Induction Delay

[madMAx4]

Hi,

i'm trying to understand why it is that the induced voltage in a coil, by moving magnets, is sometimes more than 90° after the magnetic sinus.

Let me tell you what my setup is.
Basically it is an Adams motor with bipolar magnet setup. So, when there is a north pole facing the coils, the next magnet would show a south pole and so on.
There are 8 magnets on the rotor, currently only one 60 ohm drive coil with ferrite core and one generator coil with iron core.

So when the rotor spins at 3000 rpm, and i watch the induced voltage in the generator coil, i see that the voltage is say 15° after what it should be.
This causes the rotor to accelerate when i short the generator coil out.
Because the cemf is now 2x 15° out of phase.
If the voltage is exactly 90° after the magnetic sinus, like it should be, there would be no acceleration at SC because the cemf would accelerate the magnets the same amount it slowed them down when the magnets approached the generator coil.

I can actually accelerate the motor by feeding a transformer to get a higher voltage, which i then rectify and put back to one of the input capacitors

So my question is, why does the induced voltage lag behind at higher frequencies? It only works with iron and a lot of copper, say at least 400 turns.

[EthelAether]

Hi,

i'm trying to understand why it is that the induced voltage in a coil, by moving magnets, is sometimes more than 90° after the magnetic sinus.

A commonly held misconception is that the peak amplitude of induced EMF occurs when the rotating magnet aligns with the axis of the coil (magnet at TDC). To phrase it differently, people commonly believe that the AC waveform peaks out when the magnet and coil are closest.

This is exactly the opposite of what actually happens. In actuality, the AC zero crossing coincides with the alignment. This means that the peak induced EMFs occur before & after the TDC point, and are of a polarity determined by the relationship between the polarity of the magnet and the handedness of the coil.

Now of course, there are no perfect magnets or perfect coils, and the true magnetic TDC may (or, will) be slightly offset from the mechanical TDC - but the point remains.

Induced EMF is a direct function of the rate-of-change of flux through the circuit. You can easily visualize that when the magnet reaches TDC, the RoC falls to zero ("no more magnetization to be had"). Only when the magnet is approaching or retreating is there a net RoC of flux.

[Ted Ewert]

Hi,

i'm trying to understand why it is that the induced voltage in a coil, by moving magnets, is sometimes more than 90° after the magnetic sinus.

Let me tell you what my setup is.
Basically it is an Adams motor with bipolar magnet setup. So, when there is a north pole facing the coils, the next magnet would show a south pole and so on.
There are 8 magnets on the rotor, currently only one 60 ohm drive coil with ferrite core and one generator coil with iron core.

So when the rotor spins at 3000 rpm, and i watch the induced voltage in the generator coil, i see that the voltage is say 15° after what it should be.
This causes the rotor to accelerate when i short the generator coil out.
Because the cemf is now 2x 15° out of phase.
If the voltage is exactly 90° after the magnetic sinus, like it should be, there would be no acceleration at SC because the cemf would accelerate the magnets the same amount it slowed them down when the magnets approached the generator coil.

I can actually accelerate the motor by feeding a transformer to get a higher voltage, which i then rectify and put back to one of the input capacitors

So my question is, why does the induced voltage lag behind at higher frequencies? It only works with iron and a lot of copper, say at least 400 turns.

This sounds a lot like the Thane Heins effect where a high impedance generator coil goes from inductor to capacitor and accelerates the prime mover when put under load. Voltage usually leads current in an inductor. However, it's just the opposite with a capacitor, so you may be seeing a capacitive effect occurring.

A commonly held misconception is that the peak amplitude of induced EMF occurs when the rotating magnet aligns with the axis of the coil (magnet at TDC). To phrase it differently, people commonly believe that the AC waveform peaks out when the magnet and coil are closest.

People are right if it's a coil with an iron core. Your theory is only valid with an air core coil. In an air core the flux only travels through the windings and generates voltage accordingly.
With an iron core it's different; the voltage peaks at TDC. This is because as a magnet approaches the coil, the flux is attracted to the core and polarizes it immediately. This prevents the coil from producing any polarity of charge other than what the core polarity determines. The flux is also primarily traveling through the core, and not the windings, which further reinforces this phenomenon. Try it and you'll see what I mean.

Ted

[madMAx4]

Now of course, there are no perfect magnets or perfect coils, and the true magnetic TDC may (or, will) be slightly offset from the mechanical TDC - but the point remains.

This is not the case here, i positioned the photo sensor correcly.
With a ferrite core coil the signal from the photo sensor is exacly at the zero crossing.

Induced EMF is a direct function of the rate-of-change of flux through the circuit. You can easily visualize that when the magnet reaches TDC, the RoC falls to zero ("no more magnetization to be had"). Only when the magnet is approaching or retreating is there a net RoC of flux.

I know, this is what i mean by: "If the voltage is exactly 90° after the magnetic sinus, like it should be..."

However, it's just the opposite with a capacitor, so you may be seeing a capacitive effect occurring.

The induced current is almost 90°, let's say 88°, after the voltage at short circuit. Those 2° are mostly caused by the winding's resistance i think. So there's very little capacity involved i think.

With an iron core it's different; the voltage peaks at TDC.

I've never seen that happening, that would be very strange. Have you ever measured this?
What i'm measuring is something between what EthelAether and you are saying.
Let's call the situation where the voltage is zero crossing at TDC: 0°.
Then, what you are saying, Ted, would be 90° at TDC.
What i'm measuring is 10° to 15°.
I would be very interested in how you would achieve 90°

The flux is also primarily traveling through the core, and not the windings, which further reinforces this phenomenon.

This is interesting. My experience is, the bigger the core diameter the more degree of delay you get.
I have 15mm magnets. With 10mm cores i get very little delay, let's say 5° or so. With 15mm cores i get those 10 to 15°.
This could be the reason why Adams says the core diameter should be double the magnet diameter.
I need to test 30mm cores now. This could be interesting
Thanks.

[cody]

Madmax4,

Interesting discussion. Some things are coming to mind that may or may not be correct. From what you guys are describing i am envisioning that the voltage can lead or lag tdc depending on your coil inductance and rotor frequency. The same phasing issues delt with in power factor correction for reactive components like the coil your using. Here is some info on that, look at the power factor in linear circuits
Power factor - Wikipedia, the free encyclopedia



I believe the effect is basically described is within the definition. Perhaps someone knows a more precise terminology to describe this phasing issue im talking about when dealing with these reactive components under load. Little help from the guys smarter than me It will be good to know exactly what we are dealing with so we can do some calculations to optimize preformance.


Can you try a different coil with a different inductance and tell us if the phasing is different. Also how much does rotor speed effect the shift the phasing if any?

[TanTric]

...

hi madmax,

i cannot help you but i have a doubt, lol, whats the input(V/A) of your setup, also force of magnet if you can say?

you also observe the same effects if it is homopolar rotor setup?

hugs

[madMAx4]

Can you try a different coil with a different inductance and tell us if the phasing is different. Also how much does rotor speed effect the shift the phasing if any?

The effect increases with rotor speed and core diameter. Also the thickness of the whole winding has an effect, the more copper is far away from the core the more delay you get. This, i already tested and measured.
But i think you can increase the core diameter and leave the overall coil diameter as it is and you get more delay also.

i cannot help you but i have a doubt, lol, whats the input(V/A) of your setup, also force of magnet if you can say?

you also observe the same effects if it is homopolar rotor setup?

The input is about 20W for the motor at full speed and an additional 4W for the circuit.
But this is not the issue here, i'm not saying i got OU. Just relative OU, you take something out the system and don't pay the full price.
Thus you can accelerate the motor, but it does not run on its own... yet
The effect has to be understood and a proper coil needs to be found. Thus i came here

There is no additional energy going into the system, its just that the counter EMF doesnt fully slow down the magnets.
If the voltage is zero crossing at TDC and the current lags the voltage 90° then the current would be at its maximum at TDC. That means it slows the magnet down until TDC, after which it accelerates the magnet. Thus at SC the rotor speed stays the same.
But if the voltage itself lags behind, then the current max would be after TDC, and thus the magnet would be accelerated more.
This means the more voltage lag the less you pay for the generated power

I didnt test homopolar, no need to rip out all the magnets and change the circuit . But i think the effect would be the same, maybe not as efficent. Mr Adams works only with homopolar setups i think...

[i_ron]

confirmation

Madmax,

Ether Aether is absolutely right. Many people subscribe to the TDC maximum potential theory... even Thane himself taught that, lol, But a single magnet passing a single coil generates both the positive and negative portions of the sine, with the zero crossing at TDC.

What you describe in your experiments with increased core dimensions and varying speeds I have not personally duplicated so will accept your findings as correct. My tests were done at one core size, varying numbers of turns and one speed only...1740 RPM.

What I found was that shorting the coil exactly reversed the polarity of the core.

I would post pictures but have no idea how to do that on this list

Ron P

[nvisser]

Ron.
You can post pictures when you reply to the post . Just scroll down to "Manage Attachments" and upload your photo's or files there.

[i_ron]

Ron.
You can post pictures when you reply to the post . Just scroll down to "Manage Attachments" and upload your photo's or files there.

Thanks, nvisser!

The first picture, "8e", shows the flux plot, Channel A, of a gauss meter hall probe at the outboard end of the core. The hall probe is inserted so as a north pole on the rotor gives a rising sine on the gauss meter.

It can be seen that the flux plot is exactly similar to a voltage plot.

In the second picture the coil has been shorted... the polarity of the flux field is exactly reversed!

Interesting discussion on EV gray at this time on shorting the coil at the top of the sine wave, for an increase in output.

Jules writes..."You are moving the virtual phase so that it is closer to the current node. There is a resonance point and when you cross it you get a magnetic spin reversal. Diamagnetic action from Lenz's law turns to paramagnetic as the system goes OU"

So yes, I got into this with the Thane coil, for which these tests were done... but never to the point of OU

Ron

[madMAx4]

Ether Aether is absolutely right. Many people subscribe to the TDC maximum potential theory... even Thane himself taught that, lol, But a single magnet passing a single coil generates both the positive and negative portions of the sine, with the zero crossing at TDC.

I agree, zero crossing is usually at TDC.

What I found was that shorting the coil exactly reversed the polarity of the core.

I agree also, but not in the sense that it reverses the entire core, i think it partially demagnetises the core.
EDIT: after seeing your scope shots i dont think so anymore . Seems like the cemf field is actally stronger than the magnet's field, at least at the back end. It would be interesting so see what's going on at the front end.

I got some scope shots here. They are made through an PCI DAQ with 16bit samples and 500kHz sample rate. Software is self coded on Linux.

In the first picture is the generated voltage (5V/DIV) from my 1.6 Ohm coil with M12 Screw core. You can see that the voltage lags about 10°.
Those dips at the sinus maxs are because of the bad shape of my magnets, 15mm x 8mm. They need to be longer.

In the second picture is a 10mm ferrite core coil. You can see that the zero crossing is at TDC.

EDIT: @Ron: what resistance does your coil have, and how long is it? And what core does it have, what diameter?

[i_ron]

I agree, zero crossing is usually at TDC.

@Ron: what resistance does your coil have, and how long is it? And what core does it have, what diameter?

Here is a picture of the setup... the core is about 5/8 X 7/16 X 3 3/4 inches,
about 1/2 an ohm, 4 mH, the coil is about 2 1/4 inches long.

One doesn't have to use a gauss meter, a small sensing coil over/behind the main coil will show the same thing?

Ron

edit... oh, not the same coil as in the pic, but same core, this coil above was large gauge wire wound to see if there was any difference in the reversal... there wasn't

edit 2, please note that air cored coils react differently... the scope shots only apply to steel laminate cores.

[i_ron]

I agree, zero crossing is usually at TDC.

Those dips at the sinus maxs are because of the bad shape of my magnets, 15mm x 8mm. They need to be longer.

Max,

I think you will find the it is the magnet spacing that causes the "M wave"

The incoming pole's sine must precisely overlay the outgoing magnet's sine to make an exact sinewave...in other words, magnet spacing

Ron

edit: in an NSNS rotor