Talk:Induction coil

Major rewrite required.
This is a classical textbook approach. Unfortunately the textbook was for high schools. Anybody who's looked at this with an oscilloscope knows it's a dramatic simplification. The coil and capacitor form a tuned circuit which rings like a bell, thereby providing a diminishing series of sparks with each opening of the points. I don't have time to fix this now. If nobody else does I'll come back when I can. Pawprintoz (talk) 23:23, 4 March 2012 (UTC)
 * Correct. The V2 in the diagram showing the primary current and secondary voltage is wrong.  The V2 output would be a decaying sine wave precisely because there is a capacitor across the interupter.  If there is a spark gap connected to the output, that will limit the voltage that actually appears across the gap.  The two effects together give the "diminishing series of sparks" refered to above.  The voltage that appears across the secondary coil itself will be higher due to the effects of the inductance of the conductors between the coils and the spark gap.  The technology is now old enough that good references to support any change to the article are hard to find.  I B Wright (talk) 16:18, 9 March 2014 (UTC)
 * Fixed it. There are actually plenty of references on Google Books. -- Chetvorno TALK 23:51, 19 May 2014 (UTC)

Dubious
I did some edits to fix some mechanical aspects about the armature. I put some dubious tags in for the electrical aspects, and I was tempted to add more. There are many things wrong. Putting a capacitor across the contacts is to slow the current interruption and prevent or minimize arcing at the contacts. Consequently, the capacitor prevents the voltage across the primary from going as high as it could, and that tends to limit the secondary voltage rather than increase it (the explanation is wrong). The max contact voltage would be limited by glow discharges to a few hundred volts. The arc discharge aspects are tricker. A voltage gradient of 0.5 MV/cm can start an arc discharge; that's what the capacitor stops: high voltages across the contacts at small contact separation. If the contacts do break down, then there can be an arc sustaining voltage of less than 20 V, and that will limit the primary voltage and hence the secondary voltage. Also, preventing energy loss in the primary circuit means that energy must be dissipated in the secondary. The diagrams don't seem to follow their cited reference (at least the one I checked). Glrx (talk) 19:57, 16 March 2016 (UTC)


 * The capacitor is to form a parallel tuned circuit with the primary inductance. It is true that it will prevent the back emf rising as far as it otherwise would, but its absence would only give a single spark of greater energy.  At the point od contact separation the capacitor, which is uncharged, is effectively a short circuit but a small spark still occurs at the interupter contacts as the current changes dirrection.  Not all the energy is disipated in the secondary.  The primary of the coil has resistance and this contributes to the dissipation of the energy (especially as in most coils not all the magnetic flux from the primary cuts the secondary windings due to the linear construction (magnetic circuit is incomplete).


 * A reference: a book showing a school boy how to build such a coil. The Boy Electrician, by J W Sims, "Every time the circuit is broken, a high voltage is induced in the secondary.  By means of the condenser connected across the interupter terminals, the induced emf in the primary forms a decaying sinusoid.".  He does not mention tuned circuit but that is what it is.  This description is essentially correct.


 * Interestingly before this fragment, he states, "Every time the circuit is made, the condensor causes the primary current to take a considerable fraction of time to grow".  But this cannot be correct as the capacitor (condensor in his parlance) is short circuited by the contacts at this point.-Elektrik Fanne (talk) 16:34, 17 March 2016 (UTC)


 * I've reverted your additions because they do not add sources.
 * Your first paragraph above is confused: at contact break, "a small spark still occurs at the interrupter contacts as the current changes direction." If the capacitor looks like short circuit, then the voltage across the contacts at separation is zero. There's not enough of an electric field to generate an arc. Also, if there were an arc, then it would consume all of the available current; consequently no current flows into the capacitor and the voltage across the capacitor/contacts/primary would not jump. A break arc is a performance killer to an induction coil.
 * There is a problem with using capacitor contact protection. After the contacts open, the capacitor stores energy as it charges to a high potential. When the interrupter finishes, the contacts approach each other. Now the electric field gets high enough to cause an arc, and the energy of the arc damages the contacts. See Noise Reduction Techniques in Electronic Systems, Henry Ott, 1976, John Wiley, pp 181-194.  At page 189, Ott explains a large capacitance across the contacts will supress the break arc.  Ott goes on to say, "The larger the value of the capacitor and the higher the supply voltage, the more damage the arc on "make" does, due to the increased energy stored in the capacitor."
 * The Boy Electrician is not a reliable source &mdash; especially if you condemn its teaching in the next paragraph.
 * I agree that an inductor, a capacitor, and internal resistance will form a damped sinusoid. The problem is that characterization ignores the arc in the secondary. When the arc ignites, energy is being stolen from the tank. That raises a lot of issues.  The interesting stuff in an induction coil happens during flyback &mdash; before the damped part of a damped sinusoid is manifest. The arc ignites at a high voltage, but the arc will continue to conduct until a much lower voltage. In a well-designed induction coil, what happens?  Does the initial strike take so much energy out of the system that the points will not arc again?  If that is the case, then there will be a damped sinusoid due to the primary, but the voltage level where it happens is uninteresting to the induction coil.
 * If we go to the source for a picture in the article, https://books.google.com/books?id=e-hMAAAAMAAJ&pg=PA3&hl=en#v=onepage&q&f=false, it does not show a damped sign wave. Neither does the similar illustration at http://www.alternative-technologies.org/articles/induction-coil/ which shows repetition.
 * The purpose of the capacitor is much more involved.
 * E. Taylor-Jones, The Theory of the Induction Coil, 1921, Pitman &amp; Sons, https://books.google.com/books?id=asxBAQAAIAAJ&pg=PR10&lpg=PR10&source=bl&hl=en&f=false
 * At page 17, under the Rayleigh theory, "the only use of the primary condensor is to check the formation of an arc at the interrupter". (It goes on to say that the condensor quickens the break, but that sentence can be misread.) (The page also describes some interesting experiments of breaking primary current without the usual condensor: for example, breaking the primary with a rifle bullet.)
 * Taylor-Jones has a transformer oscillator theory of the induction coil. See page 18; also fig. 8 on p 38. I get the sense that Taylor-Jones is usually describing systems without an arc because he's describing his theory. He offers oscillograms (of V2 rather than just V). He does, from time to time, specify how long an arc he can make with a setup. He pays attentino to loose coupling. I don't have time to digest it now.
 * Figs 2 and 3 of Anderegg & McEachron (1921), "An Oscillographic Study of an Induction Coil with High Frequency Load", Proceedings of the Indiana Academy of Science, Vol 32, pp 178-179, https://journals.iupui.edu/index.php/ias/article/view/13938/14023 are interesting but confusing; the HF load may confuse the issue. Some pics show lots of noise; others have parts with damped sinusoids.
 * Glrx (talk) 21:06, 17 March 2016 (UTC)


 * If I had an induction coil, I could create some oscillograms and I could guarantee that they would show a decaying sine wave across the primary as the contacts open. The secondary would also have a decaying sine wave voltage - assuming that the insulation of the coil does not break down.  Once a spark gap is connected, the voltage regulating effect of the spark modifies the voltage that appears across the secondary (and consequenty the primary).  Because of the negative dynamic resistance of the spark, the secondary voltage rises to the point where the spark gap breaks down, but the negative resistance causes the voltage to fall to a lower level where it is more or less maintained untill the voltage falls below the extinguishing point.  The voltage then follows the decaying sine wave as the spark gap no longer affects it.  The primary current (for which I cannot find an oscillogram) is a decaying sine wave from when the contacts open.  The current waveforms never mimick the voltage in these types of circuit.


 * Of your four references above, the first was written in 1922 and predates the invention of the oscilloscope, so any oscillograms that it contains are pure conjecture. Your second ref suffers from the same problem having been written in 1916.  Same for the third (1921).  I cannot open the Google books refs so they are unverifiable.  Your fourth reference contains poor quality illustrations - it's not clear whether they are annotated photographs or photographs of a blackboard.  In any case they clearly show a decaying sine wave, something that requires a tuned circuit.  My quote above from J W Simm's book was from the seventh edition published in 1955.  I also havethe fourth edition publishd in 1933.  Interestingly, this edition does not mention the decaying sine wave.  Was it that, the authors of these early works thought mentioning it was unnecessary or was it that the true circuit operation was unknown?


 * A petrol engine ignition system is, from a circuit viewpoint, identical to these induction coils - the only difference is that the interupter is moved to a separately driven contact breaker (something that was done for larger induction coils anyway). This reference contains the screen dump of a real oscilloscope conected to the coil.  It also clearly shows the decaying sine wave.  I would produce one myself, but I don't have a petrol engined car.  If there is a sine wave of any sort present, there has to be a parallel tuned circuit to create it and there is only one capacitor in the circuit so that is its function.  -Elektrik Fanne (talk) 11:52, 18 March 2016 (UTC)


 * If you made some observations on or some measurements on an induction coil, then that would be open to challenge as original research. There is no peer review. There's no guarantee that such an induction coil had been well designed.  I could guarantee a break arc by choosing a small capacitor, but that does not mean such a capacitance is a good design choice for an induction coil. The section above,, suggests there is a series of diminishing sparks. I believe such coils exist, but I suspect they are either misadjusted or misdesigned. If enough energy is left over after the first spark to create another one, then either the gap could be increased or the input energy could be decreased.


 * I have not disputed the presence of a decaying sine wave. There's going to be a little energy left in the system after the spark extinguishes, no energy is being added, so the system would show a decaying oscillation. The question is the relevance. Yes, the coil flyback is important, but that is before the decayed ringing is manifest. The energy loss in the arc is far more significant in the explanation. The ringing shown in the Pico ref figure 1.1 is at less than 1/36 of the initial stored energy (ignite at 12 kV, 3.5 kHz ring peak at 2 kV). No voltage ringing is present during the 1 ms duration spark. Your quest for a primary current diagram is apparently to show an alleged ringing at 3.5 kHz. To first order, such ringing would not be present: the spark has has shorted out the tank and killed the Q. (See also tightly coupled ignition coi issue below.)


 * A reliable reference about contact protection describes that capacitor-protected contacts incur damage on make rather than break. A reliable reference about induction coils that states a significant purpose of the capacitor is to check the arc at the interrupter. You have asserted that there is a break arc. What causes that arc to extinguish? If there's enough current to sustain a secondary arc, then there's more than enough current in the primary to sustain one. How much energy would that primary arc consume? Would it limit the primary voltage? A break arc is death.


 * "Pure conjecture"? You need a reliable source to make such a statement. Maybe oscillocopes were not available until 1921, but see Oscilloscope history and William Duddell. The given sources are not condemned. Here's an oscillograph from 1917: file:Oscillograph recorded on film.png; you can see the film's sprocket holes. Edwin H. Armstrong published some oscillograms in September 1915: "Some Recent Developments in the Audion Receiver", Proceedings of the IRE, vol 3, issue 9, pp 215-247; here's a reprint of Armstron's classic paper. Even blackboard drawings are reasonable oscillograms: Oscilloscope history.


 * It is pointless to rehabilitate The Boy Electrician when you don't even believe parts of what it says. The original author was Alfred Powell Morgan. His 1913 edition has your quotation above. The book describes construction, but it is is not a reasonable source for the theory of induction coils. There's no description of performance tradeoffs. Even the title tells us it and its later editions are not intended as professional-level or college-level work.


 * Comparing induction coils to modern automobile ignition coils is WP:SYN. Ignition coils use a different topology that includes a significant magnetic return path: "The primary winding is assembled around the outside of the secondary winding, and the laminated iron is distributed so that one portion serves as a core for the windings and the remainder as a shell around the entire subassembly." A return path increases the inductance and the coupling. Recent ignition coils use completely closed magnetic paths with magnet bias. Taylor-Jones reports that induction coils have poor coupling.


 * Glrx (talk) 19:56, 18 March 2016 (UTC)
 * Any images of oscillographs produced by anyone would be original research. However, it is original research that is allowed under Wikipedia's policies.  All images on Wikipedia are, by definition, original research and WP:OI specifically permits it.  There is no way anyone could argue the exception which is "illustrating unpublished ideas or arguments" given that we are discussing a museum piece well over a century old.


 * The primary current and voltage as well as the secondary voltage would be an exponentially decaying sine wave, if the secondary were either open circuit, or loaded with a resistor. The sine wave obviously decays faster if the secondary is loaded.  The spark gap, as noted above, is a negative resistance, and consequently holds the initial cycle of secondary current (and hence voltage) up for longer than would otherwise be the case.  But it cannot hold up the current in the primary.  This has to fall to zero as soon as the capacitor is charged (and then reverse as it starts to charge the primary inductance in the opposite direction).  The primary current will be an exponentially decaying sine wave regardless of the characteristic of the secondary load. The rate of decay is faster as energy is coupled out to the spark but the rate of decay falls once the spark extinguishes. The loose coupling of the primary and secondary allows the current waveforms to be different.


 * By the way, the modus operandi of a motor car ignition coil is exactly the same as the subject induction coil. Even as late as the 1960's they generally did not have closed magnetic circuits.  Most modern coils do have closed circuits (and some even have permanent magnets embedded).  But the primary and secondary cannot be tightly coupled.  One technique often used is to wind the secondary onto the magnetic core first, and place the primary on top, but with a space between the windings.  By this means the coupling can be controlled and designed for the greatest efficiency of spark.   85.255.232.25 (talk) 09:55, 20 March 2016 (UTC)


 * Note. has been blocked indefinitely as a sockpuppet. Glrx (talk) 20:30, 17 December 2016 (UTC)

Dubious figures
There are many problems with the two figures Primarily, the figures do not show the effects of breakdown. The comments are misleading.
 * File:Induction coil waveforms.svg
 * File:Induction coil waveforms with capacitor.svg

The first figure has an odd idealization. Apparently it wants a stray capacitance to explain a limiting peak voltage but then ignores the capacitance for the decay. The decay suggests an L/R with R being the winding resistance, but if the contacts are open, then I is zero so there's no dissipation. Toward the end, the inductor's VICs are violated because dI/dt = 0 but there's a voltage across the inductor. There's no indication of a breakdown. If we ignore leakage L, what happens is the flyback hits some peak voltage to cause arc over. The voltage then drops to a relatively constant sustaining voltage. With V now a constant, dI/dt is a constant, and the current decay has a constant slope (ignoring winding resistance). When the current falls below its sustaining value, the arc quenches. Now the residual energy is dissipated in the coil resistance as a high frequency damped oscillation.

The second figure is a coil with no breakdown. If there were a secondary arc, then the two figures should be similar. Primary resonance is different from secondary resonance; that is possible with poor coupling, but not needed for the present explanation. There's an ultra-fast reversal of i1 and rise of v2. The capacitor was supposed to slow that down. (Compare how slowly v2 reacts to primary step voltage.)

Glrx (talk) 00:53, 17 April 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 3 external links on Induction coil. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20161030062249/http://www.sentex.net/~mec1995/circ/hv/classic/classic.html to http://www.sentex.net/~mec1995/circ/hv/classic/classic.html
 * Added archive https://web.archive.org/web/20130225135733/http://www.nuim.ie/museum/ncallan.html to http://www.nuim.ie/museum/ncallan.html
 * Added archive https://web.archive.org/web/20110716172416/http://www.switchingpowermagazine.com/downloads/Oct%2001%20soft.pdf to http://www.switchingpowermagazine.com/downloads/Oct%2001%20soft.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 15:37, 13 November 2017 (UTC)

External links modified (January 2018)
Hello fellow Wikipedians,

I have just modified one external link on Induction coil. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20161009151329/http://www.worldphaco.net/uploads/CAPACITIVE_DISCHARGE_IGNITION_vs_MAGNETIC_DISCHARGE_IGNITION..pdf to http://www.worldphaco.net/uploads/CAPACITIVE_DISCHARGE_IGNITION_vs_MAGNETIC_DISCHARGE_IGNITION..pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 00:31, 22 January 2018 (UTC)

Spark gap v. Breakdown voltage
A simple test of linearity of a function is this: If (within a range) equal increments of its input value result in equal changes in the output or dependent value, then the function is linear within that range.

Within the range 4inches – 16inches, each 4inch increment in gap corresponds with a 40kV increase in breakdown voltage, so within this range the function is linear. The 1984 reference, which is widely available, has a graph on the page cited whose locus is clearly linear for the voltages listed. The correlation between these two references is surprisingly close in view of the variables which can affect the breakdown voltage for a given air gap: air pressure and humidity, size, shape and finish of the electrodes and even the material from which they are made.

References which support alternative views are always welcome but please don't blank out those which support a statement with which you don't agree. https://books.google.com/books?id=-i86AQAAMAAJ&pg=PA742&lpg=PA742&dq=%22induction+coil%22+%22spark+length%22+voltage&source=bl&ots=8rS3wA5Cp_&sig=ysVfTdZ8FiSyk-2crbh-4HIXog0&hl=en&sa=X&ved=2ahUKEwjWs9WUnInfAhWlGTQIHcTnA0EQ6AEwF3oECAUQAQ#v=onepage&q=%22induction%20coil%22%20%22spark%20length%22%20voltage&f=false (talk) 09:43, 8 December 2018 (UTC)


 * I agree. Here's a pretty authoritative source from the induction coil age who says that for metal balls at atmospheric pressure: voltage = 3000 × spark length in millimeters  (John Archibald Fleming (31 July 1903) "Hertzian Waves", Journal of the Society of the Arts, Vol. 51, No. 2645,  p.741).  He also says the voltage required to "pierce air" is about 30 kV per centimeter.  I think this source, along with your 1984 source, certainly support the article saying that the spark length is aproximately a "linear function' of voltage. --ChetvornoTALK 19:06, 8 December 2018 (UTC)

Agrees well with my youthful incomplete experimenents with a 25KV 40mA loose coupled neon sign transformer, a 1B3 high voltage rectifier tube from an old TV set, and a travelling arc built of coathanger wires.

Looks reasonable to me.

Resulting 60Hz AC travelling arc length was about 25mm at bottom arc ignition and about 60mm at top arc extinction.

I concluded that 60Hz AC arcs ignite at about 1KVRMS per millimeter, but won't sustain over 60mm due to 40mA loose coupled transformer inductive reactance current limiting.

Keep up y good work fellas!

🐱🐱🐱 FritzYCat (talk) 01:00, 7 March 2020 (UTC)