One hour CD ROM motor construction PART2 - Page 5 - RC Groups
Shop our Airplanes Products Drone Products Sales
Thread Tools
Dec 07, 2003, 04:00 PM
Registered User
galloping gimp's Avatar
I replaced the stock rubber ring magnet in my CD-ROM brushless with twelve 5x5x1 N45 magnets. What an improvement!

GWS 9x7 prop (I know -- too much prop!)
370 mAH NiMH battery pack
Jeti 08-3P controller

With ring magnet
3.5 ounces thrust @ 4.5A

With N45 magnets
4.9 ounces thrust @ 4.3A

Despite producing more thrust, the motor no longer gets too hot to touch. It's warm/hot, but not uncomfortably so.

I'll be interested to see what happens with, say, a GWS 8x4.3.

Thanks to everyone for their advice. The collective knowledge on this forum is remarkable.

- Jeff
Sign up now
to remove ads between posts
Dec 07, 2003, 10:43 PM
Registered User

Re: Re: Grinding N45s?

Originally posted by Takao Shimizu

I use 5x5x0.9t N45.

Takao.. or ALL where do we get the 5x5xx.9? I can get 5x5x1 from EngConcepts,.. George has a great place is a great seller, but didn't see the x.9

Dec 08, 2003, 02:23 AM
Grand Poobah of Nothing
Trizza's Avatar
Awesome find lads

I found an EASY and CHEAP source of 12 pole stators. Should be perfect for around 40-50mm diameter true LRK motors. The stacks are very high also, so you could make a couple of shorter motors from the one.

My first one was pulled out of an electric hand mixer - the type with the two beaters that are plugged into the body. Its a 240V 50Hz AC motor, but the stator is quite long and has 12 poles.

I also have a stack of old sewing machine motors, the stators in these are also 12 pole, but of a larger size so they'd be great for larger LRKs than my lathe could deal with. One of these motors is going to be its new power source as an upgrade to the normal 12V 4A unit.

There we have it! Small AC appliances may well have 12 pole stators in them. I'm going to go strip that hand mixer motor now..
Dec 08, 2003, 02:55 AM
Registered User

Re: Re: Re: Grinding N45s?

Originally posted by Trprhook
Takao.. or ALL where do we get the 5x5xx.9? I can get 5x5x1 from EngConcepts,.. George has a great place is a great seller, but didn't see the x.9


Dec 10, 2003, 12:23 AM
Registered User
I just got 2 computer fans and they are 4 pole. They have a very strong stock ring magnet. I was wondering if anyone has tried using these fans for our applications ?

Dec 10, 2003, 02:42 AM
Lipoly not Lipo-suction
Cranky1's Avatar
Hi AirBorne_AZ, I've seen posts before about other fans. I'm fairly certain they have to have poles in multiples of 3. (3, 6, 9 etc)

I want to try to make a larger outrunner from bits like Trizza has mentioned. :-))

Dec 10, 2003, 05:31 AM
homo ludens modellisticus
Ron van Sommeren's Avatar
Alas, number of poles is not a multiple of 3.

Met vriendelijke groet Ron van Sommeren
• diy outrunner brushless e-motor discussion.
• Electric fly-in & lrk meet, June 27th, 2004, Nijmegen, the Netherlands.
Last edited by Ron van Sommeren; Dec 10, 2003 at 05:33 AM.
Dec 12, 2003, 07:32 AM
Registered User

CD-ROM motor RC Prop. torque meter

I hope that my torque meter has good accuracy.
The calculation(Torque x RPM) says that my CD-ROM motor's efficiency is 75%.
Last edited by Takao Shimizu; Dec 13, 2003 at 02:33 AM.
Dec 14, 2003, 04:24 AM
Why not Delta?
rysium's Avatar

Brushless mystery uncovered

Hi Guys

The mystery is over. I have answers for the following questions and some others too.

Originally posted by galloping gimp
Question for all: When replacing the stock ring magnet with more powerful N45 magnets, what are the effects on RPM, torque, and current drain, assuming batteries and prop remain the same?
The Kv value, ie the rpm/V depends on the number of turns. The more the turns, the lower the Kv.
On a given turn number, the thicker wire is the better, thicker wire has lower resistance, lower resistance means lower copper loos, higher efficiency.
The conclusions given below are the results if some theoretical research and they are confirmed with some tests.

Let's just jump into conclusions first:

1. The more turns on the motor the lower rpm/V - everybody knows that, but did you ever think why?
2. The thicker wire, the lower resistance means the lower copper loss higher efficiency? It's true but in most cases it's not significant. Instead of using thicker wire, better use multiple insulated thin wires parallel. Explanation will come.
3. The stronger magnets the lower rpm/V - did you know that? In some cases the stronger magnets will not increase the motor performance for some applications. But the efficiency of the motor will be higher for the same power generated.
4. And the last one: You can't increase the motor efficiency much unless change the entire design.

If you want the long explanation please read the next message, if not, just skip it, and I have some more info to follow.

Dec 14, 2003, 04:25 AM
Why not Delta?
rysium's Avatar

Brushless mystery uncovered - part two

Thy Brushless motor theory

The most important thing to understand is, how the motor rpm is controller by ESC. ESC creates the rotated magnetic field in the motor's stator keeping it in front of magnetic fields of the rotor with the constant (or optimal) angle (motor timing). So the speed of the magnetic field is constantly adjusted by ESC to motor load. When throttle is applied only the voltage of the impulses applied to the motor is changing. When engine is running without the load (just the load from the bearings) ESC will generate higher and higher frequency impulses that will turn high rpm, when you load the motor the ESC will reduce the frequency that will result in drop of the motor's rpm. Just keep it in mind, that even the motor acts like brushed one (reduce rpm when load is applied) the speed of the motor is constantly adjusted by ESC. The speed of the motor is always the same as the frequency of the pulses generated by ESC. It's called "synchronous motor".

Thy Brushless motor efficiency

The second important thing to understand is how the energy balances in the motor that results the motor's efficiency. Input power (the applied power) = voltage times current. The output power = motor load (propeller, gearbox and bearings). The loss = Input power minus output power. It's simple. So what is the loss?

1 -The loss in the copper. Simple formula: Power = Current square times resistance. Is not as big as you may think. Giving simple example of my CD-ROM conversion the copper resistance is about 50 mohm and current draw is 5A. The power loss in the copper is 5A square times 0.05 = 1.25W. The power drawn by motor is 5A times 7V = 35W, so the copper loss is just 3.5%. So if you use the thicker wire let's say 2 times, you will reduce the copper loss 4 times so you save 2.6% - it's not much.

2. The loss in the metal stator. There ate two types of losses.
- Hysteresis losses. Iron is a ferromagnetic substance with the magnetic hysteresis. When the external alternating magnetic field is applied the magnetic field in the iron will not follow it converting some of the energy into heat. The higher the frequency of the external magnetic field, the more energy is converted to heat. At the frequency we are using in RC motors (about 3-5kHz) this is very significant amount of energy. You can calculate it exactly, but I'm sure it's much more than the loss in the copper wire.
- Eddy current losses. Iron is an electric conductor. When alternating magnetic field is applied to the conductor it will create the alternating current. Of course most of the energy is converted in the iron into the heat. Again wt the frequency we are talking about it is a significant amount of energy.

3. The same rules are applied into rotor. The magnetic field from the stator interacts with the metal rotor (magnets, bell) producing more energy loss.
4. And one more concern. Because of the Eddy current effect at the high frequency the copper is used only at the surface. Increasing the wire diameter will not result the lowering the resistance only linear. It's better to use number of thin wires, so the copper surface is bigger and wire is used more efficiently. And because the wire is round, you can fit more copper in the tight space when you use thin wire instead thick one..

Number 2 and 3 are somewhere around 20-30% of total input energy. The prove? When the motor runs without the load it still takes around 1A at high speed. The mechanical loss (bearings) is insignificant one. The copper loss is around 50mW, but motor takes 7W (at 7V). After a while touch the rotor - it's hot, touch the stator (not the wire) - it's hot. The Hysteresis loses and Eddy current loses are proportional to the current (magnetic field strength) and frequency (the speed of magnetic field changes). That's why brushless motor efficiency is low at high rpm (high frequency) and high load (high current).

Thy Brushless motor speed

Why the less turn, the more rpm? Again we have to understand, that the motor's speed is the same as the frequency of current generated by ESC. The motor's rpm is always the same as the speed of rotating magnetic field followed by the magnetic field of the rotor. So the question should be: Why the ESC is giving higher frequency on low turn motors.
The speed controller is adjusting it's frequency according to the motor load. It’s the same rule as the energy balance. Power in = Power out. We know, that power out = motor load and energy losses. Load and Energy losses are proportional to rpm (somehow). So let's calculate the input power.
Input power = Input voltage time Input current. This is the simplified rule for DC. For AC (it's an alternating power motor) the rule is: Power = Voltage times current times sinus(alpha). Alpha is the angle between voltage and current. Let's forget about it now, while the ESC is trying to keep is constant (for optimal motor run).
The Input voltage is set by the throttle position. The current is the result of Voltage and motor's impedance (with motor's very low resistance it's mostly the reactance). Reactance is the kind of resistance given by coil to the alternating current. It's proportional to the current frequency and coil's inductance. Here we go. The most two important factors are limiting our current: frequency and inductance. Inductance is proportional to the number of turns - I mean more turns, higher inductance. It's also proportional to the rotor's interaction - that means more load, lower the inductance is.
So the motor's speed is the result of perfect balance of impedance, frequency and voltage.
If motor has less turns, the result impedance will be lower, so the motor will draw more current, that will give more power to balance the loss and the load, and ESC will adjust the rpm for energy balance.

How the magnet strength affect the speed. To understand the impact we need to go back to the impedance. Part of the stator's impedance is the reaction from the rotor. When rotor moves against the stator it creates the opposite voltage in the stator coil (like the generator). For the coil it's like the input voltage from ESC applied is smaller by the voltage generated from the moving rotor magnets. The stronger magnets in the rotor the more voltage is generated. The faster rotor's speed is (the magnetic field is changing) the more voltage generated. The more turns on the stator, the higher voltage generated. This voltage generates result's like the increasing the stator's impedance. So ta-daaaa - the stronger magnets in the rotor, the lower rpm/V. To get the same output power with the same proppeler (the same RPM) we need higher voltage applied, but then the motor will draw less current (Remember? Power = Voltage times Current). Less current, means less losses in the copper. But loses in the copper are insignificant so it's not the way to go. Better use larger propeller to get the same load at lower rpm. In this case the frequency will be lower so the hysteresis losses and Eddy current loses will be lower ego higher efficiency.

Real test examples in the next post.

Dec 14, 2003, 04:25 AM
Why not Delta?
rysium's Avatar

Brushless mystery uncovered - part three

My tests

I have to almost identical motors (CD-ROM conversion).

1. 18 turns 2 x 30 G wire, Standard magnets in the rotor (no coging at all)
2. 20 turns 26 G, 5x5x1 magnets in the rotor (strong coging)

First motor measured 42000 rpm at 7V = 6000 rpm/V
Second motor measured 21000 rpm at 7V = 3000 rpm/V

The only significant difference is the magnet caused lower rpm/V

At 7V first motor turning GWS 8x4.3 draws 5A on the gearbox C (5.33:1)
At the same voltage the second motor draws the same current turning GWS 12x6 on the gearbox B (4.43:1)

You can tell the torque is much stronger on the second motor - it's like additional gearbox. The efficiency is higher because the frequency is twice lower, so the energy loss in the iron is smaller too.

If you still have a questions, just ask

Dec 14, 2003, 04:27 AM
Why not Delta?
rysium's Avatar

My new CD-Motor

Now here some pictures of the new motor

First internal parts

Dec 14, 2003, 04:28 AM
Why not Delta?
rysium's Avatar

My new motor

Mounted on the gearbox:

Dec 14, 2003, 04:31 AM
Why not Delta?
rysium's Avatar

And two motors I tested

Front view

Dec 14, 2003, 04:32 AM
Why not Delta?
rysium's Avatar

And two motors I tested

Back view


Now I'm going to bad. Good night

Thread Tools