Mostly Permanent Magnet Motor with minimal Input Power

[gyula]

Hi Luc and All,

I try to help with the calculations, if somebody does not agree with my numbers, please tell.

Assuming your 1 Ohm resistor is not inductive so we can use the simple formula of the voltage squared and devided by the resistance. (I did not consider here your capacitors in parallel with the this resistor either.)
Also, at your Ryobi motor we do not know the phase angle the motor makes but the relative current increase in the loaded cases are what counts so the voltage times the relative current increase will be considered as input power.

TEST #3
Case 1, coil is facing the magnets

Input power increase at the Ryobi when you attached the 1 Ohm load is: 67.1V*0.042A=2.818W (this can be considered as your input power in this loaded case, I considered the current increase only, the 42mA)
Power dissipated in the 1 Ohm is 1.492V*1.492V/1Ohm= 2.226W (here I use your multidigit RMS meter values, not the scope)
Efficiency, output/input power: 2.226/2.818=0.7899 i.e. 78.99% or a COP of 0.7899

Case 2, coil is sidewise to the magnets in video#3

Input power increase at Ryobi when you attached the 1 Ohm load is: 67.1V*0.005A=0.335W
Power dissipated in the 1 Ohm is 0.687*0.687/1=0.471W
Efficiency: 0.471/0.335=1.4 i.e. 140% or a COP of 1.4 Well I used the same 67.1V input voltage to the Ryobi because you mentioned in the video it was the same like in case 1, see my comments on this below.

TEST #4
Case #1 coil pair is side wise to the magnets

Input power increase at Ryobi when you attached the 1 Ohm load is: 67.1V*0.008A=0.536W
Power dissipated in the 1 Ohm is 0.609*0.609/1=0.37W
Efficiency: 0.37/0.536=0.69 i.e. 69% or a COP of 0.69

Case #2 coil pair is facing the magnets

Input power increase at Ryobi when you attached the 1 Ohm load is 67.1V*0.007A=0.469W
Power dissipated in the 1 Ohm is 0.629*0.629/1=0.395W
Efficiency: 0.395/0.469=0.8422 i.e. 84.22% or a COP of 0.8422


I think the result of Case 2 in test 3, the COP of 1.4 is doubtful because how can the Ryobi consume 0.572A (unloaded case) from the 67.1V variac voltage if in Case 1 it consumed 0.564A (also unloaded)?
The variac voltage (or the input mains to the variac) must have been increased from 67.1V to a bit higher value, to get the higher current consumption, right? but I used the same 76.1V in the calculation. So I do not think this is indeed overunity, would you mind repeating if you can?

MY questions:

If I got it right from test#4, the coils were in 'normal' parallel and not in flipped parallel, right?

The 1 Ohm load was inserted into the paralleled coils and not across the paralleled coils, right? (on the word 'across' I mean the single coil case as a comparison where you placed the 1 Ohm in parallel with the single coil)

Regarding your suggestion on using a second magnet on the other side of the coil when in the sidewise position will double the output: well I do not know, all I can say is it needs to be tested.

I tend to agree with David's comments on why you may not get similar results now like you had with your hand tests. What do you object the (perhaps less) output or the bigger drag or both?

Thanks for doing these tests.

Gyula

EDIT: I like your figuring out the possible best 'operation point' of the Ryobi motor by picking the voltage value to which the lowest current belongs because this should be the most sensitive point for testing any current increase due to the smallest load imposed to the shaft. However you have to find why the voltage or the current of the motor changes, probably the variac is not so stable, provided the mains 120V input is stable? I mean here also your test #4 where the current of the Ryobi was 0.585A unloaded in case 1 and 0.558A in case 2, also unloaded, this is too big difference when you think the input is still at 67.1V to the motor... which cannot be the case?

[gyula]

Nice setup luc!
....
I have been wondering about your wave form you have been getting. Naudins does not look like that, neither does mine. I think it may be because im using a monopole rotor, but Naudin said he was using an alternating rotor, so im not sure.
....

Hi Cody,

I think Naudin used only a single magnet for the wave form tests because he writes: "The oscilloscope pictures below, show the signal generated by a moving magnet :"

Does not use moving magnets but a magnet, and the alternating 6 magnet pairs on the rotor are meant for the Mini-Romag setup. And he also used a single air core coil for this test.

And lest we forget, member Chef above (previous page) showed a shot that looks just like Naudin's shot when the coil is side-wise. Peter also confirmed that waveform.

rgds, Gyula

[frenky]

@gotoluc:

I have found this sketch from some years ago when I built Garry Stanley anti-lenz motor.

Image shows magnetic poles and induced current when magnet is moving away from coils.


The trick with this setup is that this is ideal for motors but not for generators.

Because if lead wires are not connected, the two currents in coils cancel each other out so there is no drag.
In motor setup lead wires are connected to power supply only for a short time of the pulse.

But in generator setup the lead wires are connected all the time so current goes back into battery/capacitor and there is drag.

That is why there was no drag when you were holding coils in your hand and passing magnet by. If you would put capacitor between lead wires there would be drag and capacitor would charge.

Have fun,
Frenky


P.S.
I was so impressed by your "Mostly Permenent Magnet Motor test" youtube videos that I will start to build after quite some years of pause

[gotoluc]

I tried collecting only the collapse but it seemed to me that there was still a little bit of drag. Maybe you will have better results. Naudin seems like a very intelligent guy with lots of equipment so i tend to trust his results over mine, but i dont know. Maybe i had something hooked up wrong. And i dont have as nice as a test rig as you have.

Hi Cody,

I re-epoxied the 4 magnets on the rotor to make it a monopole and tested it on the coil with the magnets running on the side and added a diode to try to capture the collapsing spikes but I don't seem to have any luck just using a diode.

I use one diode on one leg of the coil to one side of a 10 Ohm resistor and the other leg of the coil is on the other side of the resistor with both scope grounds connected there. The green probe connected to the other between coil and yellow probe is connected between resistor (after diode)

As you can see from the first scope shot I'm collecting more then just the collapsing spike so I reversed the coil legs to see that would help and it doesn't work any better. I think a timing switch would need to be used instead. Let me know what you or others think of this.

Luc

[gotoluc]

GOTO,

the only other thing i could think of to explain what you found was this...

baring in mind that doing it by hand you had a virtually Lenz free effect, then as far as size/strength of NEO used....the bigger the better....and....by hand you were using those quite large 2x2x1 inch neos....however....on the rotor...the magnets didnt seem very large at all......could this explain the poor generation?

Hi David,

I think I got too excited the day I found this and when I quickly tested it on the small magnet rotor (holding by hand) to see if it produced power in this configuration I must of got the connections wrong since it was outputting 1.20vac on a 10 Ohm load. Maybe also couldn't tell there was drag since the magnets are so small.

I think that's maybe what happened since I tried just about everything and it doesn't get any better results.

Luc

[gotoluc]

Hi Luc and All,

I try to help with the calculations, if somebody does not agree with my numbers, please tell.

Assuming your 1 Ohm resistor is not inductive so we can use the simple formula of the voltage squared and devided by the resistance. (I did not consider here your capacitors in parallel with the this resistor either.)
Also, at your Ryobi motor we do not know the phase angle the motor makes but the relative current increase in the loaded cases are what counts so the voltage times the relative current increase will be considered as input power.

Thanks Gyula for taking the time to do this. Sounds good to me

TEST #3
Case 1, coil is facing the magnets

Input power increase at the Ryobi when you attached the 1 Ohm load is: 67.1V*0.042A=2.818W (this can be considered as your input power in this loaded case, I considered the current increase only, the 42mA)
Power dissipated in the 1 Ohm is 1.492V*1.492V/1Ohm= 2.226W (here I use your multidigit RMS meter values, not the scope)
Efficiency, output/input power: 2.226/2.818=0.7899 i.e. 78.99% or a COP of 0.7899

Case 2, coil is sidewise to the magnets in video#3

Input power increase at Ryobi when you attached the 1 Ohm load is: 67.1V*0.005A=0.335W
Power dissipated in the 1 Ohm is 0.687*0.687/1=0.471W
Efficiency: 0.471/0.335=1.4 i.e. 140% or a COP of 1.4 Well I used the same 67.1V input voltage to the Ryobi because you mentioned in the video it was the same like in case 1, see my comments on this below.

TEST #4
Case #1 coil pair is side wise to the magnets

Input power increase at Ryobi when you attached the 1 Ohm load is: 67.1V*0.008A=0.536W
Power dissipated in the 1 Ohm is 0.609*0.609/1=0.37W
Efficiency: 0.37/0.536=0.69 i.e. 69% or a COP of 0.69

Case #2 coil pair is facing the magnets

Input power increase at Ryobi when you attached the 1 Ohm load is 67.1V*0.007A=0.469W
Power dissipated in the 1 Ohm is 0.629*0.629/1=0.395W
Efficiency: 0.395/0.469=0.8422 i.e. 84.22% or a COP of 0.8422

I think the result of Case 2 in test 3, the COP of 1.4 is doubtful because how can the Ryobi consume 0.572A (unloaded case) from the 67.1V variac voltage if in Case 1 it consumed 0.564A (also unloaded)? The variac voltage (or the input mains to the variac) must have been increased from 67.1V to a bit higher value, to get the higher current consumption, right? but I used the same 76.1V in the calculation.

I can explain the differences. What I have not mentioned in the video is as the motor runs longer and longer it keeps warming up the current draw keeps going down. This is why the next test the current is lower before loading. I've kept my eyes on voltage and it's very stable, no more then + or - 0.5 volts.

So I do not think this is indeed overunity, would you mind repeating if you can?

I didn't realize the score was so good on that test. However it is not as good as your calculations above. You may of missed that I pointed out in test 4 that the side coil test in test 3 would be closer to using 0.007A and not the 0.005A which was used in you calculations above. However, I still think it's indicating OU using 0.007A so I do agree that this test would need to be redone with the most precision as possible. However, this is now a problem since I now changed the rotor to a monopole and in the process damaged the 4 magnets that I flipped around since I used a 1" square Iron bar stuck to the magnets that I heated with a torch to heat the magnet to the point so the epoxy releases but after noticed that the magnets lost 80% of their power. I don't know if heat kills Neo magnets or heat with the iron bar combination transferred their magnetism in the iron The iron bar is not magnetized though ... Anyways, I had 3 extra ones so I used those but I had to use one of the heated ones and as you can see from the scope shots above I now have a very weak pulse in the rows ... which could help to tell each rotations of the magnet rotor.

MY questions:

If I got it right from test#4, the coils were in 'normal' parallel and not in flipped parallel, right?

Yes, that is correct.

The 1 Ohm load was inserted into the paralleled coils and not across the paralleled coils, right? (on the word 'across' I mean the single coil case as a comparison where you placed the 1 Ohm in parallel with the single coil)

Yes, that is correct.

Regarding your suggestion on using a second magnet on the other side of the coil when in the sidewise position will double the output: well I do not know, all I can say is it needs to be tested.

Yes, I agree. This would need to be tested.

I tend to agree with David's comments on why you may not get similar results now like you had with your hand tests. What do you object the (perhaps less) output or the bigger drag or both?

Better read my reply to David (above)... I think this is what may of happened.

Thanks for doing these tests.

Gyula

EDIT: I like your figuring out the possible best 'operation point' of the Ryobi motor by picking the voltage value to which the lowest current belongs because this should be the most sensitive point for testing any current increase due to the smallest load imposed to the shaft.

This is exactly why I did this. When one find this balance point on induction motors they become a very sensitive and will display very small changes with RPM drops (rotor slip) and current draw. Glad you like the idea

However you have to find why the voltage or the current of the motor changes, probably the variac is not so stable, provided the mains 120V input is stable? I mean here also your test #4 where the current of the Ryobi was 0.585A unloaded in case 1 and 0.558A in case 2, also unloaded, this is too big difference when you think the input is still at 67.1V to the motor... which cannot be the case?

This was answered above.

Let me know if you have new questions or concerns with this reply

Thanks for all your time.

Luc

[cody]

Hi Cody,

I re-epoxied the 4 magnets on the rotor to make it a monopole and tested it on the coil with the magnets running on the side and added a diode to try to capture the collapsing spikes but I don't seem to have any luck just using a diode.

I use one diode on one leg of the coil to one side of a 10 Ohm resistor and the other leg of the coil is on the other side of the resistor with both scope grounds connected there. The green probe connected to the other between coil and yellow probe is connected between resistor (after diode)

As you can see from the first scope shot I'm collecting more then just the collapsing spike so I reversed the coil legs to see that would help and it doesn't work any better. I think a timing switch would need to be used instead. Let me know what you or others think of this.

Luc

Thanks for doing those tests. When you said you were going to change your rotor around i was afraid something bad would happen, i just ruined some of my neo's too , and yes, heat will kill them. The second scope shot you posted in the yellow line is how it should look with the diode. That shot appears to me that its working fine, except it looks like you have a bad magnet on the rotor or something funky is going on midway there. So if its working fine, i wouldnt think you would need a switch. Since its such a low voltage, you may try the secondary from microwave transformer to get a higher voltage to work with, but its probably not necessary. On your scope, are you getting twice the voltage on the collapse when the coil is unloaded or not hooked to the resistor or diode?

[gotoluc]

Thanks for doing those tests. When you said you were going to change your rotor around i was afraid something bad would happen, i just ruined some of my neo's too , and yes, heat will kill them. The second scope shot you posted in the yellow line is how it should look with the diode. That shot appears to me that its working fine, except it looks like you have a bad magnet on the rotor or something funky is going on midway there. So if its working fine, i wouldnt think you would need a switch. Since its such a low voltage, you may try the secondary from microwave transformer to get a higher voltage to work with, but its probably not necessary. On your scope, are you getting twice the voltage on the collapse when the coil is unloaded or not hooked to the resistor or diode?

Hi cody,

Okay, thanks for confirming that. I'll test it and post the data.

Concerning the problem. After heating the 4 neo's to un-glue them and noticing they had next to no magnetic power left. I replaced 3 with extra ones I had but had to use one of the heated ones as that's all I have. This is the low pulse you see on the scope shots above.

Luc

[nilrehob]

Hi Luc,

Good work, as always!
Probably old news for you: take a look at steorns toroids and 2 magnets/coil around 1:49
YouTube - Steorn Orbo Technology Launch 2009

/Hob

[gyula]

Hi Frenky,

I wonder if you have already seen this test by Ron on Garry's parallel connected flat (taken out from old floppy drives) air core coils:
Garry 2 flat coils motor tests by Ron Pugh

(I am aware that Ron tests are done with facing two rotor magnets in attraction, with the paralleled coils sandwiched between them, as Garry used it on his bicycle, your sketch shows one rotor magnet sandwitched between the two paralleled coils.)

I agree with you that we HAVE TO differentiate between a generator and a motor mode wrt these parallel coils, Garry always mentioned this he used this for motor mode only, Ron and others tested it for generator mode and found Lenz effect.

Have you ever managed to make output torque tests on you Garry motor in the past? That would be the best to learn on efficiency or COP numbers.

Thanks,
Gyula

@gotoluc:

I have found this sketch from some years ago when I built Garry Stanley anti-lenz motor.

Image shows magnetic poles and induced current when magnet is moving away from coils.


The trick with this setup is that this is ideal for motors but not for generators.

Because if lead wires are not connected, the two currents in coils cancel each other out so there is no drag.
In motor setup lead wires are connected to power supply only for a short time of the pulse.

But in generator setup the lead wires are connected all the time so current goes back into battery/capacitor and there is drag.

That is why there was no drag when you were holding coils in your hand and passing magnet by. If you would put capacitor between lead wires there would be drag and capacitor would charge.

Have fun,
Frenky


P.S.
I was so impressed by your "Mostly Permenent Magnet Motor test" youtube videos that I will start to build after quite some years of pause

[gyula]

Hi Luc,

I answer below.

....
I can explain the differences. What I have not mentioned in the video is as the motor runs longer and longer it keeps warming up the current draw keeps going down. This is why the next test the current is lower before loading. I've kept my eyes on voltage and it's very stable, no more then + or - 0.5 volts.

OK, I understand and it is clear the Ryobi coil resistance increases on the longer runs due to heat.

I didn't realize the score was so good on that test. However it is not as good as your calculations above. You may of missed that I pointed out in test 4 that the side coil test in test 3 would be closer to using 0.007A and not the 0.005A which was used in you calculations above. However, I still think it's indicating OU using 0.007A so I do agree that this test would need to be redone with the most precision as possible. However, this is now a problem since I now changed the rotor to a monopole and in the process damaged the 4 magnets that I flipped around since I used a 1" square Iron bar stuck to the magnets that I heated with a torch to heat the magnet to the point so the epoxy releases but after noticed that the magnets lost 80% of their power. I don't know if heat kills Neo magnets or heat with the iron bar combination transferred their magnetism in the iron The iron bar is not magnetized though ... Anyways, I had 3 extra ones so I used those but I had to use one of the heated ones and as you can see from the scope shots above I now have a very weak pulse in the rows ... which could help to tell each rotations of the magnet rotor.

Well, if I consider the 0.007A for the input current increase, it gives an efficiency of 100.1% (COP of 1.001), still just ou... I understand that in the meantime you have changed the magnet setup, so maybe later this could be done again.
Yes, Neo magnets are the most sensitive to heat among the rare earth types, the cheapest products can be used up to 80 degree C only, not higher due to degeneration. Higher temperature types are manufactured of course up to 120-150 degree C, and they are more expensive and designated with a letter H in their type name like N50H etc.

Better read my reply to David (above)... I think this is what may of happened.

Maybe you could "reverse engineer" from your earlier videos how the coils were connected when you did the very promising hand tests?

Thanks,
Gyula

[frenky]

Have you ever managed to make output torque tests on you Garry motor in the past? That would be the best to learn on efficiency or COP numbers.

Hi.

My first Garry stanley replication was made from hard drive bearing, rotor containing hard drive magnets and stator with floppy disk pancake coils. Connected like on the sketch.
The fist time that I connected it to the power it just started accelerating like creazy and a few seconds later magnets were flying allover the room.

So then I decided to build more robust version but I failed big time. I have changed design too much and have used bearings with too much friction.

After that I jumped onto J.Naudins replication of Newmans motor and went on to that: (My Newmans motor) Then just newer wen't back to the Stanleys designs.

Have fun,
Frenky

[frenky]

Anyway inspired by gotoluc's latest work I'm going to test this setup.

I will use about 50 NdFeB magnets and bifilar coil.

I'm interested to see how much enery is needed to lift magnets with mass m for a distance h.
I will use capacitor to power up the coil beacuse I can calculate stored energy.

Frenky

[rave154]

nice simple looking test there Frenky,

while youre doing it, you might want to also add in a diode to feed the BEMF from the collapsing coils field as shown in GOTOLUCS "recycling BEMF" thread, as that showed quite a dramatic increase in magnet-movement when BEMF was fed back into the coil. Of course to do this you would have to connect and then dis-connect the cap sharply...as opposed to letting the cap simply "drain" slowly through the coil.

Could you also test in a similar way to LUC's test, by testing repulsion from the ends of the coils....and...also...the sides of the coils ( even though that sounds a little silly )

David. D

[frenky]

Hi.

As soon as I get magnets all will try all of these. Tnx for the input.

This one is also in plan. Both groups of magnets joined by nonferromagnetic material and lifted with coil to measure needed energy.


Frenky