My Asymmetric Electrodynamic Machines

[JohnStone]

... So, I took it apart and I have 2-4 coils that are burnt....

OHHH! Sorry for that!
@ALL:

Your circuit needs to experience some kind of wear-in in order to find faulty conditions - but nobody does it.

I still strongly recommend:
- test with H4 bulb (OK you did it )
- test with motor but connect H4 bulbs in series to motor. Light signals show if all motor windings are ok.
- check with scope for overvoltage / back EMF (start with short pulses)
- take bulbs out and check with scope again like above - starting with short pulses

JS

[JohnStone]

...In the past there has been much talk about which Arduino and now even another choice from outside Arduino that also looks good. We have a saying that goes(You don't always get more for paying more). .....

Agree strongly!
I feel it is time to explain what we are going to do with micros:
1.
Please remember your first shirt you got as a child! Was it the size you wear today? Did your parents say: We buy a shirt for this lad once! Was this their procedure to save money?
Arduino is our first shirt here and it fits perfectly. Of course we stick our nose to the shop window (that's ok) and get big eyes watching those bigger "shirts" - but they do not fit to us just now.
2.
Like at PC the problem sits sometimes in front of the screen. Arduino is a small micro and can be clogged instantly with some few commands. Same at those bigger ones. Imagine the speed difference between Windoze and Linux.
We need to program smart code in order to speed it up. This is no criticism regarding current programmers but the good news that we can get much more out of Arduino. You programmers did a good job and we can get a decent governor later on.

3.
Interrupt! Yes that is the way!
For newbies:
a) Imagine you have a repair shop for cars and you refuse to talk to any new customer if you just do repair work.Customers need to wait until you have this car ready. -> This example stands for a micro without any interrupt.

b) In real world you will stop your repair on request for talking to new customers, agree for a date for their car repair and then you will continue the current task -> This example stands for a micro with interrupt.

c) Now imagine you have 20 customer interrupts per hour day by day! You will get serious delay and angry customers. -> This example stands for a micro with too many interrupts. And do not believe plenty of interrupt sources are a great help!

Back to Arduino:
There are only two occasions were we need an external interrupt
- Speed sensor
- amp limit sensor for emergency shut off (but there is a better way)

There is only one internal (timer) interrupt necessary in order to:
- generate pulses

Dealing with these 3 interrupts makes life even more complicated. You need to nest and interlieve code and check if one interrupt can interrupt the current one and the next one ...... It is quite easy to get a dead lock causing major damage to the setup or operator.
Ardunio is complex and any bigger micro is even more complex. In order to deal with this complexity we need to introduce an operation system .......... So please calm down and enjoy those easy times dealing with an Arduino.

4.
Do not believe that employing additional Arduinos will solve any performance problem. Any serial interface consumes computing power for data buffers, interpreting serial commands and for sheduling tasks and causes delays and we need to account for them as well.
Multiple micros need to be designed in a sophisticated way.

5.
In order to understand the computing for our setup we need to find some smart segmentation:

a) Input data: might be time critical or not.
- Speed sensor is time critical -> use of external interrupt.
- High amp emergency is time critical but Arduino might fail to react intime -> hence HW solution for shut down recommended. Any signal to Arduino in emergency case ist then NOT time critical.
- Temparture, voltage, average amp is not time critical

b) Output data: might be time critical or not.
- pulse control is time critical and shall be controlled by timer interrupts

c) Human Machine Interface (HMI) usually not time critical in our case
Especially LCDs will require computer power for constant refresh of pixels. It shall be omitted or kept to minimum.
Indicator lights, switches, pots are not time critical

d) Control program:
- There are time critical parts like pulse generation.
- Much more less time critical is the governer function because of the inertia of motor/generator setup. Nevertheless a full featured PID function consumes a lot of formula computation.
- Any HMI computing or basic serial I/F will not be time critical.

Please understand that the notions above are only a short overview on what we plan to have.

Recommandation:
Let's focus an pulse genertion and simple PI speed control first. This will be a considerable help in order to get knowledge for our setups. It is essential to get reliable data and compare achievements. It will be of no help if we discuss here problems of micros instead of useful modifications for the setups.

Some of you prefer to have a LCD with pleny of figures in order to get suffitient information. Me too but this request will kill the simplicity.

Let's decide to have 2 simple and separated Ardunio setups:

Control: with two operation modes:
1. Simple control for frequency and duty cycle (like current program)
2. Adding, rpm sensor, a mode switch for rpm control. In this mode the control for duty cycle will switch to be speed preset.

Data aquisition:
All sorts of measurements like amp, voltage torque, rpm (rpm sensor might be fed to both Arduninos) Those data is not time critical at all and mostly not related directly to genuine control function.
This Ardunino might be able to display all sorts of date via a LCD but there is a better soluition available

Dashboard:
Adding a serial I/F to Arduinos is not time critical as long it is used for raw data interchange with PC. I imagine a repetitive and non time critical frame of raw data for i.e rpm, pulse duty and frequency, amp, voltage, torque .... Any further calcualations and dashboard display can be done at PC. Do not worry for PC programming - not necessary. There are dedicated programs (like ProfilabExpert) awailable where dashboading is made very easy and intuitive along gauges, scope views, calculations and storing of data in CSV sheets (fo i.E. Excel)...... Most changes can be done instantly.

The concept presented above might seem complex at fist glance but that is not true. It segmentates the task in 3 parts. Each part is quite simple for itself and can be tested separately. The parts can be performed by different persons at differents locations and time.
I myself would like to contribute dashboards freeing all Ardunino programmers from calculating decimal values displaying them on a small LCD.

Please tip your chair back and focus on simple control for now. It will grow to a genuine part of the whole story.
js

[machinealive]

@prochiro

I assume you will be rewinding the old rotor and not waiting for some new ones

.
You are funny, funny, funny. I've already have it stripped. I'll be heading to my local motor shop tomorrow. I'll wind this one reeel purdy, mmm Yea.

[prochiro]

@ JohnStone
I am having difficulty understanding parts of your post where you use several abbreviations such as "simple PI speed control first". We are not as advanced as you are and between trying to understand the just of what you are saying and guess at the rest, I am still in the dark. I looked thru the whole post and see no other wording to lead me to know what that is.
JohnStone, I have run all the light tests, I have perfect control of the Imperial with no load, After putting just a low load on the motor I realized that loading without instant compensation at the motor load on and load off will not give any real useful information and will hurt the motor as in Machines case. The monster circuits are bullet proof once checked out. Running the motor above 3800 will heat the motor and fets. Running at 3500 the generator is only able to give a few hundred watts before loading motor real bad. This is where I am at at this time. I can not get to the pots fast enough to operate higher duty to keep things stable enough for any real testing. A real standstill. The program I have been working on is 100% governor only. I hope the term (PI) means that. If not, I really don't know where you are at. As far as the number of interrupts, I do know that one should be doable for RPM sensing and will get it to one, but as of yet, the three programs I have to learn from are using two and fudging a third. I do not have the sensors to test with yet so all I am going on is there code. I also have mentioned several times that the LCD is not on my system sinse the single push button project.
Please tip your chair back and focus on simple control for now. Again, we have simple control, on, ramp up ramp down and stop. In my opinion it is not safe to do much more without a governor to protect from motor runaway and load drop without hurting things. "You programmers did a good job and we can get a decent governor later on. "You have kept us on base thus far but we need a functioning governor in the motor shop now.
Dana

[iankoglin]

@ JohnStone
I am having difficulty understanding parts of your post where you use several abbreviations such as "simple PI speed control first". We are not as advanced as you are and between trying to understand the just of what you are saying and guess at the rest, I am still in the dark. I looked thru the whole post and see no other wording to lead me to know what that is.
JohnStone, I have run all the light tests, I have perfect control of the Imperial with no load, After putting just a low load on the motor I realized that loading without instant compensation at the motor load on and load off will not give any real useful information and will hurt the motor as in Machines case. The monster circuits are bullet proof once checked out. Running the motor above 3800 will heat the motor and fets. Running at 3500 the generator is only able to give a few hundred watts before loading motor real bad. This is where I am at at this time. I can not get to the pots fast enough to operate higher duty to keep things stable enough for any real testing. A real standstill. The program I have been working on is 100% governor only. I hope the term (PI) means that. If not, I really don't know where you are at. As far as the number of interrupts, I do know that one should be doable for RPM sensing and will get it to one, but as of yet, the three programs I have to learn from are using two and fudging a third. I do not have the sensors to test with yet so all I am going on is there code. I also have mentioned several times that the LCD is not on my system sinse the single push button project.
Please tip your chair back and focus on simple control for now. Again, we have simple control, on, ramp up ramp down and stop. In my opinion it is not safe to do much more without a governor to protect from motor runaway and load drop without hurting things. "You programmers did a good job and we can get a decent governor later on. "You have kept us on base thus far but we need a functioning governor in the motor shop now.
Dana

G'day Dana
It is some time since I did any programming mainly assembler for a 16bit processor 1974 model, I am following along with all the posts and at this stage am really not qualified to make any input to wards the programming of the Arduino but programming is programming no matter what the language is.
What JS is saying is that we need at this moment to concentrate on gathering Empirical information that can be used to make decisions when certain parameters are reached while the motor is running we need this information to know what/how we need to adjust the controls pertaining to the machine. e.g. write the Arduino program so that it modifies itself in response to the information that has been collected. I am sure you know this and am not trying to teach you in any way.

John did elaborate on this in Post 4742 Page 159 Particularly the PID
I am sure he was just implying that we all need to concentrate on one thing at a time

Dana You yourself just mentioned above that you are/ have been" working on is 100% governor only program" and this is what he is talking about when the speed governor is perfected then move on to the next segment so keep up the good work we all appreciate what you are doing

I am not in any way showing disrespect to you or anyone else as I have noticed that you along with others have progressed in the programming of the Arduino and I feel that if I do not start with the building of the P56 motors and controller as you and others are, I will be left far behind
Please keep up the good work

Kindest Regards


Trying to keep up

[JohnStone]

@ JohnStone
I am having difficulty understanding parts of your post ... ...

Sorry for confusion!

• The post was addressed mainly to those looking for high end micros and governors not knowing the additional implications they buy with them.
• The second message is that starting with simple programs and Arduino will take us very far approaching our goal perfectly.
• This simple start is no loss of effort but will fit perfectly to further elicitation if required.
• Those notions of mine were addressed to a wide area of autitorium - sorry again for confusion.

PID: Yes look at page 159 in this thread. But I want to explain in more detail and related to your daily experience..

P = proportional factor
Imagine you driving a car. You see a road hole from far and you adjust your steering wheel slightly in order to omit that hole. In this case you react with low amplification = low proportional factor.
Imagine another case where a deer jumps in front of your car.Then you might want to turn the wheel as fast as possible in order to react to this sudden obstacle. In this case you react with high amplification = high proportional factor.
If you react to the deer in the gentle "hole" mode you will be wrong and if you react to the hole in the aggressive "deer" mode you will be wrong. We can not have both reactions perfect with P reactions only. Too low reaction might prevent omitting an obstacle and too high reaction might get the system to wild and dagerous oscillations. We need to live with some flaws just now.
Regarding our setup we decide to ignore deers and accept only low P-factor = low amplified reation.

BTW: A jumbo jet reacts 90 seconds after any steering action - not before. Imagine how low and jentle rudder amplification we requrie!

I = integrating factor
Back to the example above: You adjust your steering wheel gently in order to omit that road hole and after a while you detect that you will hit that hole anyway. You intuitively readjust the wheel and the closer you approach to that road hole the more precise you will adjust the wheel - and safely omit that damn hole in the end.
An I-factor will hit in the end the exact value but it takes some time in order to adjust. It sums up all consecutive errors and thus cancels any governor error to ZERO! Understand it to get exactly 50/60Hz but it will deviate sometimes at load change.

PI = proportional / integrative
Your notion is now that you have a perfect PI governor built in and it is quite perfect for reacting to events being known in advance and as time passes the precision will increase.

PI-implemetation
Implementing PI in our setup we decide to not expect perfect automatic adjustment at sudden load changes. (those sudden deers are soft-deers and do no harm) The system might rotate for short time too fast or too slow but after a time it will adjust itself precisely to the rpm requested (omitting road holes). And then we can take our measurements.
Writing code you will get the current rpm value and check for rpm difference to the value being preset (i.e. 3600 RPM or perhaps 0.8 ms from your sensor). Depending on the value calculated you will increase or decrease the duty value.
You can play with repeating time of the action obove and / or with the increment-/decrement-value you set to the duty value. Summed up it decides for the P and I factor.
You will start with smooth and slow reaction (i.e. one rpm check every second and if necessary increment / decrement one singel count). Later on you might want to decrease time and make the governor step by step more aggressive. In the end the system needs to react smoothly at sudden change of max. load to no load and vice versa. Then you have the best tuning for simple PI behaviour. That is what we expect to be happy.
(In fact cruise control built in in cars is a PI governor -> slow, and integrative but no real dynamic behaviour (deers!)

D = differntial (just at thought food - not for implementation)
Back to our deers: You know that fast and destinct reaction is necessary while the road hole will require smooth reaction. In fact you estimate how many time is left in order to make use of your reaction. The issue is the time left along the speed of your car.
In other words: D-component accounts for dynamic behaviour. It accounts for the severity of external impact and tries to minimize system effects. D- component contributes an estimated severity preemption to the behaviour.

PID = proportional / integrative / differential
The complete classic math for such a governor indicates that an optimized behaviour will still allow for 4.3% overshot/undershot of rpm at sudden load changes. Trying less overshot will make the system instable, trying less undershot will make the system slower.
Just for information: there are other approaches available but far out of our requriements and BTW out of my comprehension.
----------------
This teaching shall give you an idea on the playground we play on and what restricted area we choose to perform our sports training. And this restriction will suffice! You are invited to ask further questions because it is essential to know what we do with our setup exactly.
JS

[machinealive]

I put the first coil on the rotor, took a while to decide what size wire. Dana, UFO, I need to know what your opinions are, they may have changed since the imperial project began, now that amps are coming down.

So I decided to go for 19.5 ga, and put 20 turns per coil. The first coil is .5-.6 ohms.

By the way, I only used 17.5 ga last time, that's why it's ohms were so low. I thought it was 18.5.

Anyway, since the brushes are for the most part contacting 2 coils in parallel, now .25-.3 ohms, maybe I should go with 20, or even 20.5 ga.

I now rpms dropped before, with coils in series, but it doesn't really seem the same, maybe it is?

I want to know what you guys think, maybe 21 ga, try to get each coil up to 1.4-1.5 ohm.

Machine

[Ufopolitics]

Hello Machine

I put the first coil on the rotor, took a while to decide what size wire. Dana, UFO, I need to know what your opinions are, they may have changed since the imperial project began, now that amps are coming down.

So I decided to go for 19.5 ga, and put 20 turns per coil. The first coil is .5-.6 ohms.

By the way, I only used 17.5 ga last time, that's why it's ohms were so low. I thought it was 18.5.

Anyway, since the brushes are for the most part contacting 2 coils in parallel, now .25-.3 ohms, maybe I should go with 20, or even 20.5 ga.

I now rpms dropped before, with coils in series, but it doesn't really seem the same, maybe it is?

I want to know what you guys think, maybe 21 ga, try to get each coil up to 1.4-1.5 ohm.

Machine

Hello Friend,

There are some things I would like to refresh to this new pulsing System...as to our Machines in general.

With this Quad Channel System Your Motor will be receiving only ONE Pulse at a time through the "timed/In Turn" Gate...and actually what rotates are the sequenced order pulsing Gates from P1 to P22. This means that only ONE Pair/Coils will be On at a time...never two Pairs like before.

This fact allows you to use a minimum of resistance per Pair/Coils...but now let's go to the Magnetic Field relation with the wire gauge...

A Heavier Gauge will produce a stronger Magnetic Field...but it would deliver a lower resistance...when compared to a finer gauge...

We need a strong magnetic field to rotate strongly that rotor mass...as to push/pull a stiff Generator Head Rotary Stator...

This is where "our Game" resumes and stays within...

So, again...I would not go too low on Resistance as below 0.6...(as this too low R Values will generate heat in the long run...and could end burning some coils at heavier loads/slower forced RPM's) and would try not to go above 1.0 Ohms per Pair of course. (Too High values of R will diminish the Magnetic Field strength, so reaching higher RPM's demand under heavier loads would be difficult) If we use here an "average in between" I will choose 0.8 or close to...as an ideal number...but then again, life is not that simple nor perfect...

Hope I helped you to take your best choice here...let me know.


Warm regards


Ufopolitics

[Ufopolitics]

Hey Dana...

@ Machine
I am sorry for the rewind but as you said, you are good at that. We are going nowhere so just take your time and do an extra special job. I assume you will be rewinding the old rotor and not waiting for some new ones for now. This new rotor will never see an RPM past 3800 and do the work of 9000 if necessary. As I said earlier, I am working on the program but it will be a short while I hope so just get er done.
Dana

Hello Dana,

Dana, there are some things I am concerned about, related to Your Set Up...

1-Mainly, I am worried about your old generator performance and its capacity to generate strong outputs...Generators loose over time their reminiscent magnetism levels. (meaning that not necessarily loose it all, but lower its strength ..to weaker values...this will excite generator fields but NOT to the original spec's when it was new...therefore the MAX (Close to Surge Out) will not be top label marked.

Of course this could only be "truly" checked/verified by comparing/rotating your generator head with a Gas engine as required from factory HP Spec's, then adding loads close to Surge values...I do not know if you could do that Test at this point.

But resuming here...if your Gen Head would not be generating the full output at the right RPM's/Torque...your tests between the two electrical machines will not produce any positive results...no matter how well tuned or coded your set up is on the Prime mover side.

2- In my opinion, Dana, construction of an almost perfect Electronic-Coded Governor...will be an awesome and very needed option we all will require to have, in order to guarantee a safe and stress-free operation between machines...However, If step One...testing both machines and manually governing speed/duty cycle...as adding SLOWLY, ONE BY ONE loads...is not "positive" then going to step Two..."governor"...I see it like out of time, or rushing it without any previous good results.

I do not recall what was your Generator full output?...but whatever it is...You do not need to take it to FULL MAX OUT...since you have so low amp draw at Imperial Motor to obtain a successful testing...


3- Another Issue am concerned related to playing with dead/off times...

Since You have so low amps draw...I feel You could play by "approaching" square wave channels or creating a small common shared time between all four...an area of Milli Seconds of time, by decreasing the "Dead/Off Times"...Or in other words to "Overlap" Square Waves just a bit...This will raise amperage and you could reduce or increase them, depending on resultant...but , for sure, it will Increase Considerably the Torque-Strength/RPM's under heavier loads...give it a try, please...as this may be an option to work with codes at heavier loads times.


Warm Regards


Ufopolitics

[lerameur]

Where to post

Hello UFO,

I am a bit confused on where to post questions concerning motors,(your motor replication) there are two threads that looks similar:
My Motors got me to Tap into Radiant Energy
and
My Asymmetric Electrodynamic Machines

Now, you have posted here:
http://www.energeticforum.com/renewa...tml#post230628
a drawing about a 21 section commutator/rotor with 4 stator section ( 4 magnets)
My stator has only 2 magnets, South and North... Do I wind the same way as a 4 magnet stator ??

is it better to have 4 magnets ? if so I could try to cut the magnets in half without breaking them.
thanks

joel

[machinealive]

I decided to use 25 turns per coil of 20.5 ga. I like it. It's measuring a solid 0.7 ohms, I've got coil 5 finished, and it's starting to stay on 0.8 - 0.7 ohms.

@ lerameur

This would be the thread for those questions. 4 magnets would be stronger, but I would use what you got, for a try. I'm not sure of pattern, but I'm sure it was posted.

[prochiro]

@joel
You could look at the seven pole diagrams but in the back of my mind I think UFO said that the 21 pole was not good. If it turns out to be OK I agree with Machine, wind and use two stators first.
Dana

[sampojo]

winding the 10-pole Quad Pent Y pattern

Hi Ufo, you really rushed this out for nuttin' took me 2 months to get here and I am a little stuck...

Hello Sampojo,

Spent all morning doing this...
It is not the same as you change all Armature structure...when you reduce the number of poles by half, must realize the "Y" Pattern "bending" will be of a more critical angle(narrower). So You MUST ALWAYS consider Stators positioning, as their bisectors related to the interacting (firing) coils at that "momentum".

It is quite different from the P5 as also from the P20...it's got its own "different look"...

[IMG][/IMG]

Explaining Diagram:

The way I chose to number Coil Pairs, was in a sequence at CCW(Starting by P1...)...and NOT following numbering per Pentagon's Groups, that is why I displayed the way it was done at the two smaller diagrams to the right.

I only set the Pairs at "work" related to brush contact not to populate too much diagram, avoiding confusion...

If you notice the Pairs at "Motor Action" are P1-P6 in Blue-Red, defining their "Y Pattern" delimited by magenta splines...as the Generating Stage are P4-P9, defining their Y Pattern by the turquoise splines(curves) [I did not wanted to hatch them(shade area) in diagram, so I did it at right images...AND if you notice they are "sharing operations here...meaning, the Motor Coils belong to Two Different Pentagon's group, as also the Generating stage.

Note that the Y Patterns between Generator and Motor are NOT Opposed(Mirrored) like in the P20 Design...which makes it easier to wind them all in same sequence.

Please notice I have set the stators bisectors, as also the firing coils bisectors for Motor Action, related to rotation sense.
Observe that P1-P6 Commutator Elements are "starting" to make contact with Motor brushes at that specific bisector positioning of their related Pairs...That, set up as is...is the more critical angle to obtain faster/strongest action...(remember the closer bisectors [Stator-Coils]are, the stronger...as the further away, the weaker machine would be.
They are very close and very well distributed, which will make a VERY STRONG Machine...

I was making a video related to a small dual pentagon I have built...making some testing to show you...but I stopped to do this...since I saw your post going for it...wanted to make sure you do it right...


Regards


Ufopolitics

I think I have gotten into a perspective problem that hopefully is OK and may just change the direction of rotation? But also I have a hunch it could make it dysfunctional. Lets see if I can describe it...

I am winding the inner S coils first, have 5 in, and now want to put in my first N coil. I went from the red P1 commutator element to S1, winding CCW, and repeated for 4 more poles, rotating the start point along the commutator elements. the problem is this diagram is the "front" of the motor, but I essentially started from the front commutator element instead of the rear (inner red) commutator element for say the P1 coil.

How might this affect your timing information?

Should I switch the wires, bit of a headache, or is there some symmetry to the asymmetry where I can easily accomodate the timing and just be getting a change in direction of rotation?

thanks a million,

Joe

[Ufopolitics]

No Problema...

Hi Ufo, you really rushed this out for nuttin' took me 2 months to get here and I am a little stuck...



I think I have gotten into a perspective problem that hopefully is OK and may just change the direction of rotation? But also I have a hunch it could make it dysfunctional. Lets see if I can describe it...

I am winding the inner S coils first, have 5 in, and now want to put in my first N coil. I went from the red P1 commutator element to S1, winding CCW, and repeated for 4 more poles, rotating the start point along the commutator elements. the problem is this diagram is the "front" of the motor, but I essentially started from the front commutator element instead of the rear (inner red) commutator element for say the P1 coil.

How might this affect your timing information?

Should I switch the wires, bit of a headache, or is there some symmetry to the asymmetry where I can easily accomodate the timing and just be getting a change in direction of rotation?

thanks a million,

Joe

Hello Joe,

First off you did not read my description "slowly"...and by doing it "too fast" you missed/jumped vital points.

Ok, you wrote:

I am winding the inner S coils first, have 5 in, and now want to put in my first N coil.

You are showing my Ten Pole Rotor Diagram here...so, assuming you are doing it and NOT a Five Poles Rotor...then you are wrong...

There are a total of Ten South Coils on the Ten Poles Config. ...and You are NOT supposed to start winding the North Coils...till done with the Ten(10) South Inner Coils.

When I wrote...

I only set the Pairs at "work" related to brush contact not to populate too much diagram, avoiding confusion...

Then look at the Two smaller diagrams at right...You see there "S10" right?


The "swapping" of Commutator order to start winding...could be easily corrected by feeding positive through the starting commutator, no matter if at top or bottom...and yes, feeding as graphic shpws...will rotate opposite to diagram...no biggie.

My main concern is the first one with not enough South Coils...

Regards


Ufopolitics

[Ufopolitics]

21 Poles...

Hello UFO,

I am a bit confused on where to post questions concerning motors,(your motor replication) there are two threads that looks similar:
My Motors got me to Tap into Radiant Energy
and
My Asymmetric Electrodynamic Machines

Now, you have posted here:
http://www.energeticforum.com/renewa...tml#post230628
a drawing about a 21 section commutator/rotor with 4 stator section ( 4 magnets)
My stator has only 2 magnets, South and North... Do I wind the same way as a 4 magnet stator ??

is it better to have 4 magnets ? if so I could try to cut the magnets in half without breaking them.
thanks

joel

Hello Lerameur,

Yes, as Machine Alive told you it is fine here...


The 21 Poles is a Four Stator, I designed for Smitev72...since his motor was a four brush/four stators assembly.
It is supposed to work fine...

The winding of a Four Stator "sometimes could" work with a two stator/two brush...(it worked for Donald Haas before on a Ten Pole design, I believe, not sure)

But it is not correct though...the windings per Pairs on a Four Stator SHOULD BE dividing 360 quadrant EXACTLY in Four Parts(90º)..as the Two Stator...Pairs MUST BE dividing 360 by Two (180º) or greater than...

Mainly, it is a matter of "matching/balancing size" of Rotor Pairs/Coils width to Stators Coils width...

Example with a 12 pole rotor:

[IMG][/IMG]

1-Two Stator will comprehend Pairs where each Coil will "grab"(wrap around) Four Poles, in order that each Pair will grab Eight Poles...but here they "share" every two poles when dividing 360 by two at every two brush revolutions...

[IMG][/IMG]

2-While a Four Stator will comprehend Pairs where each Coil will grab three Poles...Four Coils will be dividing in four the 360 .


Now , 21 is not divisible by two nor four...

And yes...a Four Stator would be stronger than just Two...


However, IF you look closely at the 21 poles Coils, note each coil grabs Six Poles, therefore Pair would wrap twelve poles...SO, I Think it should work with Two Stators...(Considering the width of stators versus Pairs) just have to "fire" Coils by setting brushes EXACTLY at center of each of the two Stators...and it should work fine...then "tune it" for better performance by adjusting them short angles.


So test this way, before going into cutting stators and adding more brushes...or "complicating it too much"...


Regards


Ufopolitics