speedmd
10 MW
Time scales are very different and the area under the curve appears much greater on the 28 pole BEMF trace. This all is fighting the controllers ability to deliver its wave form, is it not?
12 kw rc motor
- Thread starter markobetti
- Start date
speedmd
10 MW
I was under the assumption that back emf was out of phase from what the controller was trying to put into the coils. BEMF going one way and the controller trying to push against it. Certainly I understand that if the wave is the same shape, opposing it would be less of a issue to resolve accurately, but I guess I am assuming less is better.
John in CR
100 TW
bearing said:
Thanks. I guess it's good that I couldn't see it.
With one alignment having the magnets larger than the slots and the other with the magnets smaller than the slots, you'd think there would be some real different in operation other than frequency.
Personally I favor the 20 mag / 24 slot. That's partly because I have what I believe are the best hubmotors you can buy that have that alignment, which are near silent under load even with cheap chinese controllers. The other part is wrt to the commutation of brushless motor with each tooth on a phase having a different positional relationship with the next oncoming magnet. That part is still in a dense fog for me. I can understand how it works, but not how it works so exceedingly well. To me the larger magnets give a better average alignment for longer with the respective teeth.
For some reason, I take it that it's related to startup and or cogging, we don't have 24 magnet, 24 slot motors, but if my slot/magnet alignment thought process has any validity at all, why not 22 mags with 24 slots?
FWIW, something Harold pointed out when he was here the other day and I was determining the wiring for a motor, with no load (not even a wheel) and the correct wiring, then motor nudged reverse by a degree or two when turning the throttle from a dead stop and then it turned smoothly forward. That was running only half of the motor with just one one controller, so every other slot was inactive. With a wheel on and all 24 slots working with dual controllers, I can't detect that reverse bump in rotation. It probably has no bearing on the discussion, but it was interesting to an eager to learn noob in many respects like me.
John
Can you post the pictures or screen shots? I looked and I think its not the same as hooking a scope between 2 phase wires. I think the flux linkage is what you will see from a scope hooked to 2 phase wires. The back emf I seen in your files said Back emf with harmonic... and did not look like what you will see from a scope hooked between 2 phase wires... Maybe I'M not looking at the right image?bearing said:
John in CR
100 TW
Arlo1 said:
Just from the lack of sound from my motors I lean toward Arlo's view that the flux linkage is important. That's the one that jives with smooth real world results using the 6 step cheapie controllers.
Biff
100 W
it is important to note that waveform shape has alot to do with the magnet and tooth head geometry. A very subtle change in the interface can have significant impact on the BEMF waveform (but might not affect Kv that much). When I design a motor, I can spend a significant effort refining that particular geometry, and in the end you might not even notice the difference to look at difference between rev 1 and rev 40, but when you see the BEMF you would be amazed at the difference. Saying that a 28/24 will always have a certain waveform is not accurate.
The second thing to note is that Kv has nothing to do with how much torque the motor can produce. allright, that may seem a little misleading, because obviously the amount of torque that a motor can produce is defined by Kv AND how much current your controller can deliver. When it comes down to it, if you are pushing your motor to the limit of its ability, kV just defines how much current you need your controller to deliver to get the motor to produce the torque you want. A motor's maximum torque will be defined either by saturation of the iron, or when flames erupt from the coils, a good design will have both of those happen around the same current
. You can change the number of turns on a motor all you want to get it to have the Kv you want for your system, but you will not change the maximum amount of torque that the motor could deliver, if you always fit about the same amount of copper in each slot.
John in CR stated that the maximum torque a motor can produce is defined by the amount of surface area of the magnetic interaction, and from what I know that is very much true, when comparing similarly designed machines. You can't compare a motor designed to push a small fan for 30 years without a break, and a rc plane motor based on magnet area, they were designed with different intents, but if you want to compare 2 well designed motors that are both trying to accomplish the same thing, magnet area is a very good indicator as to which motor will be capable of delivering more power when setup properly.
Arlo asked why not have a 24tooth 22pole tooth design. That is a good magnetic design, I don't know why they wouldn't pick that combination. If I were going to design something that spun around 3000 rpm, or was pretty small, I would consider that. If you had budget for good iron to handle high frequency, had a controller that could handle 1kHz commutation frequency no problem, it would even be ok to push up to around 6000RPM.
One reason why brushed motors aren't great in applications where a wide range of speed is desired is because it is difficult to dynamically adjust their timing. As you start spinning faster, the inductance in the system causes a greater percentage of time to change the magnetic field in the coil. With a brushless system you just advance your timing of the FET's switching, with a brushed system you can try and rotate your brush carrier, but that is just another system that will wear out. Typically when you setup a brushed motor you advance the timing a bit to give yourself a compromise between efficiency with starting torque, efficiency at operating speed, and top speed.
-ryan
The second thing to note is that Kv has nothing to do with how much torque the motor can produce. allright, that may seem a little misleading, because obviously the amount of torque that a motor can produce is defined by Kv AND how much current your controller can deliver. When it comes down to it, if you are pushing your motor to the limit of its ability, kV just defines how much current you need your controller to deliver to get the motor to produce the torque you want. A motor's maximum torque will be defined either by saturation of the iron, or when flames erupt from the coils, a good design will have both of those happen around the same current
. You can change the number of turns on a motor all you want to get it to have the Kv you want for your system, but you will not change the maximum amount of torque that the motor could deliver, if you always fit about the same amount of copper in each slot.John in CR stated that the maximum torque a motor can produce is defined by the amount of surface area of the magnetic interaction, and from what I know that is very much true, when comparing similarly designed machines. You can't compare a motor designed to push a small fan for 30 years without a break, and a rc plane motor based on magnet area, they were designed with different intents, but if you want to compare 2 well designed motors that are both trying to accomplish the same thing, magnet area is a very good indicator as to which motor will be capable of delivering more power when setup properly.
Arlo asked why not have a 24tooth 22pole tooth design. That is a good magnetic design, I don't know why they wouldn't pick that combination. If I were going to design something that spun around 3000 rpm, or was pretty small, I would consider that. If you had budget for good iron to handle high frequency, had a controller that could handle 1kHz commutation frequency no problem, it would even be ok to push up to around 6000RPM.
One reason why brushed motors aren't great in applications where a wide range of speed is desired is because it is difficult to dynamically adjust their timing. As you start spinning faster, the inductance in the system causes a greater percentage of time to change the magnetic field in the coil. With a brushless system you just advance your timing of the FET's switching, with a brushed system you can try and rotate your brush carrier, but that is just another system that will wear out. Typically when you setup a brushed motor you advance the timing a bit to give yourself a compromise between efficiency with starting torque, efficiency at operating speed, and top speed.
-ryan
Yes, Kv on it's own doesn't tell you much, it's Kv in relation to Rm that's important. What I was struggling with is that, if increasing the pole count lowers Kv without increasing Rm then, you must get more torque for the same copper losses.Biff said:The second thing to note is that Kv has nothing to do with how much torque the motor can produce. allright, that may seem a little misleading, because obviously the amount of torque that a motor can produce is defined by Kv AND how much current your controller can deliver. When it comes down to it, if you are pushing your motor to the limit of its ability, kV just defines how much current you need your controller to deliver to get the motor to produce the torque you want. A motor's maximum torque will be defined either by saturation of the iron, or when flames erupt from the coils, a good design will have both of those happen around the same current
Actually I was asking for 20 magnets with 24 teeth but the winding factor for 22 magnets and 24 teath is a bit better. As far as I'm concerned lower magnets is better for now you will have a weight penalty because lower magnet counts will need thicker back iron (outer can) but if I had to chose between a 4kg paper weight or a awesome 4.5kg motor that can be run with a cheep 24 fet what do you think I would do? It doesn't make sense to force people to buy $1000+ controllers for a simple change like this. Yes I admit inductance will be a roll in either motor but as cheaper sine and more advanced china controllers come out they will be limited in the max electrical frequency... So I think Marko and Hall and Accountant need to understand they can make a motor that will sell if controllers are cheaper for it!Miles said:
Biff
100 W
possiblyMiles said:
John in CR
100 TW
Biff,
Thanks so much for sharing all that. Very interesting about the tooth geometry. Once the design is good, can you increase the stator and magnet lengths and expect the same good result but higher power, or are you back to only somewhere between revision 1 and 40?
Exactly my thinking, but I'm a cheapo and want the motor price down closer to the $300 range initially mentioned a few years ago. Another thing I like about 20mags/24slots is that I know 100% that it works well as the simple 6 phase. That makes the controllers even cheaper.
John
Thanks so much for sharing all that. Very interesting about the tooth geometry. Once the design is good, can you increase the stator and magnet lengths and expect the same good result but higher power, or are you back to only somewhere between revision 1 and 40?
Arlo1 said:
Exactly my thinking, but I'm a cheapo and want the motor price down closer to the $300 range initially mentioned a few years ago. Another thing I like about 20mags/24slots is that I know 100% that it works well as the simple 6 phase. That makes the controllers even cheaper.
John
Biff
100 W
John in CR said:
with a radial design for the most part you can just increase length without affecting the shape of the BEMF. The only case that I can think of where it is a little more complicated is when you have stator / rotor skew , which by the way is an easy design way of making almost any BEMF nice and sinusoid, I say easy design, because it is easy to design skewed magnets / stators, it is more difficult to actually get someone to make it, and it typically adds cost.
-ryan
bearing said:Thud said:
Probably not nonsense. You obviously get more torque (per amp) with more poles, with no penalty in added winding resistance. But the peak "saturated" torque might not rise with more poles, if Biff is right. So if you are looking to use the motor at the absolute max, more poles may be a disadvantage. That's how I conclude the latest "news" on endless-sphere
correct,
you get more efficient torque production from these high pole count motors, of course your right miles that you get more NM per amp this way without added ir so thats how its done.
but the peak torque (saturation limited) does not rise with it, so you end up with a motor that has very little difference between the max continuous torque and the max peak torque, so yes for arlo or others who like to talk unrealistic (or at the very least unsustainable) power levels like 30kw+ for these motors that could be seen as limiting.
however, the way we put a max phase amp setting in to the controller (that in real use has to be pretty sustainable) these kinda motors are gold because
they can then make the most torque you can have from its size and weight and its very made very efficiently with super low ir, saturation is not far away but you will never see it if you set the max phase amps wisely.
the ca120 is 118.4mm od and the air cooled colossus i have here is 120.4 od (can) and yes the stators are the same lams (in fact the exact same dudes hand writing 1 to 24 on the stator teeth
) so yes the ca120 has waay to thin back iron (that showed severe saturation on mileses femm diagrams in the motor comparison thread) but when i added an extra 2mm of normal steel strip attached to the outside of the can, the max (completely saturated) torque did not change even a bit.incredibly though, at more reasonable torque levels (about 15NM for the ca and 30NM for collosus) these motors can make the NM's with slightly less total loss than the same weight of Joby motor, beyond that (and everywhere else) the jobys win with a higher saturated torque even with less copper.
What Im getting at is with a rewind and using a sevcon for instance you are still limited to 400hz which is 1714 rpm max so what was that about max phase amps..... You can only build so much HP at 1714 rpm.... Yes the china controllers will run higher E-rpm. And some of us are still trying to bring a cheep controller to the game... BUT if I was going to bring a motor to market It would be something easier to run and as far as I'm concerned the new 28 poll 24 slot motor is harder to run not easier.toolman2 said:bearing said:Thud said:
Probably not nonsense. You obviously get more torque (per amp) with more poles, with no penalty in added winding resistance. But the peak "saturated" torque might not rise with more poles, if Biff is right. So if you are looking to use the motor at the absolute max, more poles may be a disadvantage. That's how I conclude the latest "news" on endless-sphere
Who is buying this motor to go slow?
SplinterOz
1 kW
I didn't buy it to go slow. So
110km/h on a full motorbike with a controller cheaper. than your sevcon.
110km/h on a full motorbike with a controller cheaper. than your sevcon.
I think Splinter got his controller for < $1000 ( we will wait for him to chime in) its the only one so far although it should not be long until more are available. But one Sine controller from Kathy on the forums here can't do it because of the E-rpm problem....swbluto said:
SplinterOz
1 kW
The 72volt 600amp controller for under 1000 and the 700amp for 1100 will get you to 14kw (rear wheel). Now this is with prototype motor,toolman2 and I have confirmed that the new one (I have an air cooled one) should get close as I am probably just short of saturating the current model.swbluto said:
This should be even more of an advantage with smaller motors and their correspondingly shorter thermal time constant. Kt falls off the closer you work to saturation, though - that might temper the gain a bit. I haven't completely got my head around pole count WRT saturation, yet........ :|toolman2 said:
I'm still unsure whether pole count increases kv. If it does, it probably has to do with things like saturation of tooth, back iron, and similar. If you look at the simple math:
e = -dΦ / dt
dt decreases with pole count. So with constant flux change, voltage should increase. But is the change in flux constant with increased pole count? When I make simple simulations with linear materials I get the same generated voltage independent of pole count, which suggests the change in flux decreases proportionally.
Another formula I remember:
e = Bvl
Pole count will not increase velocity. Magnetic field (in air gap, I assume) should not change. Length does not change. So the induced voltage/kv should not change. But this formula is maybe only for coreless motors?
Let's go back to the first formula. And let's look at the case of a magnet aligned with a tooth. With increased pole count, that magnet is smaller. This means proportionally less flux/change in flux?
e = -dΦ / dt
dt decreases with pole count. So with constant flux change, voltage should increase. But is the change in flux constant with increased pole count? When I make simple simulations with linear materials I get the same generated voltage independent of pole count, which suggests the change in flux decreases proportionally.
Another formula I remember:
e = Bvl
Pole count will not increase velocity. Magnetic field (in air gap, I assume) should not change. Length does not change. So the induced voltage/kv should not change. But this formula is maybe only for coreless motors?
Let's go back to the first formula. And let's look at the case of a magnet aligned with a tooth. With increased pole count, that magnet is smaller. This means proportionally less flux/change in flux?
Decreases, you mean?bearing said:
The only attempt at explaining this, that I've come across, is on page 32-34 of Brushless Motors & Controllers.
http://books.google.co.uk/books?id=lKs7xV_t5iIC&printsec=frontcover#v=onepage&q&f=false
Similar threads
