Triple Rewind of Unite 500W Motor

[skyl4rk]

How is no load speed measured and what does it mean for performance?

 

[duivendyk]

None of this makes much sense to me .Absent skin effect which is not a factor for the size motors or the speeds we are talking about,the only reason one would use multi stranded windings is to make winding a coil easier and get a better fill factor.The no load speed depends on the number of turns per coil (inversely proprtional),the copper and core loss and the frictional loss of commutator &bearings.So if you reduce the number of turns by two to one the no load speed will go up by a factor of two approximately.The no load speed is measured by measuring the motor rpm with no load imposed on it mechanically

 

[safe]

The Final Analysis

Okay... the data is in... performance evaluated and validated with actual on the bike experience...

So here's the comparison:

http://www.ebikehub.com/forum/download/file.php?id=399

Stock
36 volts
30 amp current limit
3000 rpm no load speed
Rated heat - 150 watts of heat (2500 rpm, 18 amps, 500 watts output)
710 watts peak power

Triple
24 volts
60 amp current limit
3371 rpm no load speed
Rated heat - 150 watts of heat (2900 rpm, 40 amps, 720 watts output)
1010 watts peak power

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

The discussion of theory will go on, but this is the "Actual" result based on the way the motor actually behaves on a real bike in the real world. Why does it behave like this? Well... I'm still refining my understanding... the short answer is that a Triple rewind as I've done DOES improve the performance over a Stock motor. However, I don't think I'm done yet in finding little ways to improve.

Without actual feedback "Research and Development" is just theory...

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

What I see in this comparison is pretty much what the RC motor people experience in that the top end power is stronger and actually runs cooler for an equivalent amount of power. The low end is worse and produces more heat than Stock. (however, despite the added heat the Triple is still stronger down low, it's just more dangerous doing so) You have to ride the bike like it was a peaky race motor and keep the rpms up. This means you are better served with multispeed gearing so that you can get a low enough gear to handle hills. If the bike is geared too tall then overheating is made worse.

So in effect you are "shifting" the useful part of the powerband upwards.

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

No load speed is just the maximum rpm the motor can achieve for a given voltage. On my bike I know the gearing and I have a speedometer, so you can get a pretty accurate estimate based on calculating backwards. I've been doing this since 2006 and have gone through several motors and many voltage and controller combinations and I've gotten pretty good at nailing down the numbers. It's this new stuff... the rewinding... that has added a whole new dimension to what I've been doing. You can effectively change the gear ratio of the bike by changing the motor windings. (that and a bunch of other stuff)

 

[duivendyk]

I analysed your results and concluded that it did not prove that this winding technique improves motor perfomance,if measured by efficiency.The power input and output are higher but the efficiency 720/960x100 = 75% is actually lower than the stock motor
500/648x100 = 77%.Admittedly it produces more power but, if the voltage on the 'stock' motor had been increased to get 720 Watts output,who knows what the efficiency would be.You changed too many parameters,namely both the operating voltage and the no load speed, to draw any valid conclusions.Theoretically if you redesign a motor for a different operating voltage but the SAME operating speed,the efficiency is not much affected,provided the same crosssection of copper is maintained in the armature slots (the fill factor).However if you redesign that motor for a higher operating speed and a correspondingly higher voltage (keeping the number of turns constant),then the efficiency will be higher.The magnetic circuit losses (eddy current&hysteresis) will reduce this somewhat.What the best rpm is for max efficiency I don't profess to know but my guess is that it is pretty high 5k or so, and there well may be other considerations (centrifugal effects brush wear etc) dictating max speeds.
If you want to make a valid comparison,keep the operating voltage and the number of turns per winding constant, but use 'trifilar' or triple wind instead.Basically the no load speed would not change much either,if it went up that would point towards increased efficiency (lower IxR drop in the armature.I predict that you might find some improvement but not more than a few percent at the most.
The moral of the story is the following.In order to keep the losses low, operate at high motor speed,high voltage and low current,use stepdown gearing if high torque is needed for hills,so that these conditions can be maintained as much as possible.Use the controller to get started and keep the current down.The motor losses determine the temperature. increase.Efficient convection cooling (internal fan+ forced air) determine sustainable power.Radiation is minimal at these temps and conduction can only be effective if a heatsink is available (not muchof that around on a bicycle except the rider).Cars are air-cooled too in final analysis.

 

[safe]

Higher Motor Speeds... a "Non-Starter"

Anyone can overvolt a motor.

That's the easiest thing to do to get more power.

But it automatically increases the motor speed and that causes all sorts of problems with gearing and with limits of how fast a motor can practically spin.

So the goal is to go the OPPOSITE way... to create a motor that keeps it's same no load speed while flowing current easier.

My mistake... was in using a thinner gauge wire... that was working against my goal. The Triple wind increases current flow, but the thinner wire decreases current flow. So that's why the results ended up mixed.

The goals can be defined as:

Want to keep the same no load speed and voltage.

Want to flow more current at the same no load speed.

Want to keep the same motor size and same amount of copper.

...the whole point is massive top end power out of a smaller motor.

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

I know for a fact that the Triple has a much higher no load current, so that means that it takes more energy to spin the motor when unloaded. But in a race motor you don't coast much... so as long as the power is good in the peak power area it's okay. (discussions about efficiency at no load speeds are irrelevant when power is the goal because you never spend any time at the no load speed)

If you rode the bike you would know... the power is solid in the middle... much more than Stock would be able to do.

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

4000 no load - Stock motor overvolted to 48 volts

3371 no load - Triple motor undervolted to 24 volts (I want the no load lower)

...the power output comes out about equal. So the goal in all this is to find that perfect "custom" setup that delivers the right no load speed while flowing as much current as possible at the lowest voltage. Using an ACL circuit to limit armature current allows the increase in voltage to whatever level you want to take it. So it's better to start with the LOWEST "k" value (the constant that is volts per rpm) so that as you do decide to ramp up the voltage you start at the lowest point.

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

Hard to believe, but on a nearby downhill I managed 49 mph with this little motor... not a record for the bike at all... but not a bad speed. Top speed on flat land is about 36 mph.

 

[safe]

The Core Question

The core question is whether these two create the same magnetic field, but with differing currents. If you had a Single wind motor that had only one pole (not realistic) then all you would look at is the coil and it's magnetic force as the current flows through it. Assuming that there is unlimited current available from the battery if we switched to a Quad design where the current is divided between the individual turns then you would get a lot more current flowing through the same amount of copper. (assume the wire thickness is constant)

What I really need to do is another rewind where I use the same gauge wire to discover whether this is true or not.
http://www.motoredbikes.com/attachment.php?attachmentid=16611&stc=1&d=1238958384

...just a note, my first rewind project used the same gauge wire and the no load speed was completely unchanged. However, the resistance was much lower and inductance lower. On that Triple I managed a speed record for my bike of 58 mph on flat land while at 72 volts. So this does seem to work because if the no load speed went up my gearing would not have been compatible with doing this. (if the no load speed went up it would have been impossible to do it)

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

A coil is inhibited from flowing more current in two ways:

One: Resistance

Two: Inductance

...the Inductance fades as the rpms rise, but the Resistance is a constant that is tied to the wire thickness. It's the BackEMF that defines where the no load speed is located.

 

[safe]

Bifilar Motors

http://www.cs.uiowa.edu/~jones/step/typesf/4.gif

http://www.cs.uiowa.edu/~jones/step/types.html#bifilar

"To use a bifilar motor as a bipolar motor, the two wires of each winding are connected either in parallel or in series. Winding 2 in Figure 1.4 is shown with a parallel connection; this allows low voltage high-current operation. Winding 1 in Figure 1.4 is shown with a series connection; if the center tap is ignored, this allows operation at a higher voltage and lower current than would be used with the windings in parallel."

I think this in essense confirms what I had in mind...

So I think everything is as I was thinking it "should be" but you just have to be careful about your selection of wire thickness because thinner wires increase the no load speed.

 

[safe]

Two Speed?

http://www.cncmagazine.com/archive01/v2i07/images/v2i07bb.jpg

In the Brushless DC motor world there is the option of Wye or Delta connections so that you can effectively "shift gears". Applying similiar logic one would figure that you could do the same sort of idea with a Brushed motor if you ran a Double wind and could somehow switch the connection from series to parallel while running.

I have no idea how you might do this... you would likely need a dual commutator with two sets of poles so that each wind had it's own commutator, so it's not something you can do easily with motors as they now exist, but it's an interesting idea.

One wonders what happens with the no load speed... in Wye / Delta the no load speed changes significantly, but with Series / Parallel winds one would figure that the current flow would change.

I really want to do a rewind with the same wire thickness and the same number of turns and just change the winds to test this... it's kind of annoying to have no way to confirm this. (I've looked all over the internet and not been able to find any clear description of turns, winds and it's relationship to wire thickness)

 

[safe]

Analyzing The Wire Thickness Error

Here's the easy way to look at why the no load speed went up...

20 AWG wire has a resistance of 0.0333 ohm/m

22 AWG wire has a resistance of 0.0530 ohm/m

...now the turns are the same length no matter what thickness you use, so you can do a simple multiplication :

0.0333 ohm/m * 18 turns * length = 0.5994

0.0530 ohm/m * 27 turns * length = 1.4310

...taking the ratio we get:

1.4310 / 0.5994 = 2.387

...no we multiply the stock no load speed by this and get:

No load speed 36V 3000 * 2.387 = 7162 rpm

...reducing to match the 24 volts I'm actually using:

7162 rpm * 24/36 = 4774 rpm

...but 27 turns of wire actually produces more magnetic force so:

18 turns / 27 turns = 0.6667 or 1/1.5

4774 rpm * 18/27 = 3182 rpm

...which just so happens to match the bike. It's interesting that the Triple wind resistance change (due to picking the wrong gauge wire) added a speed factor of 2.387 and took away a factor of 1.5. So a change in thickness has a non-linear relationship to the no load speed. (probably because a cross section is "pie r squared")


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

The overall resistance of the motor was reduced because of the Triple wind design, but the magnetic flux density remains constant because it's tied to voltage and no matter how much you divide the coil up into separate winds the DENSITY of the magnetic flux is bound by the thickness-to-voltage relationship. It's the density that went wrong... the total magnetic force goes up because of the reduced resistance overall... but the density does not and that's what actually creates the backemf and sets the limit of the no load speed.

Just keep the wire gauge constant... it makes things easier...

 

[safe]

The Better Motor

Had I done this right and used 20AWG, 9 turns in a Double wind which means a total of 18 turns that is the exact same number as the Stock motor then I would have had results more like this:

http://www.ebikehub.com/forum/download/file.php?id=401

...this is what I "should" have done.

This improved design would produce more power and less heat than either the Stock or the last Triple wind that I've done.

You learn as you go I guess...