Talk:Transformer/Archive 3

Move material to Autotransformer?
Does anyone agree that some of the material in this section would be better off now being moved to the Autotransformer article? This section is starting to have more coverage than that article. BillC 17:41, 15 January 2006 (UTC)


 * I didn't previously notice that there was an autotransformer article. Nearly all or perhaps all of the variac stuff could go there. --C J Cowie 17:59, 15 January 2006 (UTC)


 * Yes.if we leave a short description on the transformer page, and then direct readers to the main page
 * Autotransformer I think that would be satisfactory. Dont forget tho' that there are things such as fixed autotransformers (not variable).--Light current 18:28, 15 January 2006 (UTC)

I moved some of it across, though the formatting looks a little ugly right now. BillC 19:52, 15 January 2006 (UTC)


 * I think the variac picture can go also. --C J Cowie 19:55, 15 January 2006 (UTC)

Introduction
I have no doubt that high voltage transmission development continues, and I suspect that transformer development continues in that area, but I don't think the sentence that has been questioned fits very well or contributes much to the introduction. I think that I will offer an alternative. --C J Cowie 19:20, 16 January 2006 (UTC)


 * The article use to say that transformers were essential in the continuing development of the high voltage network. I dont think there is much development going on now in high voltage networks, and that that is, is not depending on transformers, but on advanced forms of switchgear automation, telecontrol, condition monitoring of plant etc.--Light current 23:19, 16 January 2006 (UTC)


 * Please sign comments. You are misinformed, high-voltage networks are ongoing topics of research, construction continues to expand high-voltage transmission, and transformers are critical elements in that expansion. Please keep up with the literature, (for example, publications of the IEEE Power Engineering Society, abstracts are available online) before making such a huge generalization.  --Wtshymanski 23:09, 16 January 2006 (UTC)

So how are transformers being used now in different ways from what they used to be? Are higher voltage types continuing to be developed?--Light current 23:20, 16 January 2006 (UTC)


 * I am not sure that you are asking about what I wrote in the intro or what Wtshymanski wrote above. What I wrote refers to "tricks" with transformers to reduce harmonics introduced in power systems by nonlinear electronic circuits and in motors by electronic motor controls. I suspect there are other examples. --C J Cowie 23:49, 16 January 2006 (UTC)

No its OK CJ, I was asking Wtshymanski about the continuing (or new) role of transformers in HV network development.--Light current 00:52, 17 January 2006 (UTC) Actually I have no problems with this part of the article: ''Its basic design, materials, and principles have changed little over the last one hundred years, yet transformer designs and materials continue to be improved. Transformers are essential in high voltage power transmission providing an economical means of transmitting power over large distances.'' --Light current 23:23, 16 January 2006 (UTC)
 * The article used to say "the transformer has played a substantial role in the development of electrical networks". --BillC 22:23, 16 January 2006 (UTC)


 * Indiana REMC Powering Customers With Soy and Envirotemp® FR3™ Fire-Resistant Fluid are a couple of examples of recent improvements to transformer oil. Maybe it would be hard to get more efficient, but environmental issues, safety and reduced costs are always being researched. (not that it needs to be in the introduction) --Dual Freq 22:32, 16 January 2006 (UTC)


 * I was actually thinking of this revision when I said it was not accurate.--Light current 22:59, 16 January 2006 (UTC)

A significant point now seems lost from the article. One can either generate electricity close to its point of demand, in which case you must transport fuel to the power station; or you can generate electricity close to its supply of fuel, in which case you must transport electricity to the point of demand. It is the fact that the transformer makes the latter significantly cheaper than the former that power stations are located by their fuel (or energy) sources, and will continue to be so. This is why the transformer has shaped the development of power grids: the grids owe their existence to it. BillC 23:41, 16 January 2006 (UTC)


 * This is agreed as a good point. I was not aware it had been removed. Can we put it back to wherever it came from.?--Light current 00:54, 17 January 2006 (UTC)


 * I removed it because a) it's not generally true - usually it's cheaper to bring coal to a power plant than to transmit power from a remote plant and b) the article is about transformers, not about electric power transmission, which has its own article and is a mouse-click away. Gotta make room for all the gearboxes. I'm also unhappy with the format of the "electrical laws" section - which I think is wordy, though at least accurate. --Wtshymanski 02:35, 17 January 2006 (UTC)


 * I agree with Wtshymanski. Power stations are located as close as practicable to points of demand considering such factors as the price and availability of large plots of real estate in urban areas, environmental concerns, and locations of rail lines etc. for coal access. High voltage transmission is essential, but the economic picture shouldn't be explained here. --C J Cowie 02:55, 17 January 2006 (UTC)


 * Not over here (UK) they're not. Theye located very near coal fields and sources of water. Nuclear ones almost always on the coast.--Light current 03:08, 17 January 2006 (UTC)

Transformer tap articles
I have proposed a merger of Tap (transformer) and Tap changer. Comments should be posted here: Talk:Tap (transformer). BillC 21:02, 17 January 2006 (UTC)

speaker transformers
http://en.wikipedia.org/w/index.php?title=Transformer&curid=30906&diff=36038301&oldid=36037749

i'm sure at least in PA type setups the tap changers on the transformers (which are usually driven from a 100V line source) are used to set the power of the individual speakers.... Plugwash 02:39, 21 January 2006 (UTC)


 * Well it may alter the power slightly but thats only because of impedance matching/ mismatching. On 100V line I think the o/p Z is normally higher than your home hi-fi for max efficiency and to avoid losses on the long leads that are used.--Light current 02:42, 21 January 2006 (UTC).


 * You are misinformed. I've usually seen it as 70 volt distribution systems. Consider two speakers on a PA system. You need a 50-watt speaker for the auditorium, and a 1 watt speaker for the janitor's office. Instead of calculating impedances all over the place and having a different impedance for every single speaker rating (and system power rating), you step up the output of the main PA to 70 volts (or 100 volts), and run that all over the building.  Each speaker, be it 4, 8 or 16 ohms impedance, gets its own transformer which will have taps on it, marked "10 watts", or "1 watt" or what have you. It alters the power delivered to the speaker a great deal, which is kind of the point. Do the math and you'll appreciate the ingenuity of the system.

--Wtshymanski 02:53, 21 January 2006 (UTC)


 * Well, if you've seen the actual markings on the transformers, I guess that settles it. But over here systems are normally 100V line and used outside in the open air with highish Z horn speakers for PA. Im not sure if transformers are actually used at the speaker end or just at the amp o/p.--Light current 03:00, 21 January 2006 (UTC)


 * See, for example, and about 2 million other Google hits on "70 Volt PA system".--Wtshymanski 03:08, 21 January 2006 (UTC)


 * Ah I see the speakers at that link have their own built in transformer (which puts the price up I suppose). Those speakers are different fron the ones I was thinking of.--Light current 03:12, 21 January 2006 (UTC)


 * Another thing to remember, not all amplifier output stages are designed for impedance matching as such. if your using a single transistor pulling against a resistor then sure matching may make sense, if you are using a push pull output stage with negative feedback taken from its output then the output impedance of your system can be very low indeed. but if you matched the load to that impedance you'd fry the transistors! Plugwash 03:28, 21 January 2006 (UTC)


 * It does appear that the same system is used over here also (but 100v line is more common here) for PA in hotels, large buildings etc. The system works by making the speaker look like lower or higher impedance depending on the power you want (or can handle) at each speaker. Saves using Hi Z exp horns!--Light current 03:32, 21 January 2006 (UTC)


 * Well i was originally assuming that the discussion was about domestic radio speakers fed from a valve amplifier in which case any tappings on the output transformer would be to match the Hi o/p Z of the valve stage to the much lower Z of the speaker.--Light current 03:38, 21 January 2006 (UTC)

Whats the good of flux?
In an ideal transformer, mag flux is not needed and is in fact equal to zero. Discuss in less than 100 words (10 marks)--Light current 01:15, 24 January 2006 (UTC)


 * I think you have confused flux with ampere-turns. The flux is the energy-coupling mechanism between primary and secondary. Without it, there is no transformer. An ideal transformer requires neglible ampere-turns to set up the flux, but it does not produce zero flux. Secondly, I have to agree with Wtshymanski, an image caption is not the place for point-making. That diagram is to show an ideal transformer, around which the arguments in the surrounding text develop. BillC 01:23, 24 January 2006 (UTC)

If that is the case, what would you measure (with a hall effect device, say) as the flux (above and beyond the magnetising flux) in an on load transformer core ? The flux may be the energy coupling mechanism, but it all canels out in the core. Otherwise the core would saturate- would it not?--Light current 01:36, 24 January 2006 (UTC)


 * Not necessarily. The ideal core material does not exhibit saturation. An ideal transformer has no leakage, no winding resistance and no magnetic losses. Loss-free does not demand zero reluctance in the magnetic path. BillC 01:43, 24 January 2006 (UTC)

But nevertheless, an ideal transformer has no residual flux in the core, not even mag flux. If it does have, how does it? Therefore the caption was wrong and thats why I changed it.--Light current 01:49, 24 January 2006 (UTC)

BTW I thought transformer cores were designed to have a very low reluctance!--Light current 02:02, 24 January 2006 (UTC)

Transformer diagram caption
Caption gives wrong impression that an ideal transformer needs flux. It doesnt!--Light current 01:19, 24 January 2006 (UTC)


 * Can you provide a reference that supports the statement that there is no flux inside an ideal transformer? You are confusing magnetomotive force in the magnetic circuit, which is affected by the presence of a current in the secondary, with flux, which is not. See, for example BillC 02:04, 24 January 2006 (UTC)

That there is zero flux in the core can easily be seen by considering the following:

H=NI, B=uH B is constant in the core.

So: Ip.Np = Is.Ns

The MMF due to the pri current is in opposition to the MMF of the sec current, and due to the above equation, they naturally cancel.

Alternative explanation:

the primary and secondary currents are flowing in opposite sense around the lines of B in the core and:

NpIp - NsIs = int(B.dl)/(u0.u) = phi.l/(u.u0. A)

Now, since u = inf in ideal trans, Ip.Np = Is.Ns

When the transformer is on load, the fluxes generated by the pri and sec cancel each other if u is infinite.

Reference: Electromagnetism, 2nd ed, IS Grant, WR Philips, pup John Wiley &Son, 1990 ISBN 0-471-92711-2 page 309 --Light current 02:33, 24 January 2006 (UTC)


 * Wtshymanski beat me to it. The argument presented there is invalid because it is dealing with a mathematical singularity: there is an inherent division by zero (the reluctance of the magnetic path) in that equation. The fact is that an ideal transformer with zero losses, perfect coupling and no reluctance is not merely physically unrealisable with practical materials, it is in fact mathematically impossible, implying as it does that there is an energy transfer from primary to secondary with no energy storage in the intervening medium and that there is a flux density B which exists without a flux Φ. The reluctance of an ideal transformer may tend towards zero, but it may not actually be zero. As a side issue, consider what the voltage induced on an open-circuit tertiary winding would be if the secondary is put on load. Zero, or non-zero? BillC 17:10, 24 January 2006 (UTC)

An o/c tertiary will have a voltage induced between its ends due to the magnetizing flux linking its turns, but of course, no current flows thro' this winding. Therfore the flux balance provided by the pri and sec currents is not upset! ;-)--Light current 06:05, 25 January 2006 (UTC)

I have changed the caption again to something less controversial but a bit more accurate. Comments?

BillC you asked for a reference. I provided one. Are you saying my reference is wrong? Can you provide a reference that says you are right? Are we going to have a reference war?--Light current 22:34, 24 January 2006 (UTC)

In an on load transformer, there is only magnetising flux in the core- right? Now in an ideal transformer ther is no magnetising flux because its ideal - right? I dont need to state the conclusion here do I? So what is your problem with my assertion? BTW whats your defn of an Ideal transformer?--Light current 22:52, 24 January 2006 (UTC)

In one of my books (Higher Electrical engineering by Sheperd Moreton and Spence (a standard work for UG level)) it clearly states that in an ideal transformer, the permeance is infinite. This must mean that the primary and sec flux must sum to zero.--Light current 20:01, 25 January 2006 (UTC)
 * But your book doesn't say the flux is zero, does it? If there was no flux, there'd be nothing linking the primary and secondary windings and no way to transfer energy. What your book *probably* says is that if the permeability of the core material is infinite, it requires negligable *current* to establish the flux. You must have flux - by definition of a transformer. Do a Google search on "ideal transformer infinite permeability" and you'll soon hit about a thousand sets of lecture notes that discuss this. Primary and secondary MMF (magnemotive force, ampere-turns) must sum to near zero in the ideal case - an ampere-turn in the secondary must be offset by an ampere-turn in the primary. Your reference may not be wrong, but you may be reading it in an entirely idiosyncratic way not accepted by the rest of the world. --Wtshymanski 22:54, 25 January 2006 (UTC)

OK Bill (WT) I think I can see where you are coming from. Now I suppose its up to me to prove that there is no resultant flux in an on load ideal transformer core.

But in the mean time, think about an ideal op amp. There is actually no input to the amplifier (inf gain), yet there is an output and an input signal. The output voltage balances the input voltage to give zero at the op amp -ve i/p. Is this not a good analogy?--Light current 17:04, 26 January 2006 (UTC)
 * Here's an analogy for you. A transformer without flux is like a gearbox without teeth (though I can think of at least one example of a toothless "gear" box so the analogy isn't perfect). --Wtshymanski 18:57, 26 January 2006 (UTC)

So you think its the teeth that transfer the torque from input to output? What happens if the teeth are infinitely small? --Light current 23:12, 26 January 2006 (UTC)


 * I think you're probably right, LC, but please leave the caption alone. Every physically realizable transformer has flux in the core.  That's how they work.  For people trying to find basic information on transformers, it would be misleading to imply that there is no flux.  The image in question is idealized, even if it's not actually an ideal transformer, because only the flux in the core is shown.  Ideal transformers can pass DC, too, and do lots of other weird things that the article says they can't, but talking about this here would jsut be confusing.  Pfalstad 01:50, 27 January 2006 (UTC)

We must certainly be careful not to confuse innocent people. But we must also be careful not to mislead them. We are trying to find a proper description of the operation of a transformer. Certainly flux plays a part, but not the large part that most people think it does. (See my op amp example above.) There has to be the possibility of flux (ie a low reluctance core,) so that flux can be cancelled. But how to explain this simply is the difficulty! We could say the primary tends to create flux, whilst the sec tends to cancel it. If there was resultant flux in the core you would have an inductance in the equ cct. This is where the diagram misleads. Thats why I renamed the flux to magnetising flux which is present in all non ideal transformers. THe flux that you talk about in a physically realisable on load transformer is the magnetising flux.--Light current 02:06, 27 January 2006 (UTC)


 * Ok, well the new caption as it stands seems fine to me, except that you misspelled magnetizing. :) Pfalstad 02:11, 27 January 2006 (UTC)

I do believe that ther are two alternative spellings: one with a 'z' and one with a 's'. Im not fussy which we use as long as we're consistent--Light current 02:23, 27 January 2006 (UTC)

Flux in the core?
In an inductor, there certainly is flux in the core and you have to be careful that you dont saturate it. In a transformer, the secondary current creates an equal and opposite flux in the core to cancel the initial flux. Now if you want to say that there is flux generated by the primary but cant be detected because its been cancelled, this reminds me of our little discussion of counter propagating waves in TLs that cant be detected because they add to dc! However, if you have ever built a HF transformer you will know that two types of ferrite cores are avaialble from the manufacturers of such things. These are Inductor cores with a gap to avoid saturation; and transformer cores with no gap (because its not needed) and therefore offer larger inductance per (turn)^2. The core wont saturate becuase these no resultant flux if the transformer is on load.(unless the frequency is too low of course)--Light current 03:38, 27 January 2006 (UTC)

Yes! Just looked up my old college notes on transformers. There it says that an IDEAL trnasformer needs no magnetising flux, and NpIp = NsIs. In this case, it is blindingly obvious that phi=0.

Now, in most texts dealing with transformers, two opposing fluxes are shown in the core: one due to the pri, and one due to the sec current. This is what I think WE should show in our diagram to avoid confusion/further controversy. Any comments?--Light current 18:47, 27 January 2006 (UTC)


 * Re saturation: there is some flux in any real transformer. The core won't saturate because the resultant flux is too small to saturate the core.


 * I think the picture is an idealized view of a transformer, as opposed to a view of an ideal transformer, since it shows the flux only in the core. Showing a picture with no flux in the core would just be misleading.  So would omitting the word "idealized".


 * I don't have any good texts on transformers, but it seems that showing two opposing fluxes in the diagram would be a good idea. Pfalstad 19:42, 27 January 2006 (UTC)

Yes the only flux in a real transfomer is the magnetizing flux. This is equal to the primary generated flux minus the secondary generated flux. If the primary inductance is large enough, then the core wont saturate. This is how we (I) design transformers. You have the correct answer:

but it seems that showing two opposing fluxes in the diagram would be a good idea. --Light current 21:22, 27 January 2006 (UTC)
 * No flux implies no time rate of change of flux, implies no voltage on the winding. See Faraday's law in the article itself. It's needlessly complex to think of "opposing fluxes" - better to think of primary ampere-turns being opposed by secondary ampere-turns. Why invoke superposition especially in a canonical example of a non-linear system? --Wtshymanski 22:34, 27 January 2006 (UTC)

As long as you agree that pri At = sec At in an ideal tranformer. In which case there is vanishingly small flux! If you dont believe me, try to measure it in a real transformer--Light current 22:53, 27 January 2006 (UTC)


 * There's no point in arguing the finer points of what an ideal transformer does because no such thing exists, and we run into contradictions trying to describe it even in theory. If the coils were perfectly coupled, the second coil would do a perfect job of opposing any changes in the core flux, so how could there be any flux?  But if the flux is zero, it doesn't change, so how could there be any voltage across the coils?  Pfalstad 02:49, 28 January 2006 (UTC)

On the contrary, I believe there is always a point in arguing- as long as were getting somewhere- and we are(slowly). Initially I was arguing with BillC and WTshymanski. They were saying the caption only depicted and IDEAL transformer. If you look back up this page youll see the whole discourse!. Anyway are you now saying its the leakage flux that makes the transformer work? I dont think so.


 * No.

Now we are all starting to see what a beautifully simple, yet complex beast the transformer really is. As to the apparent paradox, its like many physical phenomena of action/reaction. Imagine pushing against a wall. You apply a force- the wall reists. Measure the resulatant force with a strain guage (or just obseve the wall doesnt accelerate). The resultant force is zero! Voila! --Light current 03:26, 28 January 2006 (UTC)

Relativistic transformers?
No, the paradox is due to our simplified model. In a real transformer, among other issues, the secondary does not instantaneously resist changes to the flux. There is speed of light delay. With a time-varying primary current, there must be some flux even if there is no leakage. Pfalstad 04:13, 28 January 2006 (UTC)


 * How much delay do you get when pushing a wall?--Light current 04:17, 28 January 2006 (UTC)

What's your point? Are you saying there is no delay? I dont' see the relevance of the wall example, which is also a simplified model anyway. Pfalstad 04:23, 28 January 2006 (UTC)


 * Im saying your delay thing is a bout as relevant to the delay in pushing a wall. Its irrelevant--Light current 04:24, 28 January 2006 (UTC)

In actual practice, no, but I find it useful for describing the action of a transformer in more detail than you seem to want to, and for resolving the paradox described above.. (I think it's funny that you have written pages trying to convince me that capacitors and inductors are transmission lines, but now you are saying that transformers are not! :) )  Pfalstad 04:43, 28 January 2006 (UTC)


 * Well they probably are TLs, but Im not going to get into that one just yet! (Just remebered - transmission line transformers are in fact TLs! How strange!)--Light current 23:01, 28 January 2006 (UTC)

Analogies:
Just found this analogy in your link!

''The step-up/step-down effect of coil turn ratios in a transformer is analogous to gear tooth ratios in mechanical gear systems, transforming values of speed and torque in much the same way:

So Im not the only one who thinks that gears are a good description of transformer action!--Light current 02:58, 24 January 2006 (UTC)

''

inductance
Although transformer windings usually heve considerable inductance (of the order of henries), only that portion due to leakage flux that does not couple both primary and secondary windings, contributes to the series impedance of the transformer.

however the rest would surely appear as paralell impedance which would still draw VArs out of the supply. Plugwash 02:47, 25 January 2006 (UTC)

Importance of High Inductance
I think it is necessary that a transformer have high inductance in order to work; the higher the inductance of the primary of a (lossless) transformer, the more ideal it is. An ideal transformer has infinite input impedance, and a large inductance approximates that. Pfalstad 03:07, 27 January 2006 (UTC)


 * Some high frequency (RF) transformers (like transmission line transformers) have very low winding inductances. It all depends how low in frequency you want to go!--Light current 03:13, 27 January 2006 (UTC)

Sure, "high inductance" is relative. At high frequencyes, transformers can be near-ideal with a smaller inductance. Pfalstad 03:35, 27 January 2006 (UTC)


 * Agreed--Light current 03:38, 27 January 2006 (UTC)

Picture: Direction of Current
In the color picture Image:Transformer3d col3.svg, the current in the secondary seems to be going in the wrong direction. The primary winding is going CCW around the core, producing a magnetic field going up, or clockwise around the core. The secondary winding is going clockwise around the core, producing a magnetic field going down, or clockwise around the core; this is supporting the flux generated by the primary, not opposing it. See. Pfalstad 02:37, 28 January 2006 (UTC)

This old black-and-white one is correct: Image:Transformer3d.svg. I replaced the color picture with the old one. Pfalstad 20:50, 28 January 2006 (UTC)


 * Actually, Paul, the other diag is more correct if the flux is shown as non zero (as it is). I dont know if thats what BillC intended!--Light current 20:57, 28 January 2006 (UTC)

Could you explain your reasoning, as I did above? Pfalstad 21:51, 28 January 2006 (UTC)


 * A transfomer is sometimes expalined as a pair of coupled coils which both magnetise the core in the same direction. In this case there is resultant flux in the core. The last diag showed both coils magnetising the core and a resultant flux . So BINGO! it was correct!--Light current 22:02, 28 January 2006 (UTC)

I get it. :) Well is the current picture accurate, except for the fact that it only shows the primary flux?  Pfalstad 23:58, 28 January 2006 (UTC)


 * Dunno. Ive forgotten my right hand rule. But I think the direction of primary flux in both diagrams was shown wrong (current should go CW around flux to screw along the flux). Sorry Paul, Ive only just noticed this otherwise I would have mentioned it!--Light current 00:04, 29 January 2006 (UTC)

I don't think so. Current should go CCW around the flux. See, and I have a similar diagram in my freshman physics book. Pfalstad 03:47, 29 January 2006 (UTC)
 * Yes but in that diagram, current is going CW around the flux!! I agree with that diagram--Light current 03:53, 29 January 2006 (UTC)

Are we looking at the same diagram? Tilt it 90 degrees so the flux in the middle is going up. The current is going counter-clockwise. Pfalstad 03:57, 29 January 2006 (UTC)


 * Im looking at the top diagram. Yes youve chosen just the wrong way to look at the diagram. If you tilt the diagram, you are not then looking at it from the end of the core which I think you must do.


 * However, If you look at the direction of the mag field with the current going away from you- it should be CW -it is! (ie mag field follows the direction of a right hand screw being screwed when you are looking from the screwdriver end).--Light current 22:54, 30 January 2006 (UTC)

I don't understand what you're saying. Pfalstad 00:00, 31 January 2006 (UTC)
 * If you take a straight piece of wire and hold it pointing away from you (perpr to the plane of your face) and the current is travelling away from you, then the magnetic field goes ClockWise around the wire. Thats all!--Light current 00:04, 31 January 2006 (UTC)

Are we talking about about the direction of the magnetic field or the direction of the current? Whatever. So when the wire is pointing away from you, the field goes clockwise around the wire. As it is in that diagram. And so it is in the transformer image. Look at the primary (ignore the secondary), in the upper left corner of the inside of the core; the right side of the top turn of the coil. The current is going away from us. The flux is going clockwise around it. Pfalstad 00:15, 31 January 2006 (UTC)
 * Yes it appears correct for the primary winding in the diagram in the article. THe secondary, with the direction of current shown will produce an opposing flux in the core. So the flux shown is the resultant flux which is the magnetising flux. I think this diagram should definitely be altered to show two fluxes then we would avoid confusion.--Light current 00:33, 31 January 2006 (UTC)


 * Looks like this was my mistake in the first place and you were correct. Sorry!!--Light current 00:37, 31 January 2006 (UTC)