PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2019-12-09, 18:12:36
News: If you have a suggestion or need for a new board title, please PM the Admins.

Pages: 1 ... 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 [18] 19 20 21 22 23 24 25
Author Topic: Professor Walter Lewin's Non-conservative Fields Experiment  (Read 253036 times)

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
The proposed test may be wrong.

Each measurement should encompass the same angle rather than the same length of wire, correct?
   
Group: Guest
The proposed test may be wrong.

Each measurement should encompass the same angle rather than the same length of wire, correct?

The proposed test can't be wrong.  It either confirms or denies the theory.  


Edit:  When I say vary the de-coupled length, I meant the length connecting to probes normal to the loop. I know it's very tall already, but the relative tall/shortness maybe a factor affecting measurement. 





« Last Edit: 2012-04-08, 15:50:45 by GibbsHelmholtz »
   

Group: Elite
Hero Member
******

Posts: 1859

The induced voltage is not affected by series resistors nor decoupled measurement leads. That's been proven.


Nor is it affected by the resistances which are part of the loop.

The measurement leads (the meter circuit) presents a change in loop area, for one side of the -now- double loop.

Quote
I disagree. Again, a non-circular loop isn't optimal for induction. My diagram will show why.

Here is where our disagreement of solenoid position within the loop affecting or not affecting induced voltage comes into play.

In practice, performing current monitoring of a conductor via a current transformer does not change whether the cable is centered within the CT or not. CT's come in round or rectangular. You can attach a rectangular CT to a cable with the cable at one extreme inside corner. It does not have an effect upon the induced voltage. I have done this on numerous occasions. There has never been any value in centering the monitored cable within a surrounding current sensor for purposes of accuracy.

Quote
 
I don't understand this concept. Can you think of a test to prove this? The tests I've done to specifically address this issue clearly indicate that the measurement loop is not affected by the experiment. If there is no induction in the measurement loop (which I've proven for myself), and the measurement loop is high impedance (10M), how can it affect the induction loop?

There would be no induction within the meter circuit (as you have performed). The loop impedance has no effect upon the measured induced emf unless the secondary load is too high. Even then, the measured induced voltage will remain at 1V until Lenz starts to affect the amount of power being transformed.

The loop impedance isn't part of the formulas for induction. Even if the meter circuit was zero Ohms or infinite resistance, the emf induced for that loop will be 1V. The meter may show any value but the accuracy or ability to show the voltage decreases as resistance of the meter loop drops.

As testing should conclude, even when there is a diametrically connected short, the total induced voltage around the outside loop will still be 1V.
It should also be found that the induced emf for the loop circuit including the diameter short (say, on the left of the diameter short) will be 1V. The same should be found for the complete half circle on the right of the diameter short.

The confusing part is that the diameter short now has two induced currents opposing each other.

I'll have to think about a test to prove the idea. This becomes a relativity issue.


---------------------------
"As far as the laws of mathematics refer to reality, they are not certain; as far as they are certain, they do not refer to reality." - Einstein

"What we observe is not nature itself, but nature exposed to our method of questioning." - Werner Heisenberg
   

Group: Elite
Hero Member
******

Posts: 1859
The proposed test can't be wrong.  It either confirms or denies the theory.  


Edit:  When I say vary the de-coupled length, I meant the length connecting to probes normal to the loop. I know it's very tall already, but the relative tall/shortness maybe a factor affecting measurement.  







Gibbs,

The height should make no difference as long as wire R/L/C doesn't skew the results. You must look at the test from above, in your mind's eye.

The metering circuit, with leads, simply crosses from one point of the measured loop to another point. If you could place the scope within the measured loop with sensing leads straight to each of the two previous points (and the scope had no effect upon the magnetic fields), the same values would be recorded.

BTW:

These aren't my theories. The above was part of my education relating to transformer theory.



---------------------------
"As far as the laws of mathematics refer to reality, they are not certain; as far as they are certain, they do not refer to reality." - Einstein

"What we observe is not nature itself, but nature exposed to our method of questioning." - Werner Heisenberg
   
Group: Guest
...
I know that outside the solenoid the electric field drops off as a function of 1/r. And we know that the electric field is what is responsible for the induced emf. The emf being equal to the line integral of the electric field around the conductive loop of interest.
...

Only an electric field deriving from a potential drops off as a function of 1/r, not an induced emf.
An induced electric field is conservative because ∇xE = -∂B/∂t and B is conservative.

From the flux variation ∮E.dl=-dΦ/dt and Φ=∮B.dA.  B is conservative and null outside the solenoid (provided that it is long enough), so Φ is the same whatever the surface A of the loop around the solenoid.

All equations gives the same result. The first one straightforwardly explains why the emf doesn't depend on the loop diameter, provided that the same flux crosses the surface (which is experimentally not obvious to realize).

Note that the above applies only in the quasi-static fields approximation: no radiation by em waves, i.e. the circuit size must remain negligible compared to the wave length of the involved signals.

   
Group: Guest
Gibbs,

The height should make no difference as long as wire R/L/C doesn't skew the results. You must look at the test from above, in your mind's eye.

The metering circuit, with leads, simply crosses from one point of the measured loop to another point. If you could place the scope within the measured loop with sensing leads straight to each of the two previous points (and the scope had no effect upon the magnetic fields), the same values would be recorded.

BTW:

These aren't my theories. The above was part of my education relating to transformer theory.



WW,

Maybe it's true.  However, imagine my rectangular loop voltage V3 (V123-3), if we reduce the height to 0, voltage would be zero. 


   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
The proposed test can't be wrong.  It either confirms or denies the theory.  

Sure it can be. I believe measurements with comparable angle are what's called for. But I will do both.
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
Nor is it affected by the resistances which are part of the loop.
Implied, agreed.

Quote
The measurement leads (the meter circuit) presents a change in loop area, for one side of the -now- double loop.
A change in loop area with respect to what?

Quote
In practice, performing current monitoring of a conductor via a current transformer does not change whether the cable is centered within the CT or not. CT's come in round or rectangular. You can attach a rectangular CT to a cable with the cable at one extreme inside corner. It does not have an effect upon the induced voltage. I have done this on numerous occasions. There has never been any value in centering the monitored cable within a surrounding current sensor for purposes of accuracy.
I know this.

Quote
As testing should conclude, even when there is a diametrically connected short, the total induced voltage around the outside loop will still be 1V.
It should also be found that the induced emf for the loop circuit including the diameter short (say, on the left of the diameter short) will be 1V. The same should be found for the complete half circle on the right of the diameter short.

The confusing part is that the diameter short now has two induced currents opposing each other.
I am interested only in the Lewin experiment, and the test Gibbs proposed. What you have proposed, I have not done, nor is it all that relevant to these tests.

Quote
I'll have to think about a test to prove the idea. This becomes a relativity issue.
I am asking for a proposed test to prove your notion that introduction of the decoupled meter leads somehow skews the results, or enters into the equation.
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
Only an electric field deriving from a potential drops off as a function of 1/r, not an induced emf.

Not sure what you mean Ex. Let me restate it again:

The induced electric field at a point outside the solenoid diminishes as a function of 1/r.

The line integral of the E-field (which equals induced emf) at radius r does not diminish as a function of 1/r.

Better?  ;)
   

Group: Elite
Hero Member
******

Posts: 1859
A change in loop area with respect to what?

The other loop now created by insertion of the metering circuit. Please continue with the route you are taking.

Quote
I am interested only in the Lewin experiment, and the test Gibbs proposed. What you have proposed, I have not done, nor is it all that relevant to these tests.

There should be a point where you see my statement is at the heart of the Lewin experiment.

Quote
I am asking for a proposed test to prove your notion that introduction of the decoupled meter leads somehow skews the results, or enters into the equation.

Whatever I suggest must first show that insertion of the meter circuit does indeed segment the secondary loop into two loops. Since an additional meter would do exactly the same - again - ..... Well, I don't think it is that simple.

Not that the experiment is beyond you. Just that the experiment is beyond me, right now.

 


---------------------------
"As far as the laws of mathematics refer to reality, they are not certain; as far as they are certain, they do not refer to reality." - Einstein

"What we observe is not nature itself, but nature exposed to our method of questioning." - Werner Heisenberg
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
The other loop now created by insertion of the metering circuit. Please continue with the route you are taking.
I don't understand what affect that has on anything? We are adding a non-loading (essentially) non-inducing loop of wire that sits in a plane parallel to the changing flux. Two ends of that loop happen to be connected somewhere to two points on the loop that is situated normal to the changing flux. What impact does the measurement loop have on anything at all, other than the miniscule differential voltage it presents to its high impedance input amplifier/buffer at the other end of the wire loop?

Surely you're not suggesting that the tiny current in the measurement leads is skewing the experiment? Let's look at an example: Say we are measuring 400mV across two points on the loop. Let's also say the input impedance of the scope is 10M. This represents a current in the measurement leads and the loop of 40nA. This is 0.004% of the loop current of 1mA. I don't think 0.004% is anything to worry about.

Quote
There should be a point where you see my statement is at the heart of the Lewin experiment.
The Lewin experiment has been explained as far as I am concerned. The flaws in the presentation have been brought to light, and the resulting erroneous conclusions corrected. Have you not been paying attention?  >:-)

Quote
Whatever I suggest must first show that insertion of the meter circuit does indeed segment the secondary loop into two loops. Since an additional meter would do exactly the same - again - ..... Well, I don't think it is that simple.
Again, so what if there is a second loop which is parallel to the changing flux and connected to the first? You mention this, but do not explain what impact you believe it has on anything. Apologies if I've missed it.

   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
Here is an updated diagram with two points added; "00" and "11".

As promised, here are the 3 measured voltages around the rectangular loop:

V1-0 = -0.09V
V2-0 = +0.14V
V3-0 = +0.075V

Back to my proposed tests, here are the results:

V1-0 = -0.09V     |Same angle (90º) &
V11-00 = -0.09V  |same wire length for both measurements

V0b-0a = -0.03V   |Probably 30º angle
V1a-V0b = -0.06V |45º angle, same wire length for both measurements

Conclusion? Appears it IS the angle which is relevant, vs. wire length. In the first pair of measurements, the measured emfs were the same, despite one measurement taken across a 90º corner (V1-0). It also appears that the loop shape doesn't matter for the same induce emf. I won't be able to prove this beyond a doubt until I make a solenoid with a much higher length to diameter ratio.

I also re-ran the emf test (slightly differently)with loops of R and 2R, and the same result: with R, emf= 0.4V, with 2R, emf=0.3V. As Ex said, this could be due to the solenoid length. If it was much longer, the results might be more consistent. Will have to try it with the new solenoid.
   
Group: Guest

I don't know what to say.  Thank you.

Gibbs
   

Group: Elite
Hero Member
******

Posts: 1859
.99,

It doesn't matter if your scope has a 100M input impedance the same emf is induced into the loop and 'any' connection of two points to that loop segments it into more than one loop.

I t has been a very long weekend. I'll check back as soon as I can.

Good Experimenting to all!



---------------------------
"As far as the laws of mathematics refer to reality, they are not certain; as far as they are certain, they do not refer to reality." - Einstein

"What we observe is not nature itself, but nature exposed to our method of questioning." - Werner Heisenberg
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
.99,

It doesn't matter if your scope has a 100M input impedance the same emf is induced into the loop
Of course the same emf is induced. Where did I ever state that the measurement loop would cause otherwise? And you are referring to emf in the secondary loop I trust, NOT the measurement loop?

Quote
and 'any' connection of two points to that loop segments it into more than one loop.
Again, so what? Please elaborate on what you believe is the significance of this.

   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
It doesn't matter if your scope has a 100M input impedance the same emf is induced into the loop and 'any' connection of two points to that loop segments it into more than one loop.
I'm going to take a guess at what you are trying to indirectly say:

I think you are implying that the induced emf is always the same, and when it is measured as otherwise (as I have going from R to 2R), the change is due to the use of a decoupled measurement device and its leads being connected to the loop. Is that a fair conjecture?

If I've guessed right, then I can't support that notion. The measurement device and its leads and connections are having absolutely no effect on the measurements, as proven on the bench.

What I believe IS making the difference in the measurement, is the non-ideal solenoid in terms of its length to diameter ratio. My current solenoid has a Length to Diameter ratio of about 3:1 (as does Lewin's). For loops that hug the solenoid (R), this is not a problem; the induced emf is as it should be. For loops in the 2R area and larger, this is most likely a problem. The lower emf in the larger loops has nothing to do with the measurement leads at all. It simply can't. The solenoid really is inducing a lower emf in the larger loops, because the loop radius is relatively large compared to the solenoid length.
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
A hypothesis of what happens when rloop > rsolenoid, and the ratio  Lsolenoid:rsolenoid is relatively low:

In this case I believe the B-field in the region of the larger loops is non-zero. It not only has magnitude, but the opposite direction as it curls back to the other end of the solenoid. A such, a weaker and opposite electric field is produced near the loop and this cancels the effect of the primary electric field to a degree, hence the lower than expected induced emf.
   

Group: Elite
Hero Member
******

Posts: 1859
Of course the same emf is induced. Where did I ever state that the measurement loop would cause otherwise? And you are referring to emf in the secondary loop I trust, NOT the measurement loop?

Yes, I try to make it a point to call the secondary 'loop' and the metering circuit - the metering circuit.

Quote
Again, so what? Please elaborate on what you believe is the significance of this.

Regardless if the metering circuit has emf induced into it or not the metering circuit either segments or sectors the loop (if on the inside) or becomes another loop in addition to the secondary loop(if on the outside).

Yes, solenoid length will have pronounced effects and there is a point where an extended loop may see 'up' and 'down'.  I feel those issues are going beyond the initial experiment.

I keep saying 'view from above' - 'in your mind's eye'. If nothing else, when you have the metering circuit crossing over the solenoid (decoupled, if you choose) stand on the table and look directly down at the top of the solenoid and loop. Forget the height differences for a moment and imagine the whole experiment is in 2D. The metering leads cut across in some fashion. Even though they are not the subject of induction they change the Gaussian surface area. They don't need to be on the same plane.

As you now see, the angle of the metering circuit does matter. Why? Because changing the angle changes the Gaussian surface area of both sections of the loop now sectioned by the metering leads.

This is why I say there is no way to connect a metering circuit without changing something in the experiment.

The voltages you are seeing are the voltage induced on one loop minus the voltage of the other loop (because the original loop is now segmented or sectored by the metering circuit. The same voltages you are measuring have very little to do with the resistors.

If the resistors were equal in resistance, even down to zero Ohms, and you measured from original points A to D you will see .5V. The reason, both sectors of the original loop now have 1V induced into each of them.... At the joining conductor (the metering circuit) what little voltage is seen is the result of two opposite polarity 1V induced into each segment.

The only reason .4V was seen between A-D was because the resistors skewed the angle of the metering leads. Your 'normal to the loop' metering circuit was normal to the loop only in a mechanical sense.

Since my above description is exactly how angular position sensors work (as explained in the Maxim Application sheet brought up early in this thread) I really don't feel I need to argue the point. It would be a great thing to share the understanding but that didn't work for me when it was 'shared'. So, I can't blame you if you choose another path.

In these cases, the chosen path is almost always the best. Experimental results rule but observations must also be questioned until they can no longer be questioned.


---------------------------
"As far as the laws of mathematics refer to reality, they are not certain; as far as they are certain, they do not refer to reality." - Einstein

"What we observe is not nature itself, but nature exposed to our method of questioning." - Werner Heisenberg
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
WW,

Regardless if the metering circuit has emf induced into it or not the metering circuit either segments or sectors the loop (if on the inside) or becomes another loop in addition to the secondary loop(if on the outside).
I don't know what you mean by "inside" and "outside".

Quote
Yes, solenoid length will have pronounced effects and there is a point where an extended loop may see 'up' and 'down'.  I feel those issues are going beyond the initial experiment.
They go beyond the original experiment, yes. But the issue you have seems to be with the decoupled measurement technique I use.

Quote
I keep saying 'view from above' - 'in your mind's eye'. If nothing else, when you have the metering circuit crossing over the solenoid (decoupled, if you choose) stand on the table and look directly down at the top of the solenoid and loop. Forget the height differences for a moment and imagine the whole experiment is in 2D. The metering leads cut across in some fashion. Even though they are not the subject of induction they change the Gaussian surface area. They don't need to be on the same plane.
You really believe that a very slight angle (3º max) is going to skew the measurements and/or induced emfs that much?

You are saying that it is truly an amazing coincidence that the decoupled measurements of all the summed emfs equals exactly and oppositely the summed voltages across the resistors? Is it also an amazing coincidence that the decoupled measurements across the resistors equals exactly the measurements across the resistors taken in-plane? Is it yet another amazing coincidence that my simulation of this experiment also predicts exactly what I measure between all points on the loop using the decoupled measurement technique?

Quote
As you now see, the angle of the metering circuit does matter. Why? Because changing the angle changes the Gaussian surface area of both sections of the loop now sectioned by the metering leads.
I'm sorry, I don't see a significant effect on the experiment from the 3º angle at all.

Quote
This is why I say there is no way to connect a metering circuit without changing something in the experiment.
Then I challenge you to explain those amazing coincidences I outlined above. I say the detrimental effect (if any) is undetectable and/or insignificant to the results.

Does your above statement apply to ALL measurements? What is that change you speak of? How big is that change (i.e. quantify it)? Do the scopes and leads in Lewin's experiment invalidate his results too?

Quote
The voltages you are seeing are the voltage induced on one loop minus the voltage of the other loop (because the original loop is now segmented or sectored by the metering circuit.
Assuming you are referring to a decoupled measurement, what the probes are measuring is representative of precisely what voltage (both emfs and potential differences) difference there is between any two points on the loop. It exists whether the measurement probes are there or not.

Quote
The same voltages you are measuring have very little to do with the resistors.
I don't know what you mean here.

Quote
If the resistors were equal in resistance, even down to zero Ohms, and you measured from original points A to D you will see .5V. The reason, both sectors of the original loop now have 1V induced into each of them.... At the joining conductor (the metering circuit) what little voltage is seen is the result of two opposite polarity 1V induced into each segment.
Equal value resistors (reasonably larger R than Rwire) would not yield a decoupled measurement of 0.5V between D and A. My calculation indicates that 0V will be measured in such case. The reason is very simple; we know the total emf is 1V, we know that emf = PD, and we know that half the PD is across each resistor. So when you do your KVL around half the loop from point D to point A (same answer for BOTH directions), we have: -0.25V + 0.5V - 0.25V = 0V.

Now for the case with no resistors at all, i.e. a closed loop with a uniform wire, I'll let you guys ponder that one (as I have) as to what voltage would be measured half way across the loop. Show your work.  >:-)

Quote
The only reason .4V was seen between A-D was because the resistors skewed the angle of the metering leads. Your 'normal to the loop' metering circuit was normal to the loop only in a mechanical sense.
0.4V is seen between points D and A with the original value resistors, for the same reason we see 0V with equal value resistors. The KVL sum again works for BOTH directions. Starting from A going up the left side across the 100 Ohm we have: 0.25V -0.1V + 0.25V = 0.4V

Quote
Since my above description is exactly how angular position sensors work (as explained in the Maxim Application sheet brought up early in this thread) I really don't feel I need to argue the point.
I would caution against relying on MAXIM's theory in that paper of theirs, they made errors as well. For one eg. they claim that all the emf is induced in the resistors. Of course that is completely opposite to the truth. They don't fully understand the physics behind this apparatus even though they can get it to work.

Quote
It would be a great thing to share the understanding but that didn't work for me when it was 'shared'. So, I can't blame you if you choose another path.
The only path that interests me is the one that leads to the truth. I would hope that would be the case for all.  :)

I look forward to your responses.  :)
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
This post is a reminder to myself that I have a possible means to measure, non-intrusively, the actual induced emf in any of these loops. This could then be compared against the decoupled measurement technique to observe if there are any significant detrimental effects to the induced emf (or any parameter) by attaching the decoupled measurement leads to the loop at any two locations.
   
Group: Guest
This post is a reminder to myself that I have a possible means to measure, non-intrusively, the actual induced emf in any of these loops. This could then be compared against the decoupled measurement technique to observe if there are any significant detrimental effects to the induced emf (or any parameter) by attaching the decoupled measurement leads to the loop at any two locations.


Basically you have a better technique that gives the same result as decoupled ? lol
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
Gibbs,

Only to determine the true emf.

Decoupled is still the way forward.  O0
   
Group: Guest
The induced electric field at a point outside the solenoid diminishes as a function of 1/r.


The problem is that an "induced electric field at a point" has no effect onto a test charge at this point. A single electron outside a solenoid carrying a varying current, feels no force. You must have a closed path.

I was the first surprised. Outside a long solenoid, B is zero. But the vector potential is not null and we have E=-∂A/∂t. So we would be expecting a test charge to be submitted to a force F=q*E=-q*∂A/∂t. But we don't observe such a force.

I have led several experiments some years ago in order to transmit an information signal with the vector potential only (E and B being null at the receiver location). Not one has worked, a loop is always needed.

Quote
The line integral of the E-field (which equals induced emf) at radius r does not diminish as a function of 1/r.

Better?  ;)

Yes, better  :).
Now we agree that the line integral of the E-field does not diminish as a function of 1/r, and if you agree that there is no effect of this e-field on charges outside a closed path, I don't see how you could experimentally verify the 1/r effect of the "not looped e-field".

   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
Now we agree that the line integral of the E-field does not diminish as a function of 1/r, and if you agree that there is no effect of this e-field on charges outside a closed path, I don't see how you could experimentally verify the 1/r effect of the "not looped e-field".

I don't know if I agree that the induced E field has no effect on a point charge. If it did not, induction would not work.

As far as verifying experimentally the 1/r effect at a point, I don't need to, do I?
   

Group: Administrator
Hero Member
*****

Posts: 2906
It's not as complicated as it may seem...
Now we agree that the line integral of the E-field does not diminish as a function of 1/r, and if you agree that there is no effect of this e-field on charges outside a closed path, I don't see how you could experimentally verify the 1/r effect of the "not looped e-field".

Ex,

In regards to your statement in bold, I would suggest you read up on the betatron device. It uses an induced electric field to accelerate electrons in a circular orbit. I think we would agree that electrons are charges?

Also, from "Matter and Interactions Vol. 2", p.872:
Quote
No matter how an electric field is produced, it has the same effect on a charge q (that is F = qE)...
   
Pages: 1 ... 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 [18] 19 20 21 22 23 24 25
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2019-12-09, 18:12:36