The AC Induction Motor Ebike Project

[spad4me]

Safe when you get ready to make the controller board? Are you going to make it yourself?
If you are going to have it produced by some one else..
If you need a minimum number .
Count me in for a couple at least.
You may want to post a note for board orders.
Or sales.


I thought I had found a chineese connection for three phase motor controller boards .
The guy dropped off the face of the earth .

 

[safe]

If you noticed on the "Rewind" thread I've been riding the other bike a lot recently after finishing that rewind. My plan is to spend the winter researching the way to build the AC variable frequency drive given my circumstances at the lowest cost and the least complexity I can.

So don't hold your breath...

Also, I need to rewind the AC motor itself and then probably start off testing it with just regular household AC to see that it works without a controller. (to make sure I don't get the wires backwards or something)

...the whole process will take months I suspect.

I'm not planning to build more than one because I don't even know if it's going to work despite my best efforts to learn all I can. What would be great is if some guy that has 20 years of AC motor design theory took an interest and started posting here.

On that one I'm not holding my breath.

-------------------------------

There's always a "Plan B" which is rather than running the motor at 48 volts I can go to double that (96 volts) or triple (144 volts) or even more. I have 12 sets of 24 volts, so there are many combinations that are possible. It's also possible to use voltage doubling in the circuit... which is "Plan C" or so.

There are more unknowns in this project than knowns...

 

[spad4me]

Then I shall continue in my own way.
As I see fit.

I asked you if you were going to abandon this idea.
You said look here. I am here.

Just an observation.
If you really want to go for switched reluctance just buy a motor and a controller and go for it.

You moved aussie jester to rage and got him banned..
Not gonna happen to me.
LOL.
I will simply stop posting and let you fill the pages with interesting but useless minutiae.
Until I finish.

This is now the safe area.

 

[safe]

You moved aussie jester to rage and got him banned..

Aussiejester worked himself up into a rage... I really never did anything to cause it. (so don't blame me)

I've been a supporter of the concept of legal ebike racing (750 watt) since about 2007 and AussieJester has chosen the path of the "Outlaw". These things just happen. We parted ways intellectually.

This AC Induction motor project only has a reason to exist if you accept that the laws ban power levels above 750 watts. If you do not care to respect the law then the RC Brushless motor gives far better performance on a power-to-weight basis.

Frankly I wish he didn't get banned... he does great work.

(despite the abrasive personality)

---------------------------------

spad4me, you are welcome to post insights about your own project here if you want or it might be a good idea to start your own thread to document your progress. The easy thing to do is to get a regular 120 VAC motor and just use 120 VDC power as a base. Do that and it's pretty much doable with the motors and VFD's that you can buy right now. Try to redo everything to make it optimal for ebikes and safe (48 volts) will require some effort. You might try doing things the easy (but more dangerous) way first.

As someone pointed out:

"You only get electrocuted once."

...at 120 VDC that's a near certainty. At 48 VDC it's more probable you live.

 

[safe]

Chorus Motors

I was pointed to looking at this:

http://www.chorusmotors.com/technology/index.shtml

...they are big on the idea of increasing the number of phases and then using more complex harmonics to control the waveforms. It's really far out stuff and the people that are doing the work must have Phd's in electrical engineering and physics. (very intelligent)

For my feeble efforts the motor I have to use has a total of 24 grooves that I can wire up any way I want. The most obvious way is to go with Three Phase and 4 Pole because since each pole needs two grooves that adds up perfectly to 24 grooves.

Another way would be to use the lower Pole number of 2 Poles and then increase the Phase count from Three Phase to Six Phase. The advantage of this is that it turns out that the Power Factor is actually better with the LOWER pole counts because the magnetic field is smoother and less broken up into pieces. The more swirls you get into the magnetic field the more eddy current losses you get.

The negative issues with Six Phase power are that the controller gets harder to build and the base speed of the motor gets doubled compared to the 4 Pole design. Since I'd prefer not to invent more than I need to and I want that low rpm motor it makes more sense to stick to Three Phase and 4 Pole. But the idea of Six Phase is definitely one to consider... especially if you can do all this super complex multiphase harmonic stuff.

Maybe after I get a Phd...

 

[safe]

Poles And Power Factors

It would be great to think that you could get a low rpm motor by going to increasing pole counts and not pay a price for it. Unfortunately that's not reality and the more you subdivide into increasing poles the more the Power Factor falls. The Power Factor is sort of the "best case scenario" for getting high performance from the motor. It's a little different than efficiency because the AC Induction motor is... well... sort of weird compared to the very simple permanent magnet motors.

Ultimately I have to ask myself which is more important... is the most important thing to have a low rpm motor so that I can make gearing easier? Or should optimal efficiency (Power Factor) be the main focus because the larger goal is to produce the best 1000 watt machine?

If I went back to the 2 Pole design I could jump up to Six Phase power which will allow me to drive the rotor much harder. The more phases you have the less current needs to "peak" in each phase (shorter time) so that lowers heating. Also. the more spread out the phases the less distortion in the magnetic field, so the rotor actually stays cooler.

Maybe I need to accept higher rpms in order to chase after peak efficiency?

Hmmmm.....

It looks like if you want the higher power factor you have to go 2 Pole... (90%)

One needs to remember that for a brushed DC motor the typical equivalent to the Power Factor is "peak efficiency" which occurs at a specific (usually high) rpm. The best the brushed motors do is about 80%, but the average across the usable powerband is closer to 70%, so it doesn't look that hard to do better than the brushed motor even at 4 Pole. Looks like you could even do 6 Pole and equal a brushed motor. (which goes to show how inefficient they are)

How greedy for perfection do I want to get compared to practicality?

Note: A simple minimal geardown would be enough to allow the high rpm motor easily, so it's not that hard to do it. (2:1, 3:1 or 4:1 is enough)

 

[safe]

Who Wins The Race?

Begin with a 1000 watt input limit, then ask yourself:

"If you can apply any technology in the world to get the most out of this power what is the best way to do it?"

So let's think of the options:

DC Brushed Motor - Skip it, the effficiency is low and reliability bad.

DC Brushless Motor - Uses permanent magnets and so in order to attain the highest efficiency you will need multispeed gearing (mandatory) and since the motor speed tends to be high you will need a geardown. Let's assume a peak efficiency of 90% and losses for geardown of 3% and multispeed gearing 2% for a total of 5% in losses. Final "effective peak efficiency" is 0.90 * 0.95 = 0.86 (86%) However, that's only at the perfect rpm and is NOT based on load, so you will need to get an "average" efficiency as you go through the gears. We need to figure that another 3% (minimum) is lost to being "off peak". 0.86 * 0.97 = 0.83 (83%)

2 Pole AC Induction Motor - 90% Power Factor. Efficiency at full load is about 95%. In order to make the gearing work you will need two chains, one for the motor-to-rear-wheel and a second from pedal-to-rear-wheel. The direct geardown chain drive will lose about 2% efficiency. So we get 0.90 * 0.95 * 0.98 = 0.84 (84%)

4 Pole AC Induction Motor - 85% Power Factor. Efficiency at full load is about 95%. In order to make the gearing work you will need only one chain which will lose about 2% efficiency. So we get 0.85 * 0.95 * 0.98 = 0.79 (79%)

...so this might tie my hands. If I don't go with a 2 Pole design then it's possible for a highly optimized permanent magnet machine with multispeed gearing to out perform the AC Induction motor machine.

I might be forced to use two chains... (not a big deal, but it's better to realize this now rather than later) However, by using two chains I can get rid of the front freewheel and return to a freewheel on the rear. Motor power can be delivered on the left hand side and the pedal freewheel would be on the right. (which allows for regen for the motor chain) This would make for a simple and clean design.

 

[safe]

One Chain Is Better

One chain is less complicated... it just makes the machine look like it's actually well thought out and not some compromise. But the stock bicycle chain simply cannot work with a 3600 rpm motor... the numbers do not add up and so one gives up... or do they?

Break the mold!!!

No one said that there is a restriction on chain sizes, there's no reason to not switch to something like a #35 chain like the go kart racers use. After all, if these EBRR race bikes are intended to be used on go kart tracks you might as well share the same parts as they use.

So this is how you can achieve the one chain goal:

...you can see that the motor sprocket uses an 11 tooth sprocket, the pedal sprocket is larger (114 tooth) and has a front freewheel to it. The rear wheel has a 90 tooth sprocket (no freewheel), but you can vary the number of teeth depending on the top speed you desire. (most of the go karts use from 80 - 100 tooth sprockets) The idler sprocket is needed to make sure that the motor sprocket gets a "full wrap" of the chain to prevent slipping. This is something I've already learned the hard way. (you need it)

It looks possible to also maintain regen while also having a freewheel for the pedals... this might be "the solution" to the AC Induction motor gearing. From 0-10 mph the pedals allow for a strong pull off the line, (or steep hillclimbs) then from 10-30 mph the motor kicks in and delivers peak constant torque and after 30 mph the motor continues to deliver constant power at reducing torque. (unless you are forced by the law to use a frequency limiter to prevent higher speed than 30 mph) This is about as close to the ideal as I'm able to imagine.

The 2 Pole design is in fact possible without needing two chains or ANY extra geardown.

Simplicity is beauty...

.

 

[safe]

Past Lessons With Rewinding

I had posted this early on in my rewinding thread. Basically the idea I was wrestling with related to the brushed dc motor is whether you can split the coils up into multiple "winds" and let current flow in parallel. I "thought" at first that I could do that with rewinding and get higher current flow and that would translate to more power.

I was wrong, sort of.

...the REASON I was wrong was because of the way that the no load speed is determined with the brushed dc motor. No load is determined by the level of backemf that gets created in the magnet wires as the motor spins to it's maximum speed. If you split the coils up into separate sections it does increase current flow at lower voltage, but it takes more motor rpm to create the "blocking" backemf to attain the maximum speed. There is a limit to motor top speed in both an electrical and practical (geardown) sense.

The AC Induction motor seems to be different to some degree. There is no simple backemf to deal with... the rotor does create a magnetism, but it's a LAGGING field and the principle force is "slip". As a result the no load speed is based on AC frequency rather than voltage. However, the problem with AC Induction motors is that with no backemf the current can flow with virtually no limit if you aren't careful and that means the potential for runaway heat.

What does this mean?

Well, apparently some of the "silly" ideas I had with the brushed motor actually are the way things work with the AC Induction motor. In order to lower the natural voltage of a motor you would want to rewind with an eye towards more parallel coils rather than just one. There are even AC Induction motors that are classified as "Two Speed" where you can run the motor as either series or parallel wound.

Sometimes the dumbest ideas come back to be useful.

 

[safe]

Using The Online Calculator

http://hyperphysics.phy-astr.gsu.edu/HBASE/magnetic/solenoid.html#c1

Let's try to use numbers that would make sense for an AC Induction motor which is designed for 120 VAC but will run at 48 VAC. Let's assume:

120 turns of wire (to match the 120 volts)
1 amp of current
0.0120 length of the solenoid

Result: 2.51 Tesla

48 turns of wire (to match the 48 volts)
1 amp of current
0.0048 length of the solenoid

Result: 2.51 Tesla

--------------------------

Notice that in the first case I'm filling the entire groove (0.0120) with one set of coils. In the second case I'm filling just part of the groove (0.0048) with one set of coils. In order to fill the entire groove the smaller turn count (11) needs to be almost tripled. (120/48 = 2.5) This is analogous to the multiple winds approach and will in the end mean you need to use 2.5 amps to achieve the same result overall.

Assuming that you filled the entire groove (0.0120) then you would need:

48 turns of wire (to match the 48 volts)
2.5 amps of current
0.0120 length of the solenoid

Result: 2.51 Tesla

...so one can use thicker wire and lower turns or use more parallel winds with the same wire thickness. This is all the same stuff as with previous rewinding that I've done. What is different is that there is no difference in the motor's no load speed.

Inductance is the wild card in all of this... in electric motors the higher the inductance the longer the magnetic field tends to maintain itself. (it acts like a sort of dampened spring that holds onto it's magnetic state) Given the way that low rpm tends to draw a lot of current anyway it's possible that rewinding the motor with lowered inductance will mean more worries about runaway conditions.

My "hunch" is that higher voltage and higher inductance produce more efficient motors because the magnetic fields are more stable. (higher inductance smooths out the magnetic field and reduces eddy currents) In order to compensate for this the use of Six Phase rather than Three (or Single) Phase power would do the same thing as increasing inductance. This "might" be a case where the combination of using a lowered inductance rewind with a smoother running Six Phase power could negate any losses.

Much to ponder I suppose...