Talk:Transformer/Archive 12

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 5

←

Archive 10

Dot polarity

There is some good work being done here, but the polarity section is still wrong. It is true that if the source is sinusoidal and operating in the transformer’s intended frequency range (and well below self-resonance) that the voltage polarity at the dots will be the same. But it does not have to be that way. For example, suppose a primary with 1 ohm of resistance is being held at -1 Vdc for a considerable period of time (let it be an air core so we won’t have saturation effects). Then rapidly change the voltage to -0.5 V and hold it there. There will be a positive voltage at the dot on the secondary although the primary voltage stays negative all the time. Now consider the current. Using the convention that current into the dot is positive, the current is -1amp. Change it in a positive direction to -0.5 amp and you will get a positive voltage at the dot on the secondary. The polarity of the voltage at the dot on the secondary tracks the derivative of the primary current. That is what inductors do. They change the derivative of current into a voltage. To get a voltage at the secondary, you have to have a changing flux within the path of the secondary and flux is proportional to current so a changing current is required for a secondary voltage. The polarities and current direction conventions are chosen so that when current directed into the dot on the primary changes in a positive direction (even if it is negative and stays negative) then there will be a positive voltage at the dot on the secondary (even if the voltage at the dot on the primary is negative and stays negative.)Constant314 (talk) 02:04, 14 February 2014 (UTC)

The dots are there for the convenience of people hooking wires to the terminals for practical use. The dots don't need to convey every nuance of Tesla's and Einstein's theories. --Wtshymanski (talk) 17:49, 14 February 2014 (UTC)

Your tag is rather bizarre and reverted, the dots are for practical phase identification in the intended application, a normal AC operation, not some hand picked case of DC minutiae. Kbrose (talk) 21:42, 15 February 2014 (UTC)

I agree that a Wikipedia article does not have to cover every possible case, but it should be correct for every case it does cover, which, as it is written at this moment, is all cases. All that is needed to make it correct is a statement as to which cases that the definition does cover. Transformers have a transmission zero at 0Hz and a low frequency pole usually formed by the copper resistance and the magnetizing inductance. They have one or more high frequency poles usually involving winding capacitance. The circuit may have poles and zeroes related to leakage inductance and load impedances. Your definition applies at frequencies between the low frequency pole and the first high frequency pole. In this range, the phase the voltage at the secondary dot closely tracks the phase of the voltage at the primary dot.

An entirely different definition which is independent of frequency is a mechanical definition which says that starting with the dot, the windings encircle the core in the same direction. In the most common case where one winding is wound on top of the other, with the winder turning the same direction for both coils, both dots represent the start of each winding or both represent the end of each winding. This case would include bifilar windings which are particularly easy to visualize. If the windings are on separate legs of the core as shown in the Induction law section, then it needs to be understood the direction of encirclement is defined by the right hand rule with the thumb pointing in the direction of an arbitrary reference line of flux linking the two windings.

Probably the best explanation would be to start with the mechanical definition and then explain what it means in terms of voltage phase in the transformer’s normal frequency operating range. You are a better writer than I am and can probably say this much more succinctly. Wtshymanski, who is a very good writer, could probably say all this in two sentences.

As for the purpose of the dots, there are others in addition to helping the installer connecting it correctly. Probably the first purpose was for the employee at the manufacturing plant. The dot indicated which terminal was connected to the start of the winding. The dots are also used by circuit designers, especially designers of switched mode power supplies. Some topologies, such as fly-back, simply will not work if the secondary wires are swapped. These dots appear in the specification sheet and not on the physical part. Constant314 (talk) 16:51, 16 February 2014 (UTC)

Seeing no support for my point of view, I accept the Polarity section as written at this time as the new consensuses.Constant314 (talk) 17:08, 23 February 2014 (UTC)

Induction Law

There is an anachronism here that I think it is time to fix, and that is in the interpretation of Faraday’s law of induction. The equation says that the EMF induced in a winding is proportional to the rate of change on the B flux enclosed in the winding. It does not say that the changing B flux causes the EMF. If the B flux in the core could do this, it would be acting at a distance. The B field at a point can be said to cause a force on a moving charged particle at the same point, but there is nothing in the definition of the B field that can allow it to reach out from the middle of the core to create an EMF in a wire that encircles the core.

Faraday’s law merely requires that whatever causes B and the EMF must create them in such a way that they obey Faraday’s law. The main actor is A, the magnetic vector potential. Currents (presumably) in the primary cause the A field. The A field in the core causes B field there as detailed in the equation B = curl A and the A field at the wire creates the electric field there as given by E = -dA/dt + other terms.

But no matter how it works, it is still induction. Constant314 (talk) 23:16, 19 February 2014 (UTC)

Energy is transmitted from the primary circuit to the secondary circuit in accordance with the Poynting vector rules. Induction is a different phenomenon.--109.144.196.1 (talk) 01:28, 20 February 2014 (UTC)

Yes, Poynting applies. What do you think induction is? Constant314 (talk) 01:44, 20 February 2014 (UTC)

e=-Ldi/dt109.144.196.1 (talk) 01:54, 20 February 2014 (UTC)

why do you think that this does not apply to a transformer?Constant314 (talk) 02:01, 20 February 2014 (UTC)

A transformer is a constant flux device 109.144.196.1 (talk) 02:17, 20 February 2014 (UTC)

constant flux device is not an existing page. I cannot imagine anyway in which a transformer is a constant flux device. Constant314 (talk) 02:26, 20 February 2014 (UTC)

Ideal transformer when loaded by resistance R<< wL2. I1N1=I2N2 so net MMF is zero. Net flux is zero. Flux is constant at zero.--86.160.104.86 (talk) 16:31, 20 February 2014 (UTC)

I see that you are making inferences from the ideal transformer. There are no ideal transformers. The ideal transformer is a mathematical model intended for explaining what a transformer does without the complications of winding resistance, leakage inductance, magnetizing inductance and internal losses. It is only valid from an external point of view. It is not valid for inferring how a transformer works internally. In particular, it is not valid for drawing any conclusions about the internal flux. Saying that the permeability is infinite and the flux linkage is 100% is a rationalization that describes how you would have to change a real transformer to make it at like an ideal transformer. If you start with a large but finite permeability you will see that the primary amp-turns is larger than the secondary amp-turns by an amount that is inversely proportionate to the permeability. As you assume larger and larger permeability the difference gets smaller and smaller, but it is still proportional to the reciprocal of the permeability. If you multiply that small difference by the large permeability you will see that the permeability cancels and what is left is just exactly what is needed to give you the flux that satisfies Faraday’s law. Constant314 (talk) 22:51, 20 February 2014 (UTC)

The important difference between an ideal and a practical transformer as regards core flux is that a real trans former has (sec)load independent magnetising flux. Primary created flux and secondary flux are equal and opposite nad therfore cancel in the core/ This is independent of core permeablility.--86.160.104.86 (talk) 16:30, 24 February 2014 (UTC)

You are on the right track. It is necessary to consider phase. Start with the unloaded secondary. Just to be clear, assume no leakage inductance, no copper resistance and no core loss but having a core with finite permeability, $μ$ . So we have a magnetizing inductance L that is proportional to $μ$ and the square of the primary turns N_P. The magnetizing current is proportional to the integral of the primary voltage and inversely proportional to L. The magnetizing current lags the primary voltage because the inductor current is the integral of the primary voltage. Core flux is proportional to current, N_P and $μ$ . One factor of N_P and $μ$ cancel, leaving the core flux independent of $μ$ , lagging the primary voltage and inversely proportional to N_P. Secondary voltage is equal to the derivative of flux times the number of secondary turns, N_S. Because secondary voltage is equal to the derivative of the flux, it leads the flux by 90 degrees. The flux lag the primary voltage by 90 degrees, so secondary voltage is in phase with the primary voltage and proportional to N_S / N_P. So to recap: primary current and core flux lags the primary voltage by 90 degrees and the secondary voltage leads the flux by 90 degrees with the net effect that the secondary voltage is in phase with the primary voltage.

Now, connect a resistive load to the secondary. In-phase current flows in the secondary. The in-phase amp-turns of the secondary is matched by an in-phase component in the primary amp-turns. As you surmise, there is no in-phase component in the core flux. But, that is OK because the out of phase core flux is exactly what it needs to be so that the secondary voltage is equal to the derivative of the core flux times N_S. And nothing changes as $μ$ increases except the magnetizing current decreases. The core flux stays the same and stays out of phase with the secondary voltage, as it must, if it is to satisfy Faraday’s law. And the whole process is called inductive coupling. Constant314 (talk) 00:26, 25 February 2014 (UTC)

So what is the resultant value of the out of phase flux in the core? The magnetising flux?--86.160.104.86 (talk) 17:07, 27 February 2014 (UTC)

V_P / ( s N_P ) where V_P is a phasor. Constant314 (talk) 01:51, 28 February 2014 (UTC)

So the miniscule out of phase flux in the core (ie the only flux in the core) is responsible for the operation of all transformers? If so, then an ideal transformer would not work as it has no magnetising flux. The answer to this problem is that transformers operate as described in the ideal transformer section of the article. Induction only occurs as an artifact of non ideal operation bu is not essential to it. I feel far too much emphasis is being put on in duction as the operating principle for all transformers when it is but only a side effect of non ideal operation. The article is misleading because of this. --86.160.104.86 (talk) 14:04, 2 March 2014 (UTC)

1. If the magnetic core did not have anything to do with the coupling, then the transformer would work the same without the core; but it will not.

2. The magnetizing flux is not miniscule. It is V_P / ( s N_P ) regardless of the core permeability.

3. Ideal transformers do not exist, but if they did they would have the same magnetizing flux.

4. All transformers operate in a way that is consistent with Faraday’s law of induction. Ideal transformers do not exist, but if they did they would also operate in a way that is consistent with Faraday’s law of induction. In all cases there would be no induced voltage in the secondary if there were no magnetizing flux. This would be true even if the magnetizing flux did not cause the secondary voltage.

5. You are trying to make inferences about real transformers from an ideal model. That is not valid reasoning. The only valid inferences you can make are about the model. You might, for example, infer that the model was not valid because it implied a result different from reality.

6. Your inference that there is no flux in the ideal model is incorrect, probably as a consequence of not properly applying the theory of limits.

7. Even if your inference were correct, if posted it would likely be quickly removed unless you provide a reliable source saying that the inference was correct. See the Wikipedia policy on synthesis.

8. A changing current driven into the primary induces a voltage in the secondary. This is called induction regardless of the mechanism. It would be induction even if the cause were the vector potential instead of the magnetizing flux.

9. The article can probably be improved. It is not likely that we could productively discuss that if you keep posting that a transformer does not work by induction. There are literally 100’s of reliable sources that say it does. Constant314 (talk) 19:45, 2 March 2014 (UTC)

Transformer operation

I have started a new section to discuss this.

Firstly in any discussion, I wish to make a distinction between transformer operation and induction coil devices. In the latter, induction does indeed take place because the secondary is very lightly loaded and no appreciable energy transfer occurs. Faradays law of induction and Lenzs law apply. Transformers with appreciable secondary loading do not rely on induction.

Faraday's law always applies. It cannot be turned off. It applies to transformers, inductors, and loop antennas. There is only one theory of electromagnetics and it applies in all cases. Constant314 (talk) 00:37, 6 March 2014 (UTC)

Not all transformers need a core. Transformers constructed for microwave frequencies (using microstrip for instance) do not need cores. At lower frequencies, it can be seen that the purpose of the core is to direct the H field from the primary to the secondary. In conjunction with the E field, energy is transmitted from primary to secondary by virtue of the Poynting vector.

Electromagnetic energy flow is always accompanied by a net positive average Poynting Vector. This true for microwave transformers, low frequency transformers, radio antennas, transmission lines, toaster ovens, traction motors, induction heaters and loud speakers. As before there is only one theory of electromagnetics and it applies in all cases. Yes the purpose of the core is to guide, enhance, concentrate the fields. It does not matter. There is still flux in the secondary look and the induced voltage will obey Faraday's law. It is still called inductive coupling. Constant314 (talk) 00:37, 6 March 2014 (UTC)

It can be deduced from the two alternative expressions for back emf ( ie e= - Ldi/dt, and e= -NdΦ/dt), that the effective inductance, L, of a solenoidal coil is N^²/R where R is the reluctance of the magnetic circuit on which the coil is wound. Reluctance R = l/μA, where l is the length of the magnetic circuit and A is its cross sectional area. I am assuming uniform csa and a single solenoidal coil.

So, in the case of a transformer with an o/c secondary, the primary current will be

V_p/jωL. Substituting for L, I_pri= V_pR/jωN². In an ideal transformer contrary to your statement about there still being magnetising current, we see that: If μ → ∝, R→ 0 and I_pri → 0. --86.160.104.86 (talk) 18:31, 5 March 2014 (UTC)

My statement is that there is non-zero magnetizing flux in the ideal transformer. Yes, the magnetizing current goes to zero in the ideal transformer. That is one of the ideal things about an ideal transformer. Constant314 (talk) 00:37, 6 March 2014 (UTC)

Ok. Just realised that the reluctance cancels in the equation for flux and that you are in fact correct as to the existence of core flux without any current. This is a result i had not come across before and is quite surprising to me. Yet this flux is out of phase with the primary and secondary voltages and does not contribute in any way to the energy flow from primary to secondary. So are you still saying that this flux is essential to transformer operation?--86.179.250.140 (talk) 16:42, 6 March 2014 (UTC)

BTW I have learned something from you today. Thank you!--86.179.250.140 (talk) 16:45, 6 March 2014 (UTC)

The answer is complicated. The changing flux encircled by the secondary is unavoidable. That is clear from Faraday’s law. However, the flux encircled does not cause the secondary voltage because if it did, it would be acting at a distance, which it does not do. The secondary voltage is caused by the magnetic vector potential, the A field. More specifically E = -dA/dt. The reason that the flux is unavoidable is that it is also caused by the A field, specifically B = curl (A). Constant314 (talk) 00:19, 7 March 2014 (UTC)

Flux in the core does not generate EMF in secondary

It has been known for almost a century that the flux in the core does not cause the EMF in the secondary, because if it did it would be acting at a distance. It is true that the EMF in a closed path is proportional to the time rate of change of the flux contained in the path. But that equation does not describe a cause and effect relationship. The relationship is effect and effect with the cause of both being A, the magnetic vector potential. Unfortunately, the vector potential is not well known, so the term “magnetic field” is sufficiently general to include A, B and H. Of course, it is not the magnetic field in the core that causes the EMF; it is the magnetic field at the wire which is causes the E field at the wire that causes the EMF. Or more specifically:

E_WIRE = -d A_WIRE / dt + other terms.

Constant314 (talk) 00:21, 20 March 2014 (UTC)

Merge from

Compensation winding is a single sentence that could be merged here, though it needs a reference first. --Wtshymanski (talk) 19:23, 24 March 2014 (UTC)

That statement is really not accurate or complete as it stands, the most common place to find them is in motors. Kbrose (talk) 20:12, 24 March 2014 (UTC)

Which statement? My statement above, or the single line at Compensation winding? I can't find a reference for this "fractional turns" notion at the subject article. We already have many articles on motor field windings, at least one of which talks (or should talk) about compensatING windings. Maybe "compensation winding" needs to go away instead? It's too specific and would need context to decide which usage was appropriate. --Wtshymanski (talk) 20:22, 24 March 2014 (UTC)

Seems to me the best place right now would be to redirect the term to universal motor and discard the page content for lack of definitive references. Kbrose (talk) 02:02, 25 March 2014 (UTC)

Separate Transformer History article

I suggest that Transformer article's History section become a separate Transformer History article.Cblambert (talk) 05:26, 25 May 2014 (UTC)

Perhaps you might offer some reasons. The article is long, but just pulling part of it out to make the article shorter, if that is the reason, to me, does not make sense. It is OK for articles to be long.Constant314 (talk) 00:43, 27 May 2014 (UTC)

I have no strong feeling about need to split in two articles. I raise the issue because I think people may tend to intuitively prefer History at article's front end, which is not evidently necessarily felt to be optimal in Transformer article case. Hence, for example, desire to put recent new Invention section at front instead of more logical inclusion as sub-section of current History section at the end. It may be that in popular long articles such as that of Transformer, editors tend to neglect back end sections but History was in my view thoroughly fine-tuned before move to the back end. In absence of clear consensus, I say leave History as in at the end.Cblambert (talk) 03:02, 27 May 2014 (UTC)

agree.Constant314 (talk) 03:49, 27 May 2014 (UTC)

Cooling section - fast depressurization

I have delete the last 2 sentences of 2nd paragraph of Cooling section that were reading

"Another protection means consists in fast depressurization systems which are activated by the first dynamic pressure peak of the shock wave, avoiding transformer explosion before static pressure increases. Many explosions are reported to have been avoided thanks to this technology.[69]

for the following reasons:

The Transformer Protector (TP) fast direct tank depressurization system involved is proprietary technology recently developed by Transformer Protector Inc. that is very narrowly focused on very large power transformers.
The first link was a dead link and points to TP authors
The second link was an improperly formatted a TP news release
A search of IEEE Xplore 3.7+ million documents shows only one document by TP authors (doi:10.1109/PES.2008.4596521) directly related to for fast depressurization.
Assertion that 'Many explosions are reported to have been avoided thanks . . .' (TP website says 7 since May 2013 is subjected) and could be construed as advertising.

I consider any future inclusion of TP technology subject to consensus rules in this Talk section Cblambert (talk) 20:56, 27 May 2014 (UTC)

Polarity

I support the roll back of the dot convention definition because it is correct in all circumstance. I think it would be helpful to say that when a voltage transformer is operated with sinusoidal voltages in its normal frequency range and power level that the voltage polarity at the output dot is the same (plus minus a few degrees) as the voltage polarity at the input dot.Constant314 (talk) 01:28, 27 May 2014 (UTC)

I propose adding the following additional note in the Notes section along with current reference and note the end of the 2nd sentence:

"When a voltage transformer is operated with sinusoidal voltages in its normal frequency range and power level the voltage polarity at the output dot is the same (plus minus a few degrees) as the voltage polarity at the input dot."Cblambert (talk) 02:36, 27 May 2014 (UTC)

OK with me.Constant314 (talk) 02:49, 27 May 2014 (UTC)

On reflection, one of the Polarity section's 2 current notes reads,

"ANSI/IEEE C57.13, ANS Requirements for Instrument Transformers, defines polarity as the 'designation of the relative instantaneous directions of the currents entering the primary terminals and leaving the secondary terminals during most of each half cycle', the word 'instantaneous' differentiating from say phasor current."

which I think covers adequately your initial suggest above here.

Transformers are inherently for sinusoidal service so there is no need to belabour the matter beyond ANSI/IEEE C57.13 or similar terms. It is in my view beyond Transformer article's scope to provide treatment of polarity extending too far into wiring and transformer connection details, which might however be proper to get into in separate article, including possibly much-improved Dot convention article.Cblambert (talk) 05:31, 27 May 2014 (UTC)

It would be good to have the IEEE definition in the article, bu I don't think can use their definition verbatim due to copyright reasons. Constant314 (talk) 11:00, 27 May 2014 (UTC)

I've changed Polarity sub-section C57.13 note to paraphrase instead of verbatim quote and added complete reference to that no.Cblambert (talk) 19:10, 27 May 2014 (UTC)

The note "When a voltage transformer is operated with sinusoidal voltages in its normal frequency range and power level the voltage polarity at the output dot is the same (plus minus a few degrees) as the voltage polarity at the input dot." seems to have gotten lost.Constant314 (talk) 22:22, 27 May 2014 (UTC)

I thought C57.13 note would be sufficient without this addition note but I will add willy nilly.Cblambert (talk) 23:08, 27 May 2014 (UTC)

It needs to be simple. I think in the old days they told the operator of the coil winder:" connect the wire to terminal with the dot and wind it (the one direction that he machine would wind)." That is the way I think of the dots.Constant314 (talk) 02:21, 28 May 2014 (UTC)

Actually, think of it this way: if you don't put in the simple explanation, somebody will come along later and replace your correct definition with this usually correct definition because it is simpler.Constant314 (talk) 02:26, 28 May 2014 (UTC)

Improving the introduction

I think the entire third paragraph can be removed as it is completely redundant with the explanation further down.Constant314 (talk) 23:24, 31 May 2014 (UTC)

Tight coupling. I know what it means, you know what it means, but it is jargon. Let's find a better way to say that.Constant314 (talk) 04:28, 1 June 2014 (UTC)

Incorrect inference in the introduction

"A varying current in the transformer's primary winding creates a varying magnetic flux in the core and thus a varying magnetic flux through the secondary winding. This varying magnetic flux induces a varying electromotive force (emf) or voltage in the secondary winding." This is a convenient fiction that has been known to be incorrect since at least the mid 20’th century. The primary current creates a vector magnetic potential field (the A field) which creates a varying magnetic flux in the core and a varying electromotive force (emf) or voltage in the secondary winding. You can compute the secondary voltage from the flux it encircles, but if the flux so encircle caused the emf, it would be acting at a distance, which it does not do. The reason you can compute the emf from the flux encircled is that they both have the same cause which is the A field.Constant314 (talk) 03:15, 2 June 2014 (UTC)

What to you suggest instead? Change it to suit. I will think on this separately.Cblambert (talk) 06:38, 2 June 2014 (UTC)

I would simply refer to the "magnetic field" which is broad enough to cover A, B and H. Primary current causes a varying magnetic field adjacent to and a least a skin depth within the conductors of the secondary which causes an emf. I would rather tell the whole story, but I cannot put my hands on a credible reference. The best I can do is cite Feynman's statements on the vector potential and the Aharanov-Bohm experiment. But it would require WP:SYN. It is in the crack between physics and engineering; its too simple to be interesting for one and too complex to be interesting for the other. Of course when I design transformers or visualize them I think about the flux that links the primary to the secondary. It is a useful fiction. The flux in the core is like the odometer in a car. You can differentiate the number displayed to obtain the speed of the car, but we know that the odometer doesn't may the car move.Constant314 (talk) 01:48, 3 June 2014 (UTC)

Lead section is meant to be a summary, such that any lingering A, B and H issue should in my view be dealt first with in the body of the article. But go ahead and edit to suit as you see fit.Cblambert (talk) 13:32, 3 June 2014 (UTC)

done. I've made the minimum change.Constant314 (talk) 00:00, 4 June 2014 (UTC)

Do you agree with following in Induction Law section?

The voltage induced across the secondary coil may be calculated from Faraday's law of induction, which states that:

V_{\text{S}}=N_{\text{S}}{\frac {\mathrm {d} \Phi }{\mathrm {d} t}}

where V_s is the instantaneous voltage, N_s is the number of turns in the secondary coil, and dΦ/dt is the [[derivative of the magnetic flux Φ through one turn of the coil.Cblambert (talk) 06:55, 2 June 2014 (UTC)

yes.Constant314 (talk) 01:25, 3 June 2014 (UTC)