motor with the speed potential of a delta 3-phase motor all in a single motor, but at the expense 6 phase leads and a controller that runs a minimum of 12 fet banks. At least this is my meager understanding of 6-phase.

 

[John in CR]

Yes, that's exactly what I was thinking, dual 3 phase windings.

 

[Thud]

John,
It sounds like you are describing parralell/series 3 phase. I, of course have no idea what the ramifications are regarding controller pricing- I really like the price of the e-crazyman controllers (sensored) & methods has shown the way to mod them way past what an rc unit can do.(I just hope I have the chops to make the mods without a destroy in the ledger) I really think that you could build a monster motor with multi stators & multi rc controllers, Like Axel did. My sugestion would be to seperate the 3 phases about the stator.& configure the others paralell. Asuming that doesnt completely screw the back emf readings for the other controllers, that should increase the starting tourque that Axels designed lacked (from his descriptions) I am pretty sure he has progressed dramaticly for what we've seen on his videos.

 

[MitchJi]

Hi John,

Does this potentially solve some of the startup and sync issues of using R/C controllers with larger diameter lower Kv e-bike motors?

John

I'm not sure if there are any Sensorless ebike controllers are available. But I thought the problem was the other way around. Using ebike controllers with higher RPM/kv RC motors. Because when I asked Justin about using it with RC Motors the high RPM/kv was one show stopper (the other being its not a product).
http://endless-sphere.com/forums/viewtopic.php?f=29&t=7134#p107193

The Controller
7) Sensored and Sensorless: If there is a hall cable connected it will run as a sensored motor controller, if there are no hall signals present it operates sensorlessly. If the hall senor was initially connected, and then fails in the middle of your ride, the controller will detect this and switch over to sensorless mode without loosing a beat. I don't really need to say much more for people to know what this means for reliability!

 

[kenkad]

Hello John,
You are going in the right direction. Here are two JPGs that show what I will be doing. I am using some of Shane Coulton's design for the controller part. The JPGs show a 6 by 3phase drive. Also, Shane has confirmed that if you go to a sin drive instead of a square wave drive, the noise is much much less. Shane has really done some good work and his results are free. Two 3phase drives will give you much better smooth starting torque, but, the cavet is that this is also a function of the magnet orientation on the rotors and the design of the stator. Shane has been trying to point out that this is all closely tied together. You cannot solve one and not solve all the interacting factors together. My motor design is slow in coming. I have other things to work out as well for my delta trike. I am glad to see all of this discussion in this forum at this time. It is long overdue. I am sure your contribution will be appreciated by everyone of us.
kenkad

 

[John in CR]

Hello John,
You are going in the right direction. Here are two JPGs that show what I will be doing. I am using some of Shane Coulton's design for the controller part. The JPGs show a 6 by 3phase drive. Also, Shane has confirmed that if you go to a sin drive instead of a square wave drive, the noise is much much less. Shane has really done some good work and his results are free. Two 3phase drives will give you much better smooth starting torque, but, the cavet is that this is also a function of the magnet orientation on the rotors and the design of the stator. Shane has been trying to point out that this is all closely tied together. You cannot solve one and not solve all the interacting factors together. My motor design is slow in coming. I have other things to work out as well for my delta trike. I am glad to see all of this discussion in this forum at this time. It is long overdue. I am sure your contribution will be appreciated by everyone of us.
kenkad

Thanks for the vote of confidence, but I'm a follower on this stuff, especially the electronics. Thanks for pointing out Shane Colton. He's shared some great stuff on his blog, http://scolton.blogspot.com/search/label/axial motor. I'm sure we'll want to borrow some things from him in his "epic axial motor" project. FWIW, my gut too is telling me his motor will dissipate a lot more heat than his first pass calculation with that kind of air flow.

John

 

[kenkad]

Hello John,
Thought you might want to look at this 3phase drive controller module from Cirrus Logic. This is the SA306. It's available from DigiKey. I am planning to use this for the multiple 3phase drives when I do my motor. It really has well thought out pinout and at 8A continuous (15A max) and 60V, it meets all of my requirements and has builtin drive for the Mosfets. If you incorporate 2 3phase drive coil arrangements, you can get better starting torque and higher efficiency. You can get the details of their demo PCB (schematic and layout) off their website.
kenkad

 

[recumpence]

I am not an EE. But, Astro Flight told me this will not work with standard iron core motors because the pulses of each winding will confuse the back EMF signal to the other controller. I think he called it "Talk-back".

Anyway, something to think about...........

(Again, I am not an expert here. I am just repeating what was to to me).

Matt

 

[enoob]

when my wife talks back i just giver a slap . think that would work here ?. looking over my shoulder, cant have the wife see me type this ill get hurt . shes a black belt in something or other
and I blame thud as well john . i say we vote him guru of diy, that'll teach him.

 

[kenkad]

Hello Matt,
That is a function of how the stator is designed. I would suspect that the Astro Flight motor does not utilize independent (unconnected iron) coil cores (just my guess). This is what has impressed me by the discussions I have had with Shane Colton. The design of the rotor, in particular the magnet orientations, and the design of the stator determine the bEMF wave shape (his blog references a Masters Thesis document from KSU). The math in the Thesis is not what is important, the results/conclusions are important. The integrated design also affects whether the motor cogs at low speeds. I mentioned earlier, that from these discussions, he (Shane) did a sine drive implementation and immediately discovered that the motor noise was significantly reduced. His new design incorprates a microcontroller with three independent PWM outputs specifically to implement the sine drive. I am not a motor design expert either. I am hoping that these forum threads will help others think through their designs and give everyone ideas to test. This is the only way we will learn what does and does not work well. I wish everyone the best in their efforts. I hope we see some really good ideas come to light in the motor challenge thread efforts.
kenkad

 

[Thud]

Enoob wrote:

when my wife talks back i just giver a slap

I got a cuff to the back of the head just for chuckling at that (she has a black belt in frying pan)

KenKad,
I wish I had more than a simpltons perspective regarding the building of the controllers. I can visualize the sine vs squarewave & see the potential. But how to get there?....I am relying on yours & Mr Coltons work. I am still amazed at the comunity here at ES. Its good to see we are awakening all aspects of the power package up in our little group.
If there is any thing mechanicle you need, We can get it done here between Enoob, Matt or myself.

That surface mount silicon vodoo thing-a-ma-wacky, well lets just say I am way behind that technology.
(I am still working my way thrugh the bronze age!)

 

[kenkad]

Hello Thud,
Thank you for the generous offer. I keep in touch with Shane and exchange ideas. I like to think that the sine drive implementation he did recently was partly due to my harping on it to him. My technology approach and his are 'different'. Both the motor design and the control electronics. I prefer to use the 'silicon vodoo thing-a-ma-wachy' rather than a bunch of discrete components, eventhough it costs more for the single part. My design wants to pursue more 3phase groups in a motor and thus my drive current per phase requirement is lower than his. I want modularity and replacebility within the motor design, both mechanical and electrical. My mechanical design is about 98 percent finished. I have gotten a great many good ideas from this forum and others. I really appreciate your suggestion about mechanical fab support. I am in the process of getting my 4-axis mill up and running again. It has been a somewhat expensive and time consuming process. Seems most companies have cut down on staff, costs are up, and, parts are hard to find. It is easy for them to say 'we can install upgrades and get it all checked out', but I cannot afford that. Everything I am doing will be 'open source', but, not just yet. Right now, I will be spending time on the electrical design of the motor controller (schematics and PCB layouts). I am also in the process of building two delta trike frames to use as testbeds for the motor work.

I agree that we should all ask questions, contribute suggestions and exchange ideas. All of us may have different goals and work at different speeds to implement our ideas. Certainly there is nothing wrong with that and no one should be chided for that. I enjoy learning from the experience and ideas of others. I am particularly interested in the motor challenge thread. Again, thank you for the offer.
kenkad

 

[Miles]

Hi Kenkad,

I'm particularly interested in Sine controllers because of the lower noise levels they generate. I'd really like to see one with integrated control of motor direction, so that I could use it with the 2 speed gearbox I've designed...

 

[kenkad]

Hello Miles,
My requirement for motor reversing/smooth soft start torque are because I am using two SRAM geared hubs (a 3i followed by a 9i) in my delta trike drivetrain. The motor sits between (drive wise) the two geared hubs (of course there is a freewheel there also). I plan a belt drive gear reduction for the motor as well. I did not want cogging to have a distructive effect on the geared hub. Shane plans to publish all of his new stuff soon on his blog. I could not say that his motor controller is design for controlled reversing. I do not believe that his previous design did that. He uses Hall sensors and my design uses higher resolution reflective optic sensors for higher rotation position accuracy. I believe this is critical for soft start and reversing. I would be happy to discuss some particular detail if you want to PM me. All my work is still in the design phase. Have you actually built your gearbox?
kenkad

 

[Jonathan in Hiram]

There is a considerable amount of information on generating sine waves for various purposes at this URL..

http://www.tinaja.com/magsn01.asp

"Magic sinewaves" are repeating long sequences of ones and zeros.
They can get created from ordinary but extremely carefully chosen
digitally switched pulses.

Digital sinewaves with precisely controlled amplitudes and amazingly
low distortions. Compared to traditional PWM, magic sinewaves can
offer far higher efficiencies and lower distortions. With circuitry that
is elegantly simple and microcontroller friendly.

For use in such applications as induction motor speed controls, electric
autos, solar panels, power factor correction, inverters, home energy
efficiency improvers, 400 Hz avionics, UPS, PFC, and special utilities.

The newest steplocked versions of magic sinewaves let you force any
desired number of low harmonics precisely to zero in theory and to
astonishingly low levels in practice. While allowing you to do so at the
highest possible efficiency. By using the fewest switching events.

 

[Miles]

I would be happy to discuss some particular detail if you want to PM me.

Thanks. PM sent.

 

[liveforphysics]

Hello Miles,
My requirement for motor reversing/smooth soft start torque are because I am using two SRAM geared hubs (a 3i followed by a 9i) in my delta trike drivetrain. The motor sits between (drive wise) the two geared hubs (of course there is a freewheel there also). I plan a belt drive gear reduction for the motor as well. I did not want cogging to have a distructive effect on the geared hub. Shane plans to publish all of his new stuff soon on his blog. I could not say that his motor controller is design for controlled reversing. I do not believe that his previous design did that. He uses Hall sensors and my design uses higher resolution reflective optic sensors for higher rotation position accuracy. I believe this is critical for soft start and reversing. I would be happy to discuss some particular detail if you want to PM me. All my work is still in the design phase. Have you actually built your gearbox?
kenkad

Kenkad- You sound like exactly the sort of person we need more of on this board.

I wish the best of luck, and if you need parts, FETs, caps, etc, I have ordered far more than I need of most everything, and I will gladly ship you parts to prototype with free of charge.

-Luke

 

[kenkad]

Hello Luke,
Thank you for your offer. I am a very methodical person and as I said earlier, I want the most integrated components that are now available. Thank you also for the show of confidence. As you may have seen in an earlier thread of mine, I asked if there were any individuals in the North Alabama area that might want to work together and I got no responses. Similar response from all the local Universities. That does not deter me from continuing what I am doing. It will just take a little longer for me. I am not in the motor challenge because I work too slow. I will post more stuff when I can.

Thanks again.
kenkad

 

[John in CR]

I wonder how much efficiency gain exists with a sine wave controller, since there's less noise. I realize there's not much power in the actual sound, but my gut tells me that jerky commutation at startup that creates the most noise, must mean a less than optimum alignment of the rotor and stator for a portion of the rotation.

Regarding the crosstalk between coils using a pair of sensorless esc's, I don't know. Of course I have to defer to the experts, but If there's enough crosstalk to confuse the other controller, why wouldn't the same happen between the adjacent coils using a single sensorless esc too?

 

[liveforphysics]

I wonder how much efficiency gain exists with a sine wave controller, since there's less noise. I realize there's not much power in the actual sound, but my gut tells me that jerky commutation at startup that creates the most noise, must mean a less than optimum alignment of the rotor and stator for a portion of the rotation.

Regarding the crosstalk between coils using a pair of sensorless esc's, I don't know. Of course I have to defer to the experts, but If there's enough crosstalk to confuse the other controller, why wouldn't the same happen between the adjacent coils using a single sensorless esc too?

If you look at the windings in an Astro motor, they are distributed between multiple sets of teeth on the stator. This means cross talk is a real issue, because that set of windings moves around to various poles. On something wound like an RC outrunner, where each tooth only has one set of windings wrapped around it, then it's no longer an issue AFAIK.

 

[kenkad]

I am certainly not a proponent of sensorless ESCs. To me, it just does not make any sense. Clearly, at some point, assuming an ideal constant load (which in these applications is a myth), the 3phase drive signal will get 'synchronized' with the motor mechanicals (rotor/stator), but in the meantime, it is simply not efficient. How can you not get cogging? I am also not a proponent of windings like Luke described in an Astro motor. I personally would like to have been able to design a controller (for my application) that did not have fixed common nodes, like the typical delta and wye designs. Unfortunately, this is almost impossible with todays controller component ICs, simply takes too many parts which result in very bad PCB layouts and many other issues as well. Shane and I have discussed how to determine if the motor design is 'good'. The answer is simple, spin the motor up (not under power) and look at the bEMF. This is the only proof you need. If you have a sinusoidal bEMF, then your design is very good. Anything else, and you have made compromises. Then you can add things like flux concentrators/directors (siliconized iron plates on the outside of a dual rotor design, etc.). You will always have flux leakage between coils. This is not an issue if you do not share iron. There are so many small design issues that it's almost impossible to cover every item. Sometimes it is best to ignore them. In some thread, I mentioned that even if you spent $30K for MagNet software, it cannot give you the complete answer. According to a Dr Ramu currently at Virgina Tech, there is a great deal yet to be learned (check out his books at a good technical library). Unfortunately, companies need to sell product regardless if the product (motors and controllers) is less than optimum. That is why we are talking about these things here and hopefully enlightening ourselves in the process. I applaud everyone for asking questions and wish there were good answers for all of them. We all need to forge ahead and learn and share as much as we can.
kenkad

 

[liveforphysics]

I stumbled across this at the Tesla motors site:

http://www.teslamotors.com/blog4/?p=63

It seems they had an AC wave based controller, and they decided to go to a block/PWM drive setup to improve motor control, and torque.

Perhaps this is because they use an induction motor rather than a PMBLDC motor?