unime

What do you want to accomplish?

If your motor is running hot, then rewinding (with the same number of turns and thicker wire) might just help. Then again, it might not. Your losses are probably dominated by eddy currents in the core and poor connectivity in the magnetic circuit, in which case lowering resistance won't make much difference.

If you want a higher Kv constant for a faster no load speed, you can rewind with fewer turns. Adding turns will increase torque. But there's no free lunch. If you rewire a motor to pull more power at low voltage, current drain will increase, and you'll have to be careful to avoid reducing battery life.

Use multiple windings if you can't get the desired density with a single conductor, or try your luck with square cross section wire to get the same effect. There's no magic to multiple windings - it all boils down to the wire's total cross sectional area and the number of turns. You may get better density with multiple windings, especially when the wire starts to get thick (say >22 gauge for small motors - I've never worked on hub motors). Be sure to make your windings neat and tight: Wires that vibrate in use waste energy.

safe

No Magic?

The idea is that by dividing the coils into separate windings you also divide the inductance.

Actually if you play around with the spreadsheet it makes sense... you can achieve the same resistance with differing configurations, but in the Single winding case the inductance is highest and so the "flywheel effect" of the inductance is greatest and that tends to enhance low end torque.

When you increase the "Winds" you are making the inductor more "parallel" than "series".

In the Single case if you have 40 "Turns" then the inductor behaves as two sets of 20 "Turns" connected in series.

In the Double case if you have 40 "Turns" then the inductor behaves as two sets of 20 "Turns" connected in parallel.

The top speed of the motor is identical (because they share the same number of total "Turns" of the same wire) but the powerband is shifted so that the power is best up top with the Double and the low end torque is the best with the Single.

Go here:

https://www.rccartips.com/advanced-rc...motor-tips.htm


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


Number of Turns
Refers to how many times the copper wire is wound around each pole of the armature. Basic guide:

More Turns (e.g. 19T) = Higher torque, less rpm, longer battery life. Slower but easier to drive.

Less Turns (e.g. 12T) = Less torque, more rpm, shorter battery life. Faster but more difficult to drive.


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

Number of Winds
Besides the number of turns, the number of winds refer to the number of wires wound around the armature. Basic guide:

Single Wind = Most bottom end power. Power is achieved at lower rpms. For short racing tracks with a lot of turns.

Double / Triple / Quad = Less bottom end power. Power is achieved at higher rpms. For long racing tracks with long straights.



***********************************


What do you want to accomplish?


The point of the thread was to ask how one might improve on the abilities of an ebike. My experience with these motors is that trying the "brute force" approach of using more voltage and current just causes too many problems with heat. (mainly in the low end) The smarter thing to do is to increase the top end power (through something like a rewire of the motor) and then fiddle with something like an Armature Current Limiting controller so that you can take the voltage up more without getting the heat built up in the low rpms.

We could diverge a great deal in performance tips here...


Within the existing stock of ebike motors there is about a 50%-50% mix of Single and Double winds. Going beyond a Double to a Triple or Quad would require so many amps to make useful (because the resistance drops from Single 1, Double 1/4, Triple 1/8, Quad 1/16) that most people cannot achieve the necessary amps to keep the motor from "starvation". (this is why too low of a motor resistance is actually counter productive) Most stock motors tend to be setup to favor low end torque too much.

Lowering the resistance translates directly to less heat because Heat = Amps Squared Times Resistance.



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


40 "Turns" is 40 "Turns".... if you feed them current they make magnetism.... but you have a choice of feeding those 40 "Turns" from a Single source or more than that... the motor can't tell how the amps got there... it can only tell that amps exist or don't exist.

.

safe

Heat

Heat = Current ^ 2 * Resistance

...so if we looked at the most simple case of a two resistors and have the choice of placing them in series or in parallel and then apply a voltage to those circuits then the series circuit will resist the flow of the current (since V = IR) more than the parallel circuit will allow.

Series: R = R1 + R2

Parallel: 1/R = 1/R1 + 1/R2

...in our imaginary case we set the resistance of all the resistors to one and get:

Series: R = 1 + 1 = 2

Parallel: 1/R = 1/1 + 1/1 = 2 -> R = 1/2

...which means that the parallel circuit has one fourth the resistance of the series circuit.


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


The real "problem" is in holding back the current on the Double, Triple and Quad motors because they desire more current to run better. So despite the lowered resistance you can still end up with the same heat (or more) because there is less to stop it from happening. (in fact you are forced to use more current because the motor demands more)

The Single has a sort of built in current protection... but the others can draw so much current that heating is actually worse... you're getting more power because in effect the smaller voltage is "acting" like a larger voltage, but it's still increasing the heat.

So "heat management" actually becomes WORSE with the Doubles, Triples and Quads...

Make sense?



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


In a crazy kind of way the Doubles, Triples and Quads are behaving in a way that makes the battery seem like it has more voltage and just like more voltage in the classic sense you get all the problems that increasing voltage permits.

So some kind of very strict current limiting is even more important if you did this...

.

unime

No. That's not what is going on with multiple windings. When you connect several identical windings on the same core in parallel, the inductance doesn't change. That's because the same magnetic flux (flowing through the core) is generated by a given current flowing through one winding as (1/n)th the current flowing through n windings.

Honestly, I don't know what the RC car folks are talking about. Multiple windings shouldn't make much of a difference unless they are changing the total amount of copper. I'm not impressed by claims made on a random web site. I do know Maxwell's equations, but I don't see how they might explain those claims.

Maybe they are talking about doubling the amount of copper (e.g. one winding of 22 guage wire vs. 2 windings of 22 gauge). In that case, acceleration would be slowed in a double wound brushed motor due to the extra rotational inertia. Technically, this is different from torque, but I can understand someone confusing the two effects.

I don't deal with brushed motors very often. The brushless crowd does not talk about these effects, giving me more reason to suspect that this is mostly about rotating mass in the motor, not inductance, and only a little about resistance.

As I said before, you need to find the source of the inefficiencies. Look at the core of the motor, air gaps and other leaks in the magnetic circuit, etc. Adding copper (whether by using thicker wire or multiple windings, it doesn't matter which) generally help, but may only make matters worse if you don't limit current as the lower resistance will allow more power to flow at low speeds when the motor is lugging.

In practical terms, there is a lot to be said for running moderately high voltage (say 48V) to get the desired top end, with current limiting to prevent overheating. Rewiring the motor does nothing to improve efficiency (unless you increase the total amount of copper), so you get everything you need using this approach.

No, no, no! Motors are reactive devices. You cannot understand their function by looking primarily at resistance. Excepting the case of a stalled motor, your reasoning is nonsense.

alfonsopilato

can I just say I love my ebike
.. been going to work with it since last summer.. i like it.. i like it a lot

ok.. regarding the 36 volt topic: yep, anything to get you biking more is a far healthier alternative. i think i've been biking far more (actually pedalling) than i ever have, because i have an assurance that if i do get tired, i always have my trusty vroom vroom.. it's really a cool mariage (euhh. is that two r's?)

safe

I just don't understand why you aren't getting this...

Using the SAME amount of copper you can change the inductance and "effective" resistance by the choice of series or parallel. I mean this is basic stuff here. In order to achieve the same current through each wire (given the same voltage) you need to apply more current from the battery. Every time you divide the copper into more and more individual winds it reduces the resistance that the battery sees. The net effect is to make a 48 volt battery seem to act like a higher voltage battery.

It's just basic series verses parallel stuff...


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


Try to explain to me WHY you think that this isn't valid in some other terms than that you just don't want it to be.


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


The different motor options that Unite has all share more or less the SAME core. The only differences are the wiring patterns and the thickness of the copper. Some have more "Turns". Some have Single winds and others have Double winds. In the end the motor behavior is defined by this mixture of properties.

They don't do Double winds just because they were too lazy to choose a larger gauge wire.


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


From what I've read the thinking goes like this:

Inductance makes a difference at the low end of the powerband because the PWM controller can maintain the magnetic field longer. Remember how these controllers work... they are sending pulses out at differing widths depending on the throttle setting and the current limiting that is being used. As the rpms rise to the third quarter of the power band the other forces within the motor tend to dominate and the inductance reduces as a factor. It's at these higher rpms when the BackEMF starts to become the dominant factor and not the inductance and a lower resistance (due to things being in parallel) means the battery can supply the needed current more efficiently.

What is happening with the Double wind is the lower rpms are being given less attention so that the higher rpms (where all the real horsepower is found) are enhanced. The BackEMF is the constant that you can't get past because that's the relationship of the permanent magnets to the magnetic field you are generating. The maximum rpm ends up the same, but the motor is more efficient up top with the Double and the Single is stronger down low.

The RC folks are pretty smart people...



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


One last thing to ponder.... as the rpm rises the current actually DROPS with an electric motor. So while the Double, Triple and Quad looks like it should promote an "uncontrolled meltdown" of excessive current with rising rpm it really doesn't. You have to think:

Low rpm - Inductance is the dominant factor

High rpm - BackEMF is the dominant factor

...so the multiple winds are just "shifting" the motors powerband to the right and optimizing for a more narrow (higher power and efficiency) region of the motor.


A Double can make a 36 volt battery "behave" like it was 48 volt as far as power and efficiency, but at the expense of the low rpm compared to the Single. You don't get something for nothing, but you do get the best parts of the higher voltage without needing to use a higher voltage. The Double is just a better wind in many cases. Also, by choosing a Double wiring over a Single the top speed of the motor is the same... in many cases overvolting can mean that you change the top speed so that you spend more and more time in it's lower rpm region and that hurts the efficiency even more. It's sort of a trap when you go up in voltage because the heat goes up along with such a change.

.

safe

Words From The Pros... Fanthom Racing

https://www.fantomracing.com/tips/mm.htm

CHOOSING THE RIGHT NUMBER OF STRANDS

To give you a simple explanation, the number of strands you choose, i.e. single, double, triple, quad, quint, or hex, determines the power band of the motor. For example, a single produces its’ maximum horsepower in the lower RPM range, while a quad produces its’ maximum power in the mid to upper RPM range. In other words, a single will feel like it has more “punch” coming out of a corner, while a quad will seem to kick in around mid straightaway. Only you can determine what best fits your driving style, but a good place to start is with doubles. Normally speaking, a good system to follow is to base your decision on the current track conditions. Usually tight technical tracks are more suited for singles and doubles, while big open tracks may require a triple or quad. Also, traction can play an important role. For instance, if there is a lot of traction, you can get away with running a single. If the traction is sketchy, where you can’t hook up well coming out of a corner, you might be better off using a triple.

And this is the rationale that they have for "why" this actually happens... which is interesting...

https://www.fantomracing.com/cstm_mod.html

With today’s high voltage/long runtime batteries, the most common winds now being used are singles and doubles, however, there are situations where multi-strand armatures can still be utilized effectively. In simple terms, there are two main things to consider when choosing a wind; MCA and SA. MCA is the abbreviation for Mill Circular Area, and SA is the abbreviation for Surface Area. MCA is the area measurement of wire, viewed from the end, 1 mill thick, meaning larger gauge wire equals more MCA. For example, larger gauge wire, such as 15awg, 15.5awg, etc. is used on singles. SA is the amount of surface area measured around the circumference of the wire. Generally speaking, by adding multiple strands of wire the SA can be increased significantly with the more strands of wire being used. You can go even further by winding split winds (i.e. one large and one small wire), with which the smaller wire helps fill in the gaps allowing for finer tuning of the MCA/SA combination. Typically we don’t run many split wind armatures, but they are interesting to experiment with in certain situations. In very simple terms, more MCA produces more torque, and more SA produces more top end performance. Given the same MCA, more SA will move the power curve up, but use more battery at the same time.

safe

Comparing Apples and Oranges

The 750 watt 36 volt Unite motor is a 15 turn Single. 3250 rpm.

The 1200 watt 48 volt Unite is (was) a 14 turn Double. 4000 rpm.

4000 * 14/15 = 3733 rpm... so there are some other factors in play here like the width of the armature and possibly the thickness of the wire, though it's hard for me to tell much difference in cross sectional area. (the 750 watt seems a little thicker)



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


However the thing that seems to be true is that it's the "Turns" that matter in motor rpm.

"Winds" seems to effect the shape of the powerband and not the top speed.


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


So if you divide a Single into a Double you are going to have to increase the motor speed accordingly because the turn count will be cut in half. It might be necessary to regear the bike to deal with higher rpm. I would really like to see something like a "Motor Calculator" that spelled out the effects of how one might wind their motor as far as speed and torque.



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


So if someone asks:

"How can I improve my ebike so that it goes faster?"

...the answer is that they can add more battery (increasing the voltage and increasing the heat) or they could rewire their motor so that it shifts the power up higher while also increasing the motors top speed while also making low rpm torque weaker.


I just got some 22AWG 200C Magnet Wire for about $20:

unime

Safe - you can quote equations all you like, but that doesn't mean you are applying them correctly. The parallel inductance equation simply doesn't apply when the coils are wrapped around the same core - in the shared core case, inductance remains constant. Your failure to understand this does not change the laws of physics.

Here's a basic explanation of what's going on in an iron core inductor (like the electromagnets in a motor). Current flowing through the coils generates a magnetic flux through the core. The flux is proportional to the number of turns of wire times the current flowing through that wire (and depends on other things like the cross sectional area of the core, and its magnetic permeability). Flowing A amps through N turns produces the same flux as A/K amps flowing N*K turns, and it makes no difference whether the turns are wired in series, parallel (or are individual superconducting loops). Ignoring resistance, the voltage across N turns around the core depends only on the derivative of magnetic flux. Each of the K independent winds in the multi-turn motor will have the same voltage as the single wind motor, and connecting them in parallel is electrically and magnetically identical to the single winding case.

Understand?

If you understood my explanation, you will now know that the only factor influencing inductance is the number of turns. You can change resistance and weight by altering the amount of copper. That's it. Period. End of story.

This is wrong, wrong, wrong. First, you divide the current exactly as you divide the windings. Second, you cannot hope to understand the current flowing through a motor by reasoning about its DC behavior. Third, splitting a wire into separate winds has no effect on resistance (as measured from outside the motor) if you keep the same amount of copper.

Of course, not. In the end, each winding strategy will have slightly different cross sectional area, resistance, weight, packing density, amount of vibration, surface area for cooling, etc.

I get the distinct impression that they use more copper when they do double windings. This adds rotational inertia to a brushed motor, slowing its acceleration, but decreases resistive losses, increasing efficiency. The side effects (in bold) of extra copper as precisely the claims they make for multiple windings, and have absolutely nothing to do with impedance.

Note that bicycles accelerate extremely slowly in comparison to hot RC cars, so the effect of adding extra copper on acceleration is far less noticeable. It is mainly a matter of how much wire will fit, its cost, and weight.

Back EMF depends precisely on (the derivative of) magnetic flux and inductance. The concepts are fundamentally linked. Permanent magnet motors have a Kv constant that expresses back EMF as a function of speed (which is proportional to the derivative of magnetic flux seen by the cores) - there is no need to specify inductance separately when modeling these motors.

safe

Litz Wire

I admit that at first I thought you could switch from Single to Double without having to regear (doing this switch increases the motor rpm) but the effect that the "pros" in RC racing talk about really does seem to exist.

Check this link:

https://en.wikipedia.org/wiki/Litz_wire

Another way to explain the same effect is as follows: the magnetic fields generated by current flowing in the strands are in directions such that they have a reduced tendency to generate an opposing e.m.f. in the other strands. Thereby, for the wire as a whole, the skin effect and associated power losses when used in high-frequency applications are reduced. The ratio of distributed inductance to distributed resistance is increased, relative to a solid conductor, resulting in a higher Q factor at these frequencies.

In cases involving multiple wires, or multiple turns, such as windings in transformers and inductors, proximity effect causes losses to increase at high frequency even sooner and more rapidly than does skin effect.

Litz wire is used to make inductors and transformers, especially for high frequency applications where the skin effect is more pronounced and proximity effect can be an even more severe problem. Litz wire is one kind of stranded wire, but, in this case, the reason for its use is not the usual one of avoiding complete wire breakage due to material fatigue.


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


My original idea still is valid... you can alter the motor behavior by changing the motor wiring. If you switch from a Single to a Double using the same amount of wire you get a motor that goes twice as fast for the same voltage and is able to draw more current. If the goal is to go faster with the same ebike you really can choose to increase the voltage (which increases the heat) or to change the motor wiring which gives many of the same effects of overvolting but doesn't produce a situation that has as much heat. (since the Double wiring has less resistance)

The limiting factor on this "low priced upgrade" is the fact that the stock battery might not be up to the task of handling the extra current. However, for people with small stock motors that would otherwise overheat with more voltage (or possibly they don't or can't upgrade the controller or battery) this might be a secret to getting some more bang out of a bike that otherwise was slow.

If your stock motor gets you to 15 mph, then going from Single to Double would get you to 30 mph.

.

safe

Rewiring the Unite 1200 Watt Motor

The reason I've taken an interest in this issue of "Turns" and "Winds" is that I'm in the process of rewiring a motor whose wiring shorted after several thousand miles of riding. The stock magnet wires are pretty low quality and I'm hoping that the stuff I've bought (200C wire) will hold up longer under the constant heat and abuse I give it. (my bike has hit speeds of 52 mph on the flat)

Anyway... the "stock" wiring was a 14 "Turn" Double...

I'm planning to try out a Triple and with my same current limit of 40 amps at 48 volts these are the calculated powerbands and motor rpms that I expect to find. I'll be able to confirm this when I actually complete the rewind and ride it. (when the weather finally warms up)

Single - 28 "Turns" - 1 "Wind" - 2000 rpm
Double - 14 "Turns" - 2 "Winds" - 4000 rpm
Triple - 9 "Turns" - 3 "Winds" - 6222 rpm
Quad - 7 "Turns" - 4 "Winds" - 8000 rpm

.

unime

If you want to change the "gearing" of a hub motor, rewing it with a different number of turns, or (for brushless) switch it from delta to wye or vice versa.

I'm sure some of the "pros" know what they are talking about. You are simply misinterpreting their advice, which really isn't specific enough to explain what they mean. I've given you the only explanation that makes sense, but you seem only interested in arguing with me rather than learning from someone far more knowledgeable on the subject.

Why would I need to read a wikipedia article on a subject I am already very familiar with? You are grasping at straws when you suggest there is any measurable skin effect in hub motors. It doesn't occur in 15,000 RPM RC motors which draw more current (running at lower voltage) than hub motors.

Look: The 100% skin depth frequency for 20 gauge copper is 27000 Hz, a bicycle wheel spins at 6 Hz. Even though the frequency of the signal driving the motor is a multiple of 6Hz, that multiple ain't anywhere close to 4500.

Got any more crackpot ideas?

Duh! Reducing the turns increases the no load speed (at a given voltage). No secret there. I don't know why you insist in confounding that fact with your misconceptions about single and double winds.

Wrong! Wrong! Wrong! Increasing the Kv constant (allowing the motor to spin faster at a given voltage) requires rewinding with fewer turns. Naturally you'll use thicker wire, resulting in less resistance, but the motor will now draw more current (at a given load and speed). Whether you do the calculation or test it experimentally, you will find the I^2*R losses are the same, and so is the heat generated.

When you wire a permanent magnet for speed, you sacrifice torque. It's a lot like changing gears. That's why you don't have modest hub motors wired with a 50mph no load speed. The practical effect of doing so would be to reduce the actual top speed *and* acceleration.

Whether you understand it or not, what you meant to say is "going from 2N turns to N will double your speed". But that's a simplistic view of motor performance, and does not reflect real world conditions an ebike copes with, such as rolling resistance, drag, hills, etc.

safe

So we agree then that if you want to increase the top end speed you can switch from Single to Double?

That was the core point...


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


Your analysis of current requirements doesn't take into account the way things often exist in the ebike motors. Very few ebikes seem to optimize for efficiency and instead tend to want to allow more current so that there is more torque.

For example:

The Unite 1200 watt motor is designed to run at about 48 volts and 40 amps. That tends to produce an inefficient powerband shape compared to the ideal. When you do like I'll be doing and switching from a Double to a Triple it will allow the powerband shape to get closer to the ideal while maintaining the same current limit.





What about hub motors?

Well... some are actually pretty close to being efficient right off the bat and modifying them might not make much sense. In the case of the Crystalyte hub motors they've done their best to present how to get them to run as well as they can. (they seem to be the ones that people use the most)

I'm a big advocate of geared ebikes that run the motor through the gearing and that allows me to do things like change gear ratios and stuff like that. However, even the hub motor might be able to use this rewiring option if the person is restricted somehow. The Optibike is generally known as the most sophisticated ebike in existence right now and it runs the motor through the gears.

Example:

Let's say you live somewhere that the laws say that the controller is more or less "fixed" at some level like 250 watts. If you replace the controller and get caught you face some legal issues. However, if you kept the stock controller (that in concert with the stock hub motor only gets you to 15 mph) you could rewire the motor to get you to 30 mph and there would be no visible sign that you were breaking the law. In fact, the power output would be the same only it would get spread out more.

The simple fact of all this... is STILL that rewiring really does work as a way to increase speed. You lose low end torque just because of the shift in the powerband alone (not even taking the Litz Effect into account) but you do end up with more top end speed and with the Litz Effect maybe it's just slightly better than before. Then on top of that you have the potential for better efficiency because you are in effect correcting the way many motors are designed to be used.

The core point remains... you can buy more batteries and overvolt to go faster or you can rewire your motor to make it go faster. The number of "Turns" defines the motor speed and the number of "Winds" defines the resistance and also (so the experts say) has some effect on the distribution of power in the powerband. I personally do not know if this "Litz Wire" will contribute a 5%, 10% or 15% difference in the shape of the powerband... but I'm curious enough to want to try it and see.

The nice thing about changing the number of "Winds" with the same gauge wire is that you can just buy one spool of wire and try it out. If a Triple turns out to be too much, then I can just rewire back down to the Double.

All the numbers point to the Triple allowing me to be able to use less heat in producing my peak power.

Heat = I^2 * R

Heat = ( 40 amps )^2 * 0.310 (Stock resistance) ---> 496 watts @ peak power

Heat = ( 40 amps )^2 * 0.103 (Triples resistance) ---> 165 watts @ peak power


Why does this happen?

Because the Stock motor configuration isn't the most efficient, so there is room for improvement. An "ideal" motor cannot be improved upon, but most motors do NOT run in their most optimal condition and tend to be favoring low end torque.



Re-Volt or Re-Wire?

That is the question...

safe

Summary of Options

Option One: Buy more battery and add to what you have so that you have more voltage. Assuming you keep the current limit the same the heat will only increase in the lower rpm range. However, this can be substantial because the motor resistance was never altered. This option will produce a higher top speed.

Option Two: One can rewire the motor using SMALLER strands of wire in parallel to produce Double, Triple or Quad windings and this will (because of the Litz wire effect) produce some small difference in the powerband shape. However, since no one seems certain of how to estimate this it could be 1% difference or 10% difference. We just don't know. This option doesn't increase your top speed.

Option Three: Keep the wires the same diameter and change the way the motor is wound. If the motor starts off as a Single you would go to a Double by reducing the number of "Turns" in half. This will produce a higher motor speed and since the resistance is lower this will produce power more efficiently. The "catch" is that you have to have been starting with a motor that was imperfect to begin with (current limit too high) so that this tip will work. A motor that is already optimal will not be improved. However in most cases these motors are run with current limits that are way too high for their speeds so the odds are in your favor that this will work better than stock. This option produces a higher top speed.


.

unime

[quote=safe;8405544]So we agree then that if you want to increase the top end speed you can switch from Single to Double?

You misunderstand the term "winds". In an N turn motor, each winding has N turns (a total of N*K for a K wind motor). Switching from single to double winds has almost no measurable effect on the motor's properties, assuming the total cross sectional area of the wire is maintained constant in the process.

Changing windings from serial to parallel has zero effect on the power performance of a motor. None. Nada. Zilch. It doubles the current and halves the voltage, but resistive losses are exactly the same because exactly the same amount of current flows through the sames sections of wire (and the same total length) either way. I don't know where you got your graphs, but I'm sorry to tell you they are wrong.

I'm aware of Optibike. You should ask them about my capabilities ;-) (I met with most of the gang last Thursday to discuss an accessory I am planning to offer.)

Sorry, not so. TANSTAAFL, as they say. You are delusional. Check the bicycle energy calculator and tell me exactly how you expect to get 30mph out of 250 watts, when it takes over 600 watts of power to propel at bike at that speed, even with 100% efficiency.

The rule of thumb is to run a permanent magnet motor at 80% of no load speed for best efficiency. If you re-wire for an unrealistically high no load speed, you will reduce efficiency and run slower at a given power input than before.

In your case, having overheated a motor, I'd suggest filling it with as much copper as you possibly can. Of course, you may have overheated the permanent magnets in the process, rendering the exercise futile.

unime

Uh... Hello! Have you been listening? Litz wire will not help. It is wasted expense. End of story.

safe

So we agree then that if you want to increase the top end speed you can switch from Single to Double?

Assuming that the strand size is the same and you are willing to see an increase in motor rpm. You do this by cutting the "Turns" in half to get everything to fit.

Now we agree... (this was the idea I was getting at)


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


Theoretical vs Actual


Another point to address is that most motors do NOT run in their optimal efficiency configuration.

Do you agree here?


So when you do this Single to Double switch (accepting the higher rpm, dropping the "Turn" count) you are shifting the powerband so that the motor is now operating closer to the "ideal" configuration.

Let's also agree on the "ideal":

"The ideal motor will be configured so that the peak power rpm and the peak efficiency rpm are the same based on the controller limit."


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


The moment you accept the "real world" of imperfection then you start to see the added benefits that rewiring can bring. In the "perfect world" where all motors are already perfectly matched then there are no benefits because you can't improve on the "ideal". Most "real world" bikes leave plenty of room for improvement. (it seems that most ebikes are set up so that they allow twice as much current as the "ideal" configuration would need)



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


As for the "Litz Wire" these guys are actual "winners" in real racing situations and they claim a slight advantage in some situations. I tend to favor the wisdom of those who actually "win" over someone who just claims to know.



Scott Giles was the big winner for Team Fantom this year, bringing home a National Championship in Stock Truck! Scott's victory was hard fought, as he just narrowly edged out his Fantom teammate Jay Beatty for the win. Jay ended up an impressive 2nd in the class, actually winning the first round of the A Main finals. CONGRATULATIONS Scott on your National Championship victory!

https://www.fantomracing.com/


...disputing the "Litz Wire" goes against the entire RC motor community (that seem to all use it as a concept) and also goes against Wikipedia. So if you are correct (and "Litz Wire" is an illusion) then you are going against a huge installed base of belief. Good luck.

.

unime

One more time, no. The number of windings has essentially zero effect on motor speed. Changing the number of turns affects Kv. You are abusing terminology by linking these two concepts. They are independent.

If you have a motor that is working OK, but want a little more speed, then cutting turns in half is a bad idea. Why? Because you won't go twice as fast - you would need to draw 8 times the power you used before (air drag is proportional to the cube of speed) but your motor won't be able to suck that much without increasing voltage. As a result the motor will be lugging, operating less efficiently than before.

As a rule, you want to make small changes to the number of turns.

"You keep using that word. I do not think it means what you think it means."

Go ahead and rewire your motor with half the turns and let us know how that works out for you. As they say on Slashdot, good luck with that.

I have already debunked the skin effect, which you suggested was relevant, as it applies to our motors. If you really knew what you were talking about, you would have explained the existence of losses due to eddy currents in the wire induced when flux crosses the coils. Litz wire is an effective solution to this phenomenon. The problem is it does not occur to any significant extent in ebike motors. There are two main reasons for this: First, the motors we are talking about are (assumed to be) iron core motors where practically all the magnetic flux is constrained within the core; and, second, ebike motors run slowly (compared to RC motors, for example).

You don't frequent the same motor design forums I do, because there is an active debate between the litz and no-litz voices on the ones I read. That's fine, but please don't go around claiming your opinion is the only one out there. Also, let me emphasize that this is a marginal effect on high speed RC motors. As you scale up the size and slow down the rpms, the effect diminishes to irrelevance.

Finally, look at the type of wire used in solar vehicle race winners. Some use coreless motors (with no iron losses) and litz wire, others use iron cores (for a more efficient magnetic circuit) and big fat wire.

safe

Let's not forget that not all ebikes use hub motors. Hub motors run at speeds like 400 rpm. The typical motor used on electric scooters and on some brands of ebikes (like the WalMart bike) tends to have a higher rpm (2500 or so) and usually involves gearing the motor down.

At least now.... finally.... you at least confess that you know of the debate... so that's "progress".

Maybe you take one side of it or another, but at least you agree that it's an issue.





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


Let me restate the "Single to Double" again....


The concept is to be able to use the same motor and get a different performance charactoristic. Assuming that you use the same gauge wire and then decide to switch from a Single wind to a Double you will in fact increase your no load speed to double what you started with.


You say:

"Oh my gosh... in order to maintain the same level of torque at low rpm I'm going to need 8 times the current."

Yeah... but you missed the point... the idea is to SACRIFICE the low rpm torque in favor of top speed. If you need to do a lot of hill climbing then you probably don't want to in effect "gear up" your motor with a switch from Single to Double. Just because the rewired motor "wants" more current doesn't mean you need to give it the extra current. (current limits mean the motor always is running below it's theoretical maximum)

The idea here in this thread is:

"What can I do to change the performance of my ebike? Does voltage make a difference?"

...and the answer is that "yes" changing the battery and increasing the voltage is one avenue to pursue, but changing the wiring of the motor is another. You have flexibility in your "options".



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


Like it or not you have in effect "agreed" with the fact that rewiring a motor can achieve the effects that are similiar to revolting the motor. They have the same basic effect either way.


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


Finally... you didn't really address the issue of "default configuration". Most ebikes are set up so that they are inefficient because they tend to grant a higher current limit than optimal. Rewiring from Single to Double will in many cases create an "ideal" efficiency powerband if you stick with the default controller. (but I will restate that SOME ebikes are already using their "ideal" configuration and so rewiring them would be a mistake)

Just try to address the "default high current limit issue". (and ways to correct it)

Or comment on the "peak power rpm and peak efficiency rpm need to be the same" for the definition of what is "ideal".


P.S: I wired up my first hand wound motor yesterday and was amazed that it worked right off the bat. Sure is hard on the fingers though... I went with the Triple for my Unite motor and it looks to be working out as planned.

.

safe

Results of the Triple Wind

Today was the first warm day in a while and I got to try out my Triple wind rebuild of an old motor I had laying around. The results were puzzling...

Rather than getting the significantly higher no load speed like I was expecting:

48 volt - Stock No Load = 4000 rpm - Speed in first gear - 27 mph

48 volt - Actual No Load = 4148 rpm - Speed in first gear - 28 mph

...trying it at 72 volts (I have a multiconfiguration battery setup):

48 volt - Stock No Load = 6000 rpm - Speed in first gear - 41 mph

48 volt - Actual No Load = 6222 rpm - Speed in first gear - 42 mph


...these correspond to a change from 28 "Turns" in total to 27 "Turns" when I went from the Double to the Triple.


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


The motor runs really, really cool compared to the old one.


It doesn't make any sense... when you go from a 28 "Turn" Double to a 27 "Turn" Triple it's supposed to radically increase the no load speed.

What I "thought" I would get is:

48 volt - Stock No Load = 4000 rpm - Speed in first gear - 27 mph

48 volt - Actual No Load = 6222 rpm - Speed in first gear - 42 mph

...and ignore 72 volts.



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


So the actual results favor my first guess as to how this might work...


Hmmmmm............... now I really am confused.....



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


PS: Without an Armature Current Limiting circuit this Triple SUCKS AMPS at low rpms so fast that the bike literally bumps against the low voltage cutoff because of the voltage sag. With the Armature Current Limit (set at 40 amps) the Triple is not such a scary beast. In fact the power delivery is very linear from bottom to top. So you probably are going to need an ACL circuit or this idea of Doubles and Triples is going to mess with a stock controller.

The thing about "parallel paths" of current and how that can effect the motors thirst for amps seems to apply... it definitely has low resistance and pulls a lot of amps if you let it. (which you don't have to do)


Very interesting...


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

Tomorrow I'll try some speed tests. My previous best was 52 mph on the flat and 55 mph downhill. That was at 72 volts and unrestricted as far as amps which is why I burned up my motor in the first place. When I ACL limited to 40 amps I was still able to deliver 50 mph on the flat. So if I can get close to 50 mph on the flat then I'm doing okay. For 48 volts my best was 48 mph.

.

safe

"Joseph Henry's first discovery was that the power of a magnet could be immensely strengthened by winding it with insulated wire. He was the first person to make a magnet that could lift thirty-five hundred pounds of weight. Joseph Henry showed the difference between "quantity" magnets composed of short lengths of wire connected in parallel and excited by a few large cells; and "intensity" magnets wound with a single long wire and excited by a battery composed of cells in series. This was an original discovery, greatly increasing both the immediate usefulness of the magnet and its possibilities for future experiments."

https://inventors.about.com/cs/invent...ctricity_2.htm

I think the root idea is that as long as the total copper is a constant, then the parallel design of the Double, Triple, etc.. passes more current with less voltage required. The no load speed seems not to be effected by what goes on at low rpm. The Triple SUCKS AMPS at low rpm much more than the Double and I would assume it would suck even less with the Single. By adding Armature Current Limiting I "filter out" the excessive low end current and keep the top end power with the parallel winds and it's low resistance.

It's hard to disagree with something you literally ride... the no load didn't go up from Double to Triple as long as the total copper was the same.

I should have "stuck to my guns" because my first "guess" seems to be what I actually experienced. (no change in the no load speed)

.

unime

Puzzling? Your results confirm exactly what I have been telling you.

As I have been explaining, Kv is inversely proportional to turns (and independent of winds). In other words, Kv*turns is a constant. Keep in mind that Kv is simply a constant that expresses the no load rpms per volt of a given motor. Have a look at your data:

28 turns:
48 volt - Stock No Load = 4000 rpm
Kv = 4000/48 = 83.3
Kv * turns = (4000/48)*28 = 2333

72 volt - Stock No Load = 6000 rpm
Kv = 6000/72 = 83.3
Kv * turns = (6000/72)*28 = 2333

27 turns:
48 volt - Actual No Load = 4148 rpm
Kv = 4148/48 = 86.4
Kv * turns = (4148/48)*27 = 2333

72 volt - Actual No Load = 6222 rpm
Kv = 6222/72 = 86.4
Kv * turns = (6222/72)*27 = 2333

The Kv value for your stock 28 turn motor is 83.3. It is 86.4 for the 27 turn motor. As expected, Kv*turns is a constant (2333) in all cases. As I have been telling you over, and over, and over, and over, and over, and over again, the number of windings has no effect on Kv.

Now, do you understand?

It makes complete sense. I'm happy to explain what's going on if you still don't get it. Just ask a specific question so I know where to focus my answer.

safe

My First Guess Was Right After All

The humor of all this is that I had figured that constant copper means constant speed no matter how you wire it internally. (read back to my first ideas about parallel winds)

The net effect is that with more parallelism you get a lower motor resistance at higher rpms. This allows for a more efficient running motor at higher rpms and less heat. That was my goal on my project.

The downside for "normal people" without knowledge of fancy things like Armature Current Limiting circuits is that the low end rpm current demands go up significantly. So for a given battery the low end is LESS torquey than the lower numbered winds.

Better top end.

Worse bottom end.


Ultimately it's a trade from the bottom rpms to the top... sort of like a race car motor... lot's of top end power, but nothing down low...


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


Why so many theories?

One wonders why there are so many different theories:

The Parallel Paths Theory
The Inductance Change Theory
The Litz Wire Theory

...I suspect that (like with most things) they all are perspectives on the same thing. The effect is real, the no load speed remains the same, the motor runs cooler with more winds, but with less torque down low, all the effects are real... but the causes are complex enough to get mulutple theories.

The easiest theory is just to say that parallel allows more current... the Inductance and Litz theories sort of spin off from that...

But that's an ideological argument that is endless... at least I know now (by experience) how rewiring a motor "feels" to ride...


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


Rewire or Revolt?


So what if you start with some WalMart ebike with a ridiculously high current limit for the motor. The stock efficiency is bad because they wanted to allow more low end torque to give people more of a surge at low speeds. (this sells the product even though it's a bad direction technically) By rewiring the motor to a Double or Triple you improve the efficiency of the top end (around 15 mph) but without the addition of somethng like an Armature Current Limiting circuit you will experience voltage sag at low rpms when the motor sucks more amps than the battery can commit to. So it would be a two step process... rewire and then add an ACL circuit. Otherwise the Double or Triple would not work well at low rpm and would likely trigger the low voltage cutoff. (it makes the bike jerky)

Still a neat idea for those technical enough to do it. (so maybe 0.1% of the ebike population)

Plus, with this newly rewired motor with it's higher current capacity the owner could both go up in the voltage AND go up in the current at the same time. The net effect would be more speed and more power. The downside (still) is that you need to protect your low end from too much current. If you do not implement an Armature Current Limiting circuit you increase your chances of overheating by exponential levels. (very dangerous)

Armature Current Limiting is a simple idea:

Place a Hall Effects current sensor on the wires going into the motor. Read the voltage of the sensor and compare it to some baseline voltage that matches the current sensors output for your desired current limit. Use a comparator to "pull down" the throttle voltage as needed to limit the current that is going into the motor. They are pretty easy to build.

.

unime

You still don't get it! "Parallelism" is not the issue. What matters is the number of turns and the cross sectional thickness of the wire. It pretty much does not matter whether you wind with one strand or one thousand.

Wrong, wrong, wrong, wrong. You will have a lot more torque because more current flow creates more torque. You will also have better efficiency at the original full throttle torque (though efficiency gains may be negligible).

Increasing the no load speed does not necessarily improve top end!!! To optimize your motor for high speed performance, first measure your actual top speed rpms and compare the value to the motor's no-load top speed. If the actual is more than (approx) 80% of no-load, you should decrease windings. If actual is less than 80%, increase windings.

But that approach may be shortsighted. By optimizing performance at top speed, you sacrifice low speed efficiency. A hub motor bike is likely better off with more windings than you would choose for top speed.

Increasing the wire's cross sectional area (by using thicker wire or more windings) will produce worse low speed efficiency at full throttle and may be more prone to overheat at low speed, but doing so improves efficiency at any given speed and power.

Again, you are missing the big picture. If you re-wind the motor for better high speed efficiency, you will reduce efficiency at the low end. The simply truth about hub motors is they cannot have good efficiency over the wide range of speeds they cover.

There is a good chance that whoever built that WalMart piece of junk, actually know what there were doing when they designed the motor. They certainly know more about motors than you.

safe

But hub motors are kind of stupid to begin with because they only have one gear.

(but that's a whole other discussion)