Talk:Ignition system

This article is rated B-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects:

Systems Mid‑importance This article is within the scope of WikiProject Systems, which collaborates on articles related to systems and systems science.SystemsWikipedia:WikiProject SystemsTemplate:WikiProject SystemsSystems Mid This article has been rated as Mid-importance on the project's importance scale. This article is within the field of Control theory. Automobiles Start‑class Mid‑importance This article is within the scope of WikiProject Automobiles, a collaborative effort to improve the coverage of automobiles on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.AutomobilesWikipedia:WikiProject AutomobilesTemplate:WikiProject AutomobilesAutomobile Start This article has been given a rating which conflicts with the project-independent quality rating in the banner shell. Please resolve this conflict if possible. Mid This article has been rated as Mid-importance on the project's importance scale. Technology This article is within the scope of WikiProject Technology, a collaborative effort to improve the coverage of technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.TechnologyWikipedia:WikiProject TechnologyTemplate:WikiProject TechnologyTechnology

I'm doing a report on cars and Wikipedia is just one of my sources. I was just reading the entry of ignition on HowStuffWorks.com and noticed that they describe "Solid State Ignition" as "distributorless". Is this Solid State Ignition the same as Electric Ignition? If so, it seems like Wikipedia has much more information than HowStuffWorks.

I think 'solid state' is a rather ambiguous term. Really it can be applied to any type of electronics not using valves (tubes), so the term 'solid state ignition' is really pretty meaningless. Electronic ignition can be distributorless or use a distributor. The latter type is more common, since it is really just an electronic version of the classic contact breaker ignition, using a single coil and distributor. The distributorless version replaces the distributor with low voltage electronic switching, and a small separate coil at each spark plug. This type is gaining favour as it is more reliable since there are fewer moving parts and no high voltage (HT) leads needed. Hope this helps - HowItWorks is a great resource but often they don't explore a particular topic in all its variations - to do so would probably confuse more than enlighten. Here at WP we have no such qualms ;-) Graham 21:48, 23 Dec 2004 (UTC)

Thanks Graham! The extra info clears it up for me as well as giving me more to write about. (Maybe you should add this more detailed info into the article.)

The current article only really talks about piston engines, but internal combustion engines as a group include rocket and jet engines. I'm thinking the article should be enlarged a bit to cover these too. The article already mentions aircraft engines a bit. Comments?WolfKeeper 23:13, 11 October 2006 (UTC)[reply]

I doubt whether there is any standard technology for igniting rocket or jet engines which burn their fuel continuously; in any case ignition is not needed during the normal operation of these engines. Ignition is only needed to get them started and many methods could be used: electrically heated wire, electric spark, mechanical spark (as in a cigarette lighter), electromechanical sparker (like a piezoelectric starter for a gas oven) or just tossing a lighted match.

Has anyone information on how jet aircraft engines are started (or restarted in flight)? 84.210.109.151 17:12, 4 November 2006 (UTC)[reply]

This type of arrangement was quickly superseded by spark ignition, attributed to Karl Benz, a system which is generally used to this day, albeit with sparks generated by more advanced circuitry.

The above seems in conflict with the invention by Charles Kettering. The mechanical-breaker and coil ignition system is known to me as "the Kettering" rather than "the Benz" system. Can anyone settle this issue by providing dates? 84.210.139.189 20:45, 12 November 2006 (UTC)[reply]

i would most apreciate if you could expand about the Distributorless ignition system (DIS) or create a new article

Looking at an engineering book from 1935, I found this odd passage:

"Chemical ignition hy injecting an active liquid capable of ignition on contact with one of the fuel components into the chamber, is doubtlessly the most convenient method of non-recurrent and recurrent ignition. Such liquids ere known (e.g, the solution of phosphorus in carbon dlsulfide) and they have been recommended many times for internal-combustion engines."

Any idea what he is talking about? DonPMitchell (talk) 08:57, 8 November 2008 (UTC)[reply]

Anyone have any knowledge or information on the Leburg electronic ignition system? There was a little section in the article that was removed by a unregistered user mentioning it was an advertisement(On May 3, 2009).

The section removed

More recently a retrofit Leburg electronic ignition system was made available for VW aero engines. This system does away with the distributor and other mechanical parts entirely. The variable ignition timing is based on the instantaneous RPM, which is measured with an electronic sensor.

The Leburg Electronic Ignition System is suitable for most aircraft engines to provide improved and consistent dual ignition to replace magnetos. For details see www.leburg.com ( Sept 17th 2009 )Henry Mickleburgh (talk) 16:24, 17 September 2009 (UTC)[reply]

--Rent A Troop (talk) 08:16, 18 May 2009 (UTC)[reply]

I found an Ignition system diagram on the Wikimedi Commons in French and arranged for it to translated to English.

• http://commons.wikimedia.org/wiki/File:Car_ignition_system.svg

deleted (moved to new section: Ballast resistor not shown in diagrams.Dcebr (talk) 03:28, 30 July 2014 (UTC)[reply]

Thank you 129.215.149.98 and PiRK

I will be integrating this graphic into the article soon, if no one beats me to it!--220.101.28.25 (talk) 15:27, 28 October 2009 (UTC)[reply]

We had an edit conflict there! (I think I nodded off over my keyboard, local time is very late!)

I have gone back to the 'original' diagram but full size because:

• It fits the text better, which still needs re-writing

• The text here is specifically about the mechanically ('points') switched system

• I had it translated specifically to use here
• It (or a version of it) may fit better under 'Electronic Ignition' --220.101.28.25 (talk) 20:27, 28 October 2009 (UTC)[reply]

In case it's needed again, may put this back in.

• Switch should probaly link to contact breaker/ points not switch
• Thyristor seems to be for an 'analog' electronic ignition system?

• Hall effect or optical 'points'??

• Needs a 'legend' of what component labels mean ie. Lp = Coil Primary Winding

--220.101.28.25 (talk) 21:16, 28 October 2009 (UTC)[reply]

I missed the little diagram ,. sorry . Do what you think is worthwhile,it all looks good,it is your ball .Wdl1961 (talk) 21:30, 28 October 2009 (UTC)[reply]

Ford is developing a system that uses LASERS in place of spark plugs to ignite the fuel.

I know Wiki is not a cystallball, but this is apparently a serious proposal. Researching reliable source/s required.

Refer to http://greenprophet.com/2009/07/14/10523/fords-laser-spark-plugs/ URL from Spark plug article

NB Spark plug article is longer than Ignition System article! --220.101.28.25 (talk) 23:47, 28 October 2009 (UTC)[reply]

After further consideration this has been removed.

For an ignition coil, one end of windings of both the primary and secondary are connected together. That's a quote from the Modern Ignition Systems section. I'm not seeing the connection point. I think this is shakey. Is there a clearer way to describe this?Longinus876 (talk) 01:47, 20 August 2011 (UTC)[reply]

I don't understand your point. They're connected. They're connected at the same point as the contacts and condenser. Andy Dingley (talk) 10:22, 20 August 2011 (UTC)[reply]

If the primary is wired in series with secondary, then the output voltages add and you get greater effective output. — Preceding unsigned comment added by 161.185.151.131 (talk) 08:32, 25 March 2015 (UTC)[reply]

As i read the articles "ingnition system" and "ignition coil" in the german and english Wikipedia, i noticed that most of the circuit diagrams (of classical ignition) do not match the "real thing". In fact I think only the one about "wasted spark" ist correct.

(1) The classic ingnition coil has only three connectors (numbered "15" - 12 V switched by car key from battery/generator, "1" via contact breaker to ground and finally "4" high voltage via spark plug to ground). And that implies that there ist no connection to GND, when the contact breaker is open. Therfore the only closed loop for the ingnition current (the current which "feeds the spark") is through the battery (its internal impedance is very small). Unfortunately no circuit diagram shows that. This circumstance is important, because alternators (the rectifier in it) will be destroyed by the high voltage which drives this (quite small) current of the spark, when the battery is disconnected in case the motor is running.

(2) Most of the classical ignition systems have two resistors of some importance: One with several kOhm in the high voltage loop, integrated either in the cable or in the sparkplug. This resistor has to effects: it makes the duration of the spark a little bit longer (but the spark a little "weaker") and it reduces interferences of the ignition system with radio frequencies. The other (second) resistor in series with the primary coil of the ignition coil, usually between +12 V an pin 15 of the coil (it has only about 0,5 to 1 Ohm). This resistor reduces the current when the ignition is on and the breaker is eventually closed (motor isn't runnig) and it improves the engine start, because the resistor is bypassed (by the starter switch) when the starter is operating and the voltage of the system drops due to high start current.

I did some rework of one of the diagrams which looks like that: http://www.brix.de/elektrik/_images/Batterie-Spulenzuendung_Schaltplan.jpg

Stefan Brix, sx@brix.de (sorry, currently no user-account in the english Wikipedia ...)

--77.187.23.139 (talk) 11:29, 30 March 2013 (UTC)[reply]

The purpose of the capacitor is merely to protect the breaker contacts from high voltage spikes induced in the primary winding of the coil. It does not form an oscillating lc circuit with the coil to store energy for the spark as stated in the article. The energy for the spark is stored in the magnetic field created in the coil windings by the battery current. — Preceding unsigned comment added by 186.247.70.97 (talk) 01:06, 6 May 2014 (UTC)[reply]

Ah yes, that will be because on a non runner after trying everything, you cranked it with the distributor cap off, saw the spark, changed the condenser, et voila. Now, why did it not run at all in this state? And with a good condenser, why is there still a tiny spark? And how does the secondary circuit complete its path back to the coil? There is no continuity on a coil between the HT connection and the casing, the drawing is wrong. Just get a multimeter and test one. Some of them are even painted for heavens sake!!! All these years, every publication I have ever seen has been wrong on this, but they all agree with each other so the myth goes on. Come on, someone independently think of the only possible logical answer to the last question and rewrite the books, please!!!James Turner, Birmingham UK — Preceding unsigned comment added by 86.25.123.52 (talk) 19:23, 16 September 2016 (UTC)[reply]

Nearly all coils encountered have three connections on them. The large one is the high voltage output to the distributor. One of the small ones connects to the contact breaker and capacitor. The other small one connects to the battery positive via the ignition switch. Since there is no other terminal provided, the only possible return route for the high voltage current is via the coil mounting lug(s). This makes sense because the return side of the spark plug gap is earthed to the metal engine casing and so is the coil's mounting lugs.

The exception to this is coils where there are two outputs designed to connect to two spark plugs (as in the second diagram in the article). But even here the centre point of the spark plugs is still earthed to the engine casing, even though there is no other return path. --Elektrik Fanne 17:05, 16 October 2016 (UTC)[reply]

I used to think this too, but there is no continuity between the coil casing or lugs and the high voltage terminal. I have a fibreglass bodied car and the coil is mounted to fibreglass only. It still works. — Preceding unsigned comment added by 86.25.123.52 (talk) 07:06, 11 January 2017 (UTC)[reply]

There are good diagrams here, yet unfortunately I didn't see a ballast resistor in any of them. The electronic ignition schematic (with electronics symbols) can probably omit one because there is a timing capacitor to control the maximum dwell time to limit the current through the coil.Dcebr (talk) 03:23, 30 July 2014 (UTC)[reply]

whoever is deleting "purpose" (as if that isn't pertinent) and editing main article. the purpose is not to be so generalistic as to say nothing, where the same number of words describes exactly. your welcome.

please watch youtube about adjusting points before flatulating about diagrams

someone is promoting the idea a condenser is a capacitor - it IS NOT (the name differs for a reason).

someone is flooding the article with lengthy descriptions that are really just trivia of part cost / efficiency related. AND are also incorrect and not revealing of electron direction and force anyway.

while the Kettering or other historical references are interesting. listing every every CONTRIVED electronic circuit that does the job:

needs to be a different topic - one for the auto hobby channel

and already one can see diagrams on google for particular cars - there is no need to list what each current car manufacturer is doing currently. just reference it or even show a few diag out of (chilton's)

my tone here is someone is deleting ANY addition without considering purpose or any rules of articles, simply to command the topic: while accusing me of being non-encyclopedic and even IP attacking me

this deletion harassment and IP attacks by "admins" or other deranged users needs to stop or be by reader vote - be done with psychopathic admins

72.219.202.186 (talk) 15:21, 19 August 2014 (UTC)[reply]

The problems are first, your writing is nearly incoherent. Basic English language writing ability is required to make significant edits. Others could copyedit for you to fix the problem, but your writing is so bad that nobody actually knows what you're trying to say. The essay Wikipedia:Competence is required explains this further.

Second, you are adding your own opinions and observations without clearly citing verifiable sources. This violates the Wikipedia policy no original research.

It is rude to call your contributions "crap", and Wikipedia doesn't allow attacking other editors. But honestly, if you can't be more clear and can't name sources coherently, then you're not helping the article. --Dennis Bratland (talk) 20:13, 19 August 2014 (UTC)[reply]

This article is wrong. The early engines had no spark plugs. The points were mounted inside the cylinder and a single coil was used as a inductor to generate the high voltage. This was the system used by the Wight brothers on their first aircraft. I have seen a replica of that engine at Oshkosh. Please get it right.

see; http://airandspace.si.edu/exhibitions/wright-brothers/online/fly/1903/engine.cfm

Arydberg (talk) 18:49, 16 September 2015 (UTC)[reply]

Why do you think that one engine, relatively late (1903), built unusually lightweight for aircraft use, has such relevance as redefining the invention for other and earlier engines? Lenoir was forty years before this, Benz' first car twenty years earlier. Both had a spark plug. Andy Dingley (talk) 20:31, 16 September 2015 (UTC)[reply]

The statement "The simplest form of spark ignition is that using a magneto: is not correct. The wipe ignition is simpler. I do not claim it is a good system. I find Wikipedia people to be wrong on many counts and when corrected they destroy you. That is my experience. Arydberg (talk) 17:14, 19 September 2015 (UTC)[reply]

I think the editor was probably trying to say simplest of current systems, not obsolete ones, but they weren't very clear about that. Hot tube is even more primitive than wipe, but they're kind of a different category. Can you fix it, if you've got the sources handy on this subject? --Dennis Bratland (talk) 19:27, 19 September 2015 (UTC)[reply]

Here are some suggestions. There are many more needed as the subject is explained in a style that jumps around too much and explains some thing over and over. In the text below capitals are used to show changes

An ignition system is a system for igniting a fuel-air mixture.

Ignition systems are well known in the field of internal combustion engines such as those used in petrol (gasoline) engines used to power the majority of motor vehicles, but they are also used in many other applications such as in oil-fired and gas-fired boilers, rocket engines, etc.

The first ignition system to use an electric spark was probably Alessandro Volta's toy electric pistol from the 1780s.

Virtually all petrol engines today use an electric spark for ignitionEXCEPT Diesel engines THAT rely on air compression for ignition, XbutX THEY usually also have glow plugs that preheat the combustion chamber to allow starting of the engine in cold weather.

EARLIER FORMS OF IGNITION INCLUDE THE HOT ROD AND WIPE IGNITIONS.

ONE EARLY TYPE USED BY THE WRIGHT BROTHERS WAS A INDUCTOR IN SERIES WITH TWO POINTS CONNECTED TO A BATTERY. THE POINTS WERE PLACED INSIDE THE CYLINDER AND SEPARATED TO PRODUCE THE SPARK.

here are two additions I suggest. The entire article needs to flow much better starting with primitive spark systems leading to CD systems and transistorized systems.

HOW A SPARK IS PRODUCED BY A BATTERY In a battery operation ignition system the coil is a high voltage transformer. As such it has 2 coils wound around a iron core. A primary winding with a low number of turns of heavy wire and a secondary which is a high number of turns of very fine wire. One end of each coil is connected to ground.

The second end of the secondary coil is connected to the spark plug. The second end of the primary coil goes to one of the points. The other point is connected to the battery. Before the plug is to fire the points are closed and a current builds up in the primary coil. This causes a magnet field to build up. At the moment the plug is to fire the points are separated. This causes the magnetic field to collapse and this collapse causes a high voltage in the secondary. It is this voltage that causes the spark. A capacitor is connected across the points to extend their life. This type of ignition was in use for many years.

HOW A SPARK IS PRODUCED IN A MAGNETO A magneto engine also has a spark coil as above with two windings on a iron core and a set of points. In this example the coil is assumed to be stationary and the magnet is assumed to be moving but the reverse is also possible. One point is grounded and thus connected to one end of the primary winding. The other point is connected to the free end of the primary winding. Before the spark is needed the points close then the magnet rotates by the iron core of the spark coil building up a current in the primary winding. This current is accompanied by a magnetic field. When the spark is to occur the points are opened and the magnetic field collapses which produces a high voltage spark.

As for sources they are a bad word to me I was banned from a wikipedia topic for suggesting sources.Arydberg (talk) 13:41, 23 September 2015 (UTC)[reply]

Note that HT systems (with a two winding transformer) were a later development from LT systems, which had a single inductor or magneto winding. Andy Dingley (talk) 15:21, 23 September 2015 (UTC)[reply]

A recent unsourced change has replaced transformer with "inductor with a step-up winding". Why? Andy Dingley (talk) 10:53, 4 March 2016 (UTC)[reply]

And an unsourced change has also switched it back to a transformer!

It is incorrect to describe an ignition coil as any form of transformer, including a pulse transformer- it is an inductor with a step up winding, just as in switch mode power supplies. Also most ignition coils are wound like- I hate to say this- an auto transformer, with the low voltage winding common to the high voltage winding, unlike the schematic currently shown in the article at the moment.

A transformer, at a very basic level, has the characteristic, Vp*N2= Vs*N1 (corrected).

Where N1= number of primary turns, N2= number of secondary turns, Vp= primary voltage, and Vs= secondary voltage.

Another characteristic of a transformer is that even when a primary voltage is applied, if no current is flowing in the secondary, no current will flow in the primary (assuming a perfect transformer that is. In practice a small magnetizing current will flow and will be lost as heat).

For this reason, a transformer is designed to have a high primary inductance. If a coil had the same inductance as a similar voltage transformer, little current would flow, and only a relatively low voltage would be produced and a low energy, insufficient to ignite a fuel air mixture. Also switching at engine speeds would be difficult.

A typical turns ratio of a car (auto) coil is 1:50 to 1:100 (corrected). This means if a 100:1 ratio coil were functioning as a transformer, and taking the battery voltage to be 12V, the high voltage winding would produce 100*12V= 1.2KV. This voltage would not be sufficient ignite a petrol air mixture reliably and, in any case, it is at odds with a typical HT voltage for a car (auto) of between 12KV to 30KV. In theory, the coil in a Kettering ignition system does not need a step up winding. It is not fundamental to the functioning of the circuit and is only there for practical considerations.

From the moment that the points (or switching element) closes and applies 12V DC to the coil, the current builds aiming towards infinity. The rate of current increase is proportional to V/L. and the energy stored in an inductor is (I↑2*L)/2 joules (corrected), where I is the final current in the inductor in amps and L is the inductance in Henrys. If the points remain closed for a long time, the coil low voltage current is limited by the ballast resistor (if fitted), and the self resistance of the low voltage coil. This current is typically 4A.

When the points open the coil attempts to oppose any current change by generating a voltage heading toward plus infinity (V= -L*di/dt) (this sentence has been corrected). Before this voltage can be reached the plug arks over at a high voltage depending on the plug gap, temperature, pressure, and fuel air mixture. This dissipates much of the energy in the coil which previously had no where to go. The coil then oscillates at a frequency defined by its inductance and parasitic capacitances.

Normally though, a capacitor of between 100nF and 470nF is connected across the switching element to protect it and to lower the resonant frequency of the coil (the capacitor is effectively in parallel with the low voltage winding on the coil). The lower resonant frequency gives a longer spark duration to ensure a good burn.

In view of the above, at present this Wikipedia article gives a completely misleading description of how a Kettering ignition system works. It is especially misleading because of the superficial similarity between an inductor (coil) and a transformer; you see this error frequently in electronics, and when I was a newbee electronics design engineer I needed to learn about ignition systems. It took me a while to untangle all the miss information before I could start designing.

Worse still, the article at present says that an AC signal is applied to the primary of the transformer. That is absolutely not true: a DC voltage is applied to the low voltage winding which causes a current to grow in the low voltage winding.

This Wikipedia article should be changed to give a correct description of the normal spark ignition system. CPES (talk) 11:53, 4 March 2016 (UTC)[reply]

Agreed, the image is wrong. The coil internals need fixing.

You have confused me though. You cite a simple turns ratio for voltage and then talk about "the points apply 12V DC to the coil" for what is absolutely a pulse transformer and needs to be treated as such (i.e. considering current, not voltage), more than a simple AC transformer. Yet you have some electronics background? The Kettering system is a system, and it just doesn't work as isolated components (the condenser isn't simply there as a snubber).

Why is this not a transformer? It's not an auto-transformer as the two windings are connected, not shared.

What's the definition of a transformer? How is this not meeting that? Andy Dingley (talk) 14:45, 4 March 2016 (UTC)[reply]

In what way is either of the images wrong? Both diagrams are valid ignition systems. The arrangement in the larger diagram is for a Citröen 2CV but it often turns up in other two cylinder engines. In this design the secondary of the coil has no earth connection. Both plugs fire once per crankshaft revolution, but the sparks that occur between the exhaust and induction cycle in each cylinder are wasted. Elektrik Fanne (talk) 15:19, 4 March 2016 (UTC)[reply]

(edit conflict)You have got the principles wrong, and you have multiplied the wrong voltage.

The circuit consists of a transformer (often called a coil for historic reasons), a contact breaker (often called points) and (very importantly) a capacitor (often called a condensor for historic reasons). The contact breaker is in series with the primary winding of the transformer and the capacitor is in parallel with the contact breaker - you already knew this. The system will not work well, if at all, without the capacitor.

When the contact breaker closes, the current through the primary builds up to a value limited by the resistance of the primary and any other resistance in the circuit. The inductance does not affect this final current, only the speed with which it builds up. The magnetic circuit of the transformer is now storing energy (0.5 x L x I² Joules) - your formula was wrong.

When the contact breaker opens, the current does not fall to zero. This is impossible with an inductance in the circuit. The inductance makes the current continue to flow (it has to in order to disipate the stored energy). For the inductor to cause the current to continue to flow, it induces a driving EMF as result of Lenz's law that is much larger than the original battery voltage and it attempts to oppose the decrease in current (this means that the induced EMF is of the opposite polarity to the battery voltage as seen at the transformer primary).

It is this much larger EMF that is stepped up by the transformer action to 400 times this value. The current charges the capacitor until it stops. The energy then flows back and forth between the capacitor and the inductor in a decaying sinusoid oscillation. As noted, this gives a better spark and more reliable ignition. The article does not make this clear and I shall clarify the point.

In a practical ignition coil one terminal of the primary and one terminal of the secondary is usually a common connection, but that does not stop it being a transformer. Elektrik Fanne (talk) 14:48, 4 March 2016 (UTC)[reply]

(1) Reply to (Andy Dingley)

Agreed, the image is wrong. The coil internals need fixing.

That's good

You have confused me though.

I don't see how, typos notwithstanding. I have explained everything at length, and quite clearly.

You cite a simple turns ratio for voltage and then talk about "the points apply 12V DC to the coil"

I can't see any problem with either statement. The closing points do apply 12V DC to the coil low voltage winding.

for what is absolutely a pulse transformer and needs to be treated as such (i.e. considering current, not voltage), more than a simple AC transformer.

You don't seem to appreciate the difference between an inductor and a transformer even though I have gone to some length to explain this very basic principle. Just insisting something is so does not make it so. It is important to differentiate between form and function. There is no fundamental difference between a pulse transformer, current transformer or any other kind of transformer for that matter. Quite simply a transformer does not store energy in it's core like an inductor does. It supports a flux. But if you drive a transformer like an inductor it is then functioning like an inductor, and in terms of circuit function, is an inductor.

Yet you have some electronics background?

Not sure if that is a genuine question. If it is, you haven't read/understood a word of my long and comprehensive post. But just to say, I am an electronics design engineer of 35 years standing and have designed, developed, and trialed a number of electronic ignition systems for road vehicles. All used standard ignition coils and capacitor discharge where, you will be pleased to know, the coil is used as an auto transformer with a 400V or so pulse being fed to its primary.

The Kettering system is a system, and it just doesn't work as isolated components (the condenser isn't simply there as a snubber).

I have no idea what you are getting at. The Kettering Ignition system is a... system and I didn't say otherwise. The important thing is that it functions by storing energy at one VI ratio in an inductor and discharging it at another VI ratio. This is not a transformer action. Further more, I said that the high voltage winding is only there for practical reasons and is not theoretically necessary. This can be proved by drawing a simple schematic of the Kettering system with a simple one coil inductor and analyzing that with a 12V pulse. It will still produce a theoretically infinite voltage when the switching element opens As there is no secondary involved even you should see that no transformer action is involved either.

I didn't say that the capacitor is only a snubber. I specifically said that it protects the switching element and lowers the resonant frequency of the inductor with it's parasitic capacitances, to provide a longer spark. I get the impression that you have not read my post above.

Why is this not a transformer? It's not an auto-transformer as the two winding are connected, not shared.

The question is, why is it a transformer. There has been no statements to support this. Sure, it looks like an auto transformer and also as an adjunct uses inductive coupling as both transformers and inductors do, but that does not make it a transformer when used in the Kettering ignition system.

I do not see how, in one breath, you can claim that a coil is a transformer and in the same breath claim it is not an auto transformer. The low voltage winding and the high voltage winding are connected and they are both on the same core and their voltages sum. That is the definition of an auto transformer- except when it is an inductor with an inductively coupled winding that is.

What's the definition of a transformer? How is this not meeting that? (talk)

I have already explained that in some depth.

(2) Reply to Elektrik Fanne (but this Wikipedian seems to have resigned)

In what way is either of the images wrong? Both diagrams are valid ignition systems. The arrangement in the larger diagram is for a Citröen 2CV but it often turns up in other two cylinder engines. In this design the secondary of the coil has no earth connection. Both plugs fire once per crankshaft revolution, but the sparks that occur between the exhaust and induction cycle in each cylinder are wasted.

Conventional ignition coils normally have the low voltage and high voltage coils connected. This has no electrical impact but simplifies coil construction and the car wiring loom, because it eliminates one terminal connection. As you imply, when the lost spark technique is used the high voltage coil must be isolated from the low voltage coil because plugs are connected to both ends of the high voltage coil.

You have got the principles wrong, and you have multiplied the wrong voltage.

I have not got the principle of a transformer wrong, as you can quite easily see from my subsequent calculations about secondary voltage and turns ratio. You are being deliberately obtuse and trying to score brownie points. I do admit to a typo though which, had you been objective, you could have simply pointed out and cleared up the issue. By the way, I have corrected the error in the original post.

The circuit consists of a transformer (often called a coil for historic reasons), a contact breaker (often called points) and (very importantly) a capacitor (often called a condensor for historic reasons). The contact breaker is in series with the primary winding of the transformer and the capacitor is in parallel with the contact breaker - you already knew this. The system will not work well, if at all, without the capacitor.

A transformer has never been called a coil, except maybe in prehistoric times.

I can't see the point of you telling me something I already know, and have already stated.

You are making statements without any supporting explanations.

When the contact breaker closes, the current through the primary builds up to a value limited by the resistance of the primary and any other resistance in the circuit. The inductance does not affect this final current, only the speed with which it builds up. The magnetic circuit of the transformer is now storing energy (0.5 x L x I² Joules) - your formula was wrong.

Only at low revs or when the engine stops, is the current limited by resistances. As the revs increase there is less and less time for the inductor current to grow and hence the stored energy in the inductor drops radically, especially as the stored energy is proportional to the square of the current. This is one drawback of the Kettering system, which can be more easily overcome by a capacitor discharge (CD) system where, incidentally, you can claim that the coil is acting as an auto transformer.

Yes, my formula for energy stored in an inductor was missing the square on the current, but you surely can see that it was only an omission and did not affect the underlying principle anyway. Once again you are being deliberately obtuse.

When the contact breaker opens, the current does not fall to zero. This is impossible with an inductance in the circuit. The inductance makes the current continue to flow (it has to in order to disipate the stored energy). For the inductor to cause the current to continue to flow, it induces a driving EMF as result of Lenz's law that is much larger than the original battery voltage and it attempts to oppose the decrease in current (this means that the induced EMF is of the opposite polarity to the battery voltage as seen at the transformer primary).

You are correct about the current not falling to zero. As you quite rightly point out that would be impossible- my error (I do like your explanation). But I was correct in principle as supported by the formula shown. By the way, you have just described an inductor perfectly- just as I did.

It is this much larger EMF that is stepped up by the transformer action to 400 times this value. The current charges the capacitor until it stops. You have also called the 'transformer' an inductance because it is necessary for your explanation.

I have no idea what this means. What larger EMF? By the way the coil turns ratio is 50 to 100, not 400 as I originally said. Are you saying that the coil resonates with the capacitor. If so I agree- that is exactly what an inductor does. As I have said before, practical aspects aside, the Kettering circuit will generate the same high voltage without a coil with a high voltage winding, just as, in principle, boost switch mode power supplies do.

The energy then flows back and forth between the capacitor and the inductor in a decaying sinusoid oscillation. As noted, this gives a better spark and more reliable ignition. The article does not make this clear and I shall clarify the point.

I already said that the inductor will resonate with the capacitances. Yes, providing there is no back connected diode across the switching element, there will be an exponential decaying sinusoid, after a fashion. But if a diode is fitted the waveform will normally be suppressed on the first negative excursion (in a +12V system).

In a practical ignition coil one terminal of the primary and one terminal of the secondary is usually a common connection, but that does not stop it being a transformer.

It does not make it a transformer either.

(3) Summary

Apart from insisting that a coil is a transformer neither of you have explained how the high voltage of around 18Kv is generated. I have done this. As I have explained, the turns ratio of the coil, treated as a transformer won't do it, especially as the high voltage winding is not theoretically necessary, and a simple two terminal inductor will generate the required high voltage necessary to spark the plug.

The worry is that, at present, this Wikipedia article, perpetuates a fallacy about the basic functioning of the the Kettering ignition system, which is used on practically all spark ignition engines, and will lead the average reader down the wrong path when trying to get a simple understanding of its modus operandi. Unsigned comment by CPES talk 21:04, 4 March 2016

Why do you say that I seem to have resigned?

Why do you believe that your calculation of voltage and turns ratio is correct? As you stated in a previous post you multiplied 12 volts by the turns ration of 400:1 and got nowhere near the required voltage. You have changed the ratio in your post making the goal even further away. You must not refactor posts once people have replied to them, it might look like a deliberate attempt to make the replier look wrong. Whatever the turns ratio is (I have no information which is why I accepted your figure of 400:1), the 12 volt battery supply is the wrong voltage to use. The right voltage is the induced emf at the moment the contacts open. This is much larger than 12 volts, so your latest calculation is still wrong.

There are many examples of the construction which, though clearly relying on transformer action, have been historically called coils. The subject ignition coil is but one example. A Tesla coil, Oudin coil and Ruhmkorff coil being just three notable examples - the latter the forerunner of the ignition coil. It is true that the more conventional transformer is not generally called a coil (though they are constructed using coils).

The stored energy is dependant on the time allowed for the current build up. The parameters of the circuit are designed to allow that energy to be as large as possible over the rpm range. You cannot 'claim' that the coil in any system is acting as an auto transformer, because it doesn't and is not constructed as such (and there is practically no advantage in doing so). In a CD system, the coil has practically the same function. The only difference is that the initial stored energy starts off in a capacitor rather than the inductor (and it has to be charged to about 400 volts to match the voltage produced by the coil in an LD system).

Unfortunately, my crystal ball is away for its annual service. It is therefore not possible for me to tell whether an error in a formulae is a typo or a failure to understand the subject being discussed. Unfortunately, I get used to my students making errors of this sort so I assumed the later. I do accept (because you readily accepted my correction) that it was a typo.

You accepted that the inductor attempts to keep the current flowing once the contact breaker opens. In order for it to do that there must be a driving voltage from somewhere (with the contact breaker open, it cannot be the battery). This voltage is produced by the inductor. As the current collapses, the change in current induces a voltage in the inductor winding (equal to L(di/dt)). What I cannot understand here is why you are having problems understanding this when you state in your summary, "...a simple two terminal inductor will generate the required high voltage necessary to spark the plug.". Where do you believe that this 'required voltage' will magically come from?

You are correct that a simple (two terminal) inductor will produce a high voltage. The problem is that such an arrangement would require a capacitor with a much larger breakdown voltage which would be more expensive to manufacture. The second problem is that the spark plug would dissipate a large part of the energy on the first oscillation and subsequent oscillations would not exceed the spark plug break down voltage (the spark gap limits the voltage). The double winding solves both problems. However, having said that: the first magneto systems did indeed use a single winding. The double winding made a substantial improvement in the operation because limiting the secondary voltage did not limit the primary current and a longer spark duration resulted.

Where has this 'back connected diode' come from? Has anyone seen an ignition circuit where anyone has been fool enough to connect a commutating diode across the primary of the coil? Good luck trying to get the engine to start.

Two windings that have mutual inductance are a transformer regardless of the construction. Indeed, a single winding with a tap is also a transformer (see Autotransformer).

I would like to take up one point that Andy Dingley made. He stated that a transformer does not store energy like an inductor does. As I said, any construction that has two (or more) windings that have mutual inductance is a transformer. While it is true that a power transformer does not rely on stored energy in the core, that is only a result of the way it is used. It is quite possible to use a transformer in a manner in which energy is stored in the core. The most obvious example that springs to mind (where it is called a transformer) is the line output transformer used in obsolete cathode ray tube display systems. These circuits have a provision to recover this stored energy to (among other things) derive the high voltage required to operate the tube. Another example is the transformer in switch mode power supplies (although they don't have a mains transformer, they do have a flyback transformer later on in the circuit). Elektrik Fanne (talk) 15:24, 5 March 2016 (UTC)[reply]

Some references that an ignition coil is a transformer: Within an automobile's ignition system, an ignition coil plays an essential role in transforming the battery's 12 volts to the 20 or 30 thousand volts or more needed to spark the spark plugs., Therefore, the voltage is transformed and multiplied., The coil is a simple device -- essentially a high-voltage transformer made up of two coils of wire.

Also see here, Figure 1, waveform b shows the voltage across the coil primary, the large reverse voltage at the time the contact opens is clearly visible. The description is for a coil without a capacitance but there is a hint of decaying oscillation from the stray capacitances in the coil and circuit.

I have just had a another look at the diagram of the circuit with the distributor (the smaller diagram). Some minimal research finds that the primary and secondary do not have a common terminal (as I stated) as the contact breaker and its parallel capacitor are, in general, in the earth side of the primary coil. It could just as easily be in the 12 volt wire with no affect on operation. I assume that the arrangement used has some cost saving attached to it. Coils for sale online are all 4 terminal devices (one side of the secondary being the coils metal shell (for metal shelled coils)).

It also seems that the open circuit peak voltage produced in the secondary circuit is between 32 and 40&nbsp:kV (though a coil should never be operated open circuit). The spark plugs limit the peak voltage to a lower value as the spark is essentially a crude voltage stabiliser. This means that the peak induced emf from the primary of the coil works out around 400 volts which would tally with the voltage required for the capacitive discharge system.

Finally, a properly designed system should have a period of time where the current in the primary has reached a steady state current (the 'dwell time') over the speed range of the engine. -Elektrik Fanne (talk) 17:26, 5 March 2016 (UTC)[reply]

@Elektrik Fanne: You recently edited the article to change series resonance to parallel. The primary of the ignition coil is definitately in series with the contact breaker/capacitor. I was about to revert your edit, but your discussion above suggests that you are more familiar with this subject than I am. I therefore have to ask, why do you think it is a parallel resonant circuit and not a series? 85.255.232.169 (talk) 23:05, 5 March 2016 (UTC)[reply]

Your assumption that the circuit is a series resonant circuit is understandable, given the configuration, but incorrect. Series resonance and parallel resonance are different manifestations of the same phenomenon. You cannot get a decaying oscillation in a series resonant circuit - what you get is a voltage magnification across the reactive parts and a whole different discussion. In order understand what is going on, it is necessary to analyse the circuit into its active parts at any point in the operating cycle. There are three phases to the operation of the circuit.

Phase 1: starts the moment the contact breaker closes and the current starts increasing through the coil.

Phase 2: starts when the current reaches is final steady state value (the dwell time).

Phase 3: starts the moment the contact breaker opens.

Let's examine each phase. For a start, the ignition switch itself remains closed throughout the entire operating cycle of the circuit, so for analysis purposes, it can be considered a short circuit and ignored.

Phase 1: the battery +ve terminal is connected to the coil, and the other coil terminal is connected to the contact breaker/capacitor which is connected in turn via the chassis to the battery -ve terminal. Since the contact breaker is closed, it is a short circuit and is shorting the capacitor which therefore has no effect. The coil secondary is open circuit, as there is no spark in this phase, and therefore also has no effect. The circuit thus simplifies to the primary of the coil (which consists of an inductor in series with its resistance) in parallel with the battery.

Phase 2: the equivalent circuit is the same as phase 1, but because the current has reached its steady state value and is no longer changing, the inductance disappears from the equivalent circuit and it becomes just the coil's resistance in parallel with the battery. Again the coil secondary has no effect because it still open circuit.

Phase 3: Now everything changes. I will put the battery and the secondary coil to one side for the moment. The contact breaker is open circuit which means that there is now a capacitor in the circuit. This also means that system is an open circuit as far as the DC supply from the battery is concerned, and no net DC current flows. However, we have to consider the equivalent AC circuit. The coil's primary inductance is connected to one side of the capacitor. The other side of the coil is connected to the battery +ve. The other side of the capacitor is connected to the battery -ve via the chassis.

As far as the AC analysis is concerned, the battery has no reactive parts and is a very low resistance (so low in fact, that it can deliver around a killowatt to start the engine). The battery is thus a short circuit (to AC) and the capacitor is therefore in parallel with the coil - a parallel resonant circuit. The energy stored in the magnetic circuit of the inductance of the coil oscillates back and forth between the inductance and the capacitance. In this case the secondary is not open circuit due to the arcing across the spark plug(s). The secondary couples the energy out of the inductor every time it charges and the oscillation decays more rapidly than if the secondary was not there. The primary and secondary resistances also dissipate some of this energy. -Elektrik Fanne (talk) 12:54, 6 March 2016 (UTC)[reply]

OK. I had to draw it out, but I can see what you are getting at. Thank you for the explanation. 85.255.234.100 (talk) 17:12, 9 March 2016 (UTC)[reply]

The above explanation by Elektrik Fanne (talk · contribs) is unsourced; it is correct in many parts but wrong in some others. "You cannot get a decaying oscillation in a series resonant circuit" is wrong. Both are resonant circuits with losses; both exhibit decay as energy is lost to heat. The series/parallel label is one of viewpoint. When the points open, the low impedance of the (series-connected) battery sees the low impedance of a series resonant circuit. When the points open, the (parallel-connected) points see a high-impedance parallel resonant circuit. The LC oscillation is an artifact rather than a significant feature of the ignition circuit. See Talk:Induction coil#Dubious. Glrx (talk) 18:06, 26 March 2016 (UTC)[reply]

The above explanation by Glrx (talk · contribs) is unsourced. It is also contradictory as it claims that when the points open there is both a series and parallel resonant circuit simultaneously. If my explanation is unacceptable then neither is yours. --Elektrik Fanne 18:11, 26 March 2016 (UTC)[reply]

Sources are given if you go to Talk:Induction coil#Dubious. Glrx (talk) 23:11, 28 March 2016 (UTC)[reply]

This series of edits goes back to a flawed explanation and adds (restores) poor sources.

All but one of the sources are poor.

• http://www.fas.harvard.edu/~scphys/courses/E1b/E1b_3b.pdf is notes from a physics II class taught at Harvard Extension School. It is interested in showing physical phenomena. A prof looks at an ignition system and making some observations about it. He's not an expert on ignition systems. It is, for the most part, unsourced (there are refs to the class's physics text). There are many issues with what it shows
  1. Figure 1 shows a setup with a Pasco low-output impedance Pasco function generator. The low impedance avoids the current interruption issue; low Z is not on-off in the switch sense. The notes do not specify the model number or the output impedance. A Pasco 9587C is a possible candidate because it has low and high Z outputs. You can find copies of the manual on the net, but I won't WP:COPYLINK vio them here; the info I found does not specify output impedance.
  2. At page 2, the prof does not trust his students to understand derivatives; he use V = −L (ΔI / ΔT). (You may cringe here for other reasons, too.)
  3. At the bottom of page 4, the notes start to explain the purpose of the capacitor. Without the cap, the current changes instantly, the back emf is infinity, and there's a breakdown. That's good, but after focusing on dI/dt, there's not a good discussion about how the capacitor can control dV/dt.
  4. Figure 3 is an RLC. There's no spark gap indicated in the figure. The section is just discussing the primary. The spark plug/spark gap is not added until the bottom of page 7, point 15. Note figure 3 does not have a capacitor across the switch as did figure 1.
  5. The notes describe the breakdown as happening "near the point the current is interrupted". That's the switch contacts because the section is only talking about the primary circuit. The arcing explanation at the bottom of page 5 is half right; see Ott. The prof ignores the dynamics of opening the contacts. Using larger C will mean lower peak V on the damped wave, but there is more to the story. The "resulting in less ionization and arcing through the air" comment is cringe-worthy. The refs that the edits deleted explained that the points' break arc is eliminated.
  6. Look at point 15, which asks, "If you remove the capacitor across the switch, do you still get a spark? Is the spark stronger or weaker? Why?". This is a lab exercise. The work is not trying to explain what is going on but rather challenge the students.
  7. The article now claims the capacitor's "main function is to form a parallel resonant circuit with the ignition coil". The notes actually state, "To minimize the arcing, a capacitor is put across the points, as shown in Figure 1." (page 5, emphasis in original.) A bold comment should be the main function. Searching the notes for "parallel" finds nothing. "Resonance" is mentioned in a few places, but it only says RLC. Resonance is emphasized in the primary-only section of the notes. Many of the notes' implications about resonance are shown to be naive by the worldphaco.net ref below.
• http://www.rtftechnologies.org/emtech/coil-driver.htm is not a reliable source; it appears to be Andrew Seltzman's personal website; personal websites/blogs are not reliable sources. There's little explanation about what the author thinks what is going on. Furthermore, the circuit is suspect: 2N3055 are not high-voltage transistors.
• https://www.physicsforums.com/threads/role-of-capacitors.330883/ is not a reliable source. Anybody can chime in.
• http://www.worldphaco.net/uploads/CAPACITIVE_DISCHARGE_IGNITION_vs_MAGNETIC_DISCHARGE_IGNITION..pdf is much, much, better, but the oscillatory comment is restricted. "Equation 2 and 3 will come in handy later in calculating the expected peak voltage on the primary or secondary winding of the coil when the magnetic field has collapsed and given all of its stored energy to the electric field of the ignition coil's capacitances." (page 5) The work goes on for an extensive discussion of leakage inductance, resonances, and coupling issue. There's a nice figure 8. However, the section states (emphasis in original), "No spark is allowed to occur for this test (coil HV output wire is disconnected)." (page 13) Figure 9 is also interesting; there's some confusion about primary/secondary labeling and the offscale aspect. "Initially there are oscillations in the spark current, but the oscillations do not stop the spark as the value during the spark time is never zero." The spark current lasts for "2.3&nsbp;ms" (I see about 2.8). These oscillations are due to leakage inductances. "When the spark occurs the ignition coil's secondary is loaded and the primary leakage inductance Lip resonates with the primary capacitance." (page 15) That 8.7 kHz is not the capacitor selection that was done to limit "the expected peak voltage on the primary"; the primary resonance has a simple value of 3.4 kHz and more involved reason to be at 2.4 kHz. (page 10.) This is a good reference, but it is being misused.
• http://www.powerlabs.org/igncoildrivers.htm is another personal website: "Here you can see a reconstruction of my first High Voltage device. It is an ignition coil driver; I built the original when I was 11 years old from plans I found in an old electronics magazine". "Here is the final version of a 555-based ignition coil driver I built." (I don't doubt he's done some interesting things. There is an EE Web interview with him (http://www.eeweb.com/spotlight/interview-with-sam-barros) where he says, "In my previous job I worked for a company that specialized in High Voltage special effects; we used large Spark Gap driven Tesla Coils to produce real lightning bolts that could be up to several million volts, and incorporated these into shows, movies, etc…".)
• http://mgaguru.com/mgtech/ignition/ig108.htm I've already complained about Barney Gaylord's personal website. Gaylord apparently believes in multiple sparks ("Assuming the initial spark ignites the fuel air mixture, ...").

The worldphaco.net is a reasonable source, but it is giving a much more involved description of what is going on. It is also not addressing the points because they no longer break the high current. Many of the references above discuss the spark is diminished or absent if the capacitor is removed. There's no clear explanation why that happens. The Harvard physics class notes leave that to the student but probably don't understand. Harvard explains about maximizing the energy transfer from the primary to the secondary by making the resonances equal ("To take advantage of this resonance phenomenon and huge voltage gain, we choose the capacitor across the points so that the primary RLC circuit of the coil has a resonance frequency close to that of the secondary coil." page 7). Well, the energy has already been put into the coil; if the primary didn't conduct at all (it is open), then all of the energy in the flux collapse will be directed to the secondary. There's something else going on, and that is why the deleted Taylor-Jones reference is important. Another poor ref, MGAguru, describes the result ("If you do not connect the capacitor the spark may be so small as to be hardly visible, and lucky if it will jump 1/32 inch gap.") but gives a different faulty explanation (the importance of "ringing" for raising the primary voltage; without the ringing, the voltage would go higher).

Here's what happens without a capacitor.

1. Points are closed and energy builds in the coil.
2. The points open. Let's say they instaneously separate to 1 μm. Assume no capacitance (or just stray capacitance). The back emf rises rapidly. Ott, page 181, figure 7-7, shows an arc discharge breakdown voltage of 30 V at 1 μm (the electric field is 30,000 V/mm). So the arc discharge starts and collapses to the arc sustaining volage (about 12 V and independent of separation distance). If we view the coil as a simple transformer (ignore leakage inductances and assume everything is infinitely fast) with a 1:100 turns ratio, then the spark gap voltage peaks at 3000 V and then falls to 1200 volts. That's not enough to induce an arc breakdown at standard pressure at a separation of 1 mm. It might be enough to start a glow discharge at standard pressure (but that requires somebody else to provide some initial electrons). It doesn't look like enough for a glow discharge at 2 mm. Things will be worse in the combustion chamber because the mean path length is smaller. That's why Taylor-Jones teaches eliminate the break arc.

The situation is much different with the capacitor.

1. Points are closed and energy builds in the coil.
2. The points open. The capacitor limits the voltage across the points. The primary current was fixed; (i/C) = dV/dt; choose a C to limit dV/dt. It's a foot race. There won't be an arc until the ionization potential is reached (say 10 volts). The contacts are opening at some speed, and the voltage across the contacts is increasing at some rate. Choosing the capacitance large enough keeps the electric field low enough to prevent an arc discharge break down. By the time the contacts open to 10 μm, the contacts are safe for any voltage below about 300 V.
3. If the secondary is a simple ideal transformer, then its secondary voltage has been rising 100 times as fast. The spark gap breaks down before the primary reaches 300 V.
4. The spark goes to its sustaining voltage. The primary voltage is 100 times smaller. Most of the energy now gets sucked out of the coil. Instead of a resonant LC circuit, the capacitance charging currents are diverted to the arc; V is constant, so there's no capacitive current. Instead, there's a constant voltage and linear current discharge: V = L di/dt; V and L are constant, so di/dt is constant. See worldphaco.net, figure 9, spark current discharge, E to F, which has a nice linear discharge.
5. The spark dies. The small amount of leftover energy dissipates as a damped wave.

Throwing in leakage inductances makes more squiggles, but the basic story doesn't change.

It's not about resonances but rather energy and break down voltages. Yes, there's a primary circuit resonance, but it is more about limiting voltage / keeping the primary current going rather than making a damped sinewave.

Glrx (talk) 23:11, 28 March 2016 (UTC)[reply]

Basically WP:TLDR. You are displaying all the signs of being a tendentious editor. You seem to have made your mind up about the modus operandi and no amount of discussion or citations is going to change that. You resort to writing extensive essays, to undermine any references that are provided (in spite of the fact that your own are severely wanting).

(At risk of sounding like a parrot.) The coil and capacitor form a parallel tuned circuit, because the other connection is via the (very) low impedance battery for exactly the same reason as this circuit arrangement of an oscillator is a parallel tuned circuit. What gets me is that in a recent post above, you actually admit it is a tuned circuit (even if you cannot make your mind up which type).

The position is complicated because there are references that mention the purpose of the capacitor to supress any arc but do not mention the resonant circuit or its true function (probably what you keep turning up). This is mainly because such texts are written for motor vehicle mechanics who neither understand resonance nor even care what it is. All they need to know is that signs of excessive arcing mean the capacitor needs replacing and that if the capacitor fails completely, the circuit will not work at all (something your theory fails to explain). Your theory only claims arc suppression which would mean that the circuit would work without the capacitor, but the contacts would have a short life. Your theory also does not explain why a solid state switch replacing the contact breaker still requires the capacitor even though there is no arc to suppress.

I have found a better reference and will add it to the article. As we do not need seven references for one point, I shall leave only the two authoritative book based references. 85.255.232.231 (talk) 08:04, 29 March 2016 (UTC)[reply]

I have always admitted there is a resonant circuit. A capacitor and an inductor in a mesh can trade energy back and forth. Labeling it a series or a parallel network depends on perspective and doesn't matter in this application. The battery sees a series resonance (see 85.255.232.169 post above); the battery must carry all of the circulating current. If one transforms the battery to an equivalent short circuit, then the resonator looks like a parallel circuit. Well, series circuits can be transformed to equivalent parallel circuits and vice versa. The bullet is the label does not matter here, and my article edits have not advocated for a series or a parallel label but rather dropped the label. The important issues are the energy storage in the coil flux and the induction at flux collapse.

"Your theory only claims arc suppression which would mean that the circuit would work without the capacitor, but the contacts would have a short life." No, Taylor-Jones explains the break arc must be suppressed to develop higher seondary voltages. The explanations above are not about extending the contact life (contact life is not even mentioned) but rather preventing the primary break arc that both steals energy from the magnetic field (so the secondary cannot use it) and prevents the secondary from achieving a high voltage by clamping the primary to a low voltage; the clamping would prevent a spark in the secondary. Consequently the primary break arc implies failure. (Yes, a failed/missing capacitor would direct much more energy toward contacts and cause them to fail more quickly.)

The article currently claims that the capacitor has the dual function of resonance and contact protection. There's no explanation of why the resonance is important. What resonant frequency should be chosen? The article claims that there are many cycles of this resonance: "During resonance, energy is repeatedly transferred to the secondary side until the energy is exhausted." The resonance is used to explain the duration of the spark: "As a result of this resonance the duration of the spark is sustained". A resonance would not maintain an arc, it would destroy the arc as the current and voltage reverse. The spark would have to reignite at each reversal.

The story is simpler. The break arc must be suppressed. A capacitor of sufficient size will do that. It sets an initial dV/dt so the points can open without arcing. That is explained in Ott and summarized above. The side-effect of using the capacitor is a primary side resonance, but the resonance was not a goal.

"Your theory also does not explain why a solid state switch replacing the contact breaker still requires the capacitor even though there is no arc to suppress." Solid-state switches also need to be protected from breakdown, and a capacitor will serve that purpose. Ott, Noise Reduction Techniques in Electronic Systems, John Wiley, 1976, page 193, states, "If an inductive load is controlled by a transistor switch, care must be taken to guarantee that the transient voltage generated by the inductor when the current is interrupted does not exceed the breakdown voltage of the transistor." A capacitor will limit the primary back emf and protect the transistor. Simple contacts and solid-state switches confront similar problems; the solution can be similar.

An additional function of the capacitor is system breakdown protection. The capacitor is chosen so the damped oscillation stays within the voltage breakdown limits of the system. If the system is operated without a rotor or plugs, then the coil and other parts of the system will not be immediately destroyed. The capacitor needs to be big enough to suppress the break arc and big enough to suppress unwanted breakdown, but still small enough that it will permit the desired sparks.

I have been unable to find your book references with Google or WorldCat:

• https://www.worldcat.org/search?q=ti%3Aautomotive+electrical+systems+au%3Aemerson&qt=advanced&dblist=638
• https://www.worldcat.org/search?q=ti%3AAn+engineer%27s+guide+to+vehicle+electrical+systems+au%3AWilliamson&qt=advanced&dblist=638

Please provide the publisher and the date or LCCN.

Glrx (talk) 18:52, 29 March 2016 (UTC)[reply]

Gentlemen. This seems to be a discussion that has been going on for some while (here and elsewhere). There is a problem that you are each fighting from a particular view point. The big question is, who is right and who is wrong? Some of your submissions suggest that you are each aware that something is not quite right with your viewpoint. Well, the answer to the big question is – that all three of you are right. You are just are not joining enough of the dots of evolution to get the complete picture. Some of this is repetition, but I aim to pull all the relevant points together. Also, it is difficult to be brief, so bear with me. However, I promise that there is no maths in this anywhere.

Consider for a moment that the ignition coil is built as an ideal transformer but with the resistive losses. Also consider that the primary circuit is exactly the one under discussion. As has been pointed out elsewhere, if the secondary is not connected to anything, it has no influence over what happens in the primary. When the contacts open, the primary voltage and current form a decaying sine wave by virtue of the LC circuit formed, the energy of which dissipates in the primary circuit losses. Both Elektrik Fanne and 85.255.232.231 are correct on this point.

Now consider that the secondary is short circuited. The behaviour of the primary side is now very different. When the contacts open, the primary side attempts to oscillate (again because it is an LC circuit). However, in this mode, a significant part of the energy is coupled across to the secondary and is dissipated in the secondary losses. The primary current does not oscillate, but just decays to zero over a protracted period. The short circuit secondary has totally damped any oscillation that attempted to occur. The secondary current mirrors the primary (because this is an ideal transformer) with a magnitude in inverse proportion to the turns ratio. This last behaviour is what I suspect Glrx is focussing on.

In both circumstances above, one function of the capacitor is to suppress the arc that would appear across the contacts if it were not there (a position that you all seem to accept). In both cases, if the arc were to occur, the energy would dissipate in the arc (again, a point you all seem to accept). In the second scenario, it would be wrong to claim that the capacitor’s only function is to suppress the arc just because no resonance actually occurs. It is still a resonant circuit, just completely damped. The capacitor is an essential element of that resonant circuit. An analogy would be a bell submersed in treacle. It is all but impossible to get the bell to ring (ignoring the impracticality of actually striking it), but the bell is still there and it is remains a resonant structure.

So what of the alternate viewpoint (which is apparently supported by Emerson and Williamson)? There are clues in the references themselves. Emerson predates ISBN numbering so that means the reference is pre 1967 (when ISBN started). Williamson is old enough that it still uses ‘thousands of cycles per second’ instead of kilohertz for frequency. Hertz was adopted by the Conférence générale des poids et mesures in 1960, though it did not become universal for many years after that. From a literary point, technical texts used ‘thousands of cycles’ before ‘kilocycles’ became common place so Williamson may be even older than that.

The significance of this is that, historically, the design of ignition coils then was very different to today. The first battery powered ignition systems, used nothing more than a Ruhmkorff coil often mounted in a nice polished mahogany box! The car industry renamed the ‘interrupter’ to ‘trembler’, but it was essentially the same. The basic design of the coil remained unchanged until the 1980’s though the trembler became the contact breaker in the distributor; the mahogany box had long since been replaced by an unpolished aluminium cylindrical container and doubtless the materials used in the internal manufacture had improved.

The coil was constructed using a soft iron cylindrical bar core with the primary wound onto it and the secondary wound on top of that. In more recent times, the primary and secondary are wound onto a plastic former which is placed over the core on assembly. Importantly, the ends of the iron core were not closed into a complete magnetic circuit. This cylindrical structure fitted nicely into that cylindrical can (and is probably what dictated the shape of that can). A not so obvious problem with this method of producing a coil assembly is that not all of the magnetic flux produced by the primary intersects the secondary. In fact less than 40% of it does (exact figure depending on coil geometry). The position is complicated because any current in the secondary produces its own magnetic flux which ‘pushes’ the primary flux from the windings reducing the coupling even further.

This means that the secondary has far less effect on what happens in the primary than would be the case with our ideal system above. Obviously, for an open circuit secondary, nothing changes apart from the somewhat lower inductance relative to resistance that the primary will possess. If the secondary is short circuited, when the contacts open, the secondary cannot damp the oscillations in the primary in the same way as the ideal case. The primary displays the decaying sinusoidal voltage and current, with a proportion of the energy determined by the coupling factor coupled out of the primary system during each half cycle. The primary displays a partially damped decaying sine wave.

If the secondary is connected to a suitable spark gap (call it a spark plug), the voltage applied follows the decaying sine wave. However, the position is complicated because the spark gap is an open circuit up to the point the gap breaks down and a spark is established. The result is a series of sparks each time the secondary voltage exceeds the gap breakdown voltage. This mode of operation is not ideal because each spark of the series occurs for a relatively short time and the time taken for the voltage to build up to breakdown each time is effectively wasted. This is the mode of operation that those old sources are referencing.

For decades, this method of operation was considered adequate, because it worked well enough for the engines of the time. Little thought was given to improving the performance of engines (except in racing circles). In any case, the car industry has always demonstrated a reluctance to change anything that does work well enough. But times were a changing. Concern was being expressed over pollution from engines and, as more time passed, the escalating pace of fossil fuel consumption. There was now the necessary impetus that engine design had to change and ignition systems along with it.

In order to produce engines that had lower pollution and better fuel economy, a better and longer burning spark was required. Turning the ignition coil into as close to an ideal transformer as possible, was not going to work because the primary side was too over-damped during the spark for effective operation. What was required was a coil design where more of the primary flux intersected the secondary, but not all of it. Increasing the coupling will increase the energy transfer, but only up to a point. Increasing the coupling increases the energy transfer but increasingly damps the decaying sine wave. The energy transfer continues to improve after the primary's decaying sinusoid is no longer significant. But past the optimum point, the energy transfer falls off as the primary becomes over-damped. In reality, the best operation occurs with a coupling between primary and secondary of a little over 90% (the exact value is dependant on circuit parameters).

With this degree of coupling, the secondary has almost as large an influence over what goes on in the primary as in the ideal case. Although the capacitor and primary inductance still form a resonant circuit, the primary current does not follow a decaying sine wave (at least to start with) although it does try to do so. Once the spark strikes, the energy is coupled out of the primary to the secondary, but the deliberately imperfect coupling does not over-damp the primary, but allows the maximum energy to couple to the secondary giving a single sustained spark until the secondary voltage drops below the quench level. From this point on, the residual energy forms a decaying sine wave because the secondary no longer influences the primary as its current has fallen to zero. Of course this decaying current induces a similar decaying sinusoidal voltage in the (otherwise) open circuit secondary which remains open circuit as there is insufficient voltage to restrike the spark.

Modern ignition coils have a more or less closed magnetic circuit otherwise the required degree of coupling cannot be obtained. The better coupling also means that a coil of a smaller physical size has a similar primary inductance to the cylindrical coil described above.

Now an important point here is that it still cannot be claimed that the only purpose of the capacitor is to suppress the arc at the contacts, simply because there is no oscillation while the spark is burning. The capacitor forms a resonant circuit with the coil. It is just that any resonance is heavily (but not totally) damped (perhaps putting that bell above in jelly rather than treacle).

As for references. There are plenty of references describing the traditional coil design and its operation. There are a few references for the modern designs (and, of course, they are different). What seems to be missing is anything covering how we got from one to the other. There are plenty of blogs and discussion forums where the participants are arguing over whether the capacitor and coil form a resonant circuit (generally claiming that it is not based on the evidence of a lack of decaying sinusoid while the spark is burning, but ignoring the give-away decaying sinusoid once the spark has extinguished).

The above is straightforward physics and is based on my over 40 year’s experience in the world of electrical engineering with the evidence of the available references. These usually inconveniently describe what is happening but not why it happens the way it does. The older design of coils is a case in point. Few people understand the significance of the relatively poor coupling between the primary and secondary and the severe limitation to the influence that the secondary has on the primary. Without the intervening references, I'm not sure how you are going to reconcile the article with reality though you could document the old and the new, though I can see that could cause problems with those that have not joined the dots either. 86.149.141.166 (talk) 14:27, 31 March 2016 (UTC)[reply]

Thanks for taking the time to delve into this matter. I'll give a quick reply right now.

I believe you misunderstand both my position and Elektrik Fanne/85.255.232.231's position.

All sides believe there are damped oscillations. The question is their significance: what function do the resonances serve?

Neither are talking about early ignition systems with Ruhmkorff coils. I haven't looked too closely because it isn't needed for the current debate, but Kettering's '492 patent seems to be an ordinary induction coil (21, 22) with separate interrupter (41 thru 45) and capacitor (46). There's a rotor (23, 24) that drives a switch (25-29) that looks like the rotor/points in a more modern system, but the function of the rotor and switch appears to be turning off the spark generator during most of the engine cycle. The rotor switch (25-29) is not protected by a capacitor, and it does not need to be because it is not the switch that ordinarily initiates the spark -- that task is done by switch (42, 43). This is not the system that EF/86.149 and I are discussing. It has similar characteristics, but it is not the typical 1960s inductive ignition system.

EF knows about the construction of the coils and that the typical pre-CDI ignition coils have better magnetic coupling than induction coils. EF knows about leakage inductance.

Your explanation using an ideal transformer is flawed. Ideal transformers do not store any energy. In an ideal transformer, if there is a secondary current, then there must be a matching primary current. Please reconsider your statement that "Once the spark strikes, the energy is coupled out of the primary to the secondary"; (to first order,) the energy is not stored in the primary or the secondary; it is stored in the coil's flux. I believe EF knows about magnetizing inductance.

Glrx holds the purpose of the capacitor is to suppress the break arc completely so the secondary arc can happen. The capacitor keeps the contact voltage low enough during separation so they do not arc. The capacitor also introduces a lower frequency RLC / damped wave, but that is an artifact rather than a primary purpose of the capacitor. The damped wave nature is never a big issue. The secondary arcs before the first quarter cycle has occurred. The waveforms do show damped oscillations, but they are not fundamental to the operation.

EF believes there is a break arc: "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."[1] EF does not explain how this break arc is extinguished or how there can be a secondary arc while there is a primary arc. EF does understand that the arc initiates at a high breakdown voltage and then falls to a lower sustaining voltage. EF does not explain the purpose of the resonance.

85.255 does not understand the break arc issue; 85.255 believes my capacitor explanation is only about contact life: "Your theory only claims arc suppression which would mean that the circuit would work without the capacitor, but the contacts would have a short life."[2]

Do you have dates/publishers for Emerson and Williamson?

Glrx (talk) 19:21, 31 March 2016 (UTC)[reply]

Thank you for your response. It has to be said, that it is difficult keeping track of who says what. Forgive me if I attribute something incorrectly but the essential point is that smeone said it.

My reference to the Rhumkorff coil was only in the context that the 'transformer' part of the coil design design had not fundamentally changed for decades. I never discussed its operation in this context.

I do not follow your point about the ideal transformer explanation being flawed. I said, "The secondary current mirrors the primary (because this is an ideal transformer) with a magnitude in inverse proportion to the turns ratio", a point you reiterated, but in a manner that suggested I violated this principle. At no point did I discuss this ideal transformer driving a spark gap.

Your statement that an ideal transformer cannot store energy is not only flawed but incorrect. When the contacts close, a current builds up in primary. This current creates a magnetic field in the core. When the contacts open, the primary inductance attempts to keep the current flowing. Lenz's law tells us that the collapsing magnetic field induces an emf that opposes the change. If there is emf and current there is power. With no other power source present (because the contacts are open, there is no power from the battery at this time), this power can only come from stored energy within the transformer. Thus, in this application, the transformer (be it ideal or otherwise) stores energy.

The protagonists to this debate all have their own ideas about what does what and it is clear that none of you are seeing the complete picture over the evolution of the system, just some fragments and some of those are not necessarily correct (which is why, I suspect, that there is such a contradictory picture, even from a single contributor).

I concur that the presence of the capacitor suppreses the arc that would otherwise occur at contact separation. However, I am adamant that that is not its only function. [Putting aside the older loosely coupled coil] I maintain that its presence forms a resonant circuit with the primary inductance (and most of the secondary inductance while current flows in it, which lengthens the discharge). The resonant circuit is there even though while the spark is present it is not permitted to act as a resonant circuit (but only in a relatively tightly coupled coil). I can say that the circuit would not work well, if at all, if there was no capacitor, but there would be an arc at contact separation. Any statement that the sole purpose of the capacitor is to suppress the arc is just plain wrong, even though that is how it seems to manifest itself during the relevant phase of the operating cycle.

As to whether there is a small spark with the capacitor present? Theory would suggest not as the capacitor is effectively a short circuit at this point. However, in the back of my mind, I do have some recollection of a small spark at the contacts, but as it has been a couple of decades since I have owned (and taken to bits) a petrol engined car, my recollection could easily be faulty on the matter. As you suggest, what happens after the spark extinguishes is immaterial to the operation as such, but it does provide clues as to what is trying to happen.

I have no information on either Emerson and Williamson (and the internet draws a complete blank, but this is not unusual for very old works of this type. Even my university text books are unknown but I know that they did exist when I bought them 'cause I still have most of them!), but I have to assume that they say what is claimed. Of course, there is always the question of whether the context changes the interpretation. I might find out if my local library can rustle up copies. 86.149.141.166 (talk) 17:26, 1 April 2016 (UTC)[reply]

Thanks for your response.

Ideal transformers do not store energy. If i×v is shoved in the primary, then i×v pops out of the secondary. If there is no current in the secondary, then there is no current in the primary. With no primary current, there is no magnetic field and no stored energy. There's no flyback with an ideal transformer.

The model you want is an ideal transformer with a magnetizing inductance but no leakage inductances. For simplicity, ignore all the winding resistance -- they only turn undamped oscillations into damped ones.

(An even simpler model deletes the ideal transformer and just uses a flyback inductor. That model makes it clear that the switch contacts must separate quickly and not break down.)

I disagree about your complete picture assessment.

The primary purpose of the capacitor is break arc suppression. The statement is sourced:

Taylor-Jones, E. (1921). The Theory of the Induction Coil. Pitman & Sons. 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".

How does the capacitor participate in a primary resonance when the arc is active? To first order, the secondary voltage is clamped at the arc sustaining voltage. With an ideal transformer, that means the primary is also clamped at a fixed voltage; dv/dt=0. The capacitor is parallel with the primary, so the capacitor voltage is also fixed. i = C dv/dt means that no current flows in or out of the capacitor. The capacitor is not resonating with anything. There isn't a primary resonance to "lengthen the discharge".

To first order, the duration of the discharge is set by the removal of energy from the magnetizing inductance. The fixed magnetizing inductance L sees the fixed arc voltage V; V = L di/dt means di/dt is a constant. The current through the arc decays linearly until it falls below a sustaining value. EF provided a ref with a beautiful picture of that linear decay. The duration of the arc can be increased by increasing the inductance or the initial energy. There no resonance. (And there are not multiple sparks as your initial post claims.)

My previous posts discussed the make arc that discharges the capacitor. The make arc has no impact on the spark.

If a book is a significant authority, it should be held in a library somewhere. WorldCat indexes many library holdings, but has neither book. Taylor-Jones (1921) doesn't have an isbn, but it can be easily located.[3]

Glrx (talk) 22:57, 1 April 2016 (UTC)[reply]

Thanks again for your response.

Maybe my attempts at trying to simplify my points using an ideal transformer were too simplified. I reckon you could also argue that since all of the energy input to the primary is output from the secondary (because it is ideal), there is none left to create a magnetic field in the first place. However, this is theoretical component that does not exist in rerality. I did suggest that my ideal transformer was not actually ideal.

This does demonstrate why a near ideal transformer (which can be constructed) is not actually very useable in this rôle.

I am not following your paragraph how the contact arc stops any resonance between the capacitor and the primary inductance. If the capacitor suppresses the arc then there is no arc. This is the case whether the transformer is near ideal or far from ideal.

If we were to use a coil that has no secondary, would you agree that on contact opening, the LC circuit would produce a decaying sine wave? You should because it does. If a secondary coil is then provided, but does not couple particualrly well with the primary (and in the linear construction typical of the older style ignition coil, remember that the coupling is less than 40%), you must conceed that the secondary however loaded can only partially damp the primary oscillation which will succeed in manifesting itself (and we do appear to have sources - if only they could be located). Only as the coupling is improved is a point reached where the oscillation fails. Of course, the oscillation does start to occur until the spark strikes (and you did state this yourself).

Isn't it amazing how a circuit so simple can generate much debate?

I found a book cataloguing website, but unfortunately it allows a search by book title or author, but not both. 'Automotive electrical systems' generated over 3700 hits (but it is a rather obvious book title). J T Emerson pops up 678 hits. Pity I can't cross-reference them. 'An engineer's guide to vehicle electrical systems' does something very strange and turns up a whole raft of audio recordings that have no relationship with the search phrase (or subject) that I can discern. Hmm. 86.149.141.166 (talk) 17:39, 2 April 2016 (UTC)[reply]

Hello, again.

Yes, if you push any power into the primary, then all of that power pops out of the secondary. There's no power deficit to integrate dt into an energy storage. Your model did not include a proper energy storage element.

I don't follow this remark. Your ideal transformer model does work if you include the magnetizing inductance. All real world transformers have a magnetizing inductance. It's also OK to assume the coupling is unity and the leakage inductances are zero. That will make a decent flyback transformer. It would be very useful in this role.

In an LC resonance, the L and C move the stored energy back and forth. If the voltage across the capacitor is fixed, then the amount of energy it stores is not changing -- there's no more back and forth energy transfer or resonance. It also means that the current through the capacitor is zero. While the points are closed, the capacitor is discharged to 0 V and the inductor is fluxed. When the points open, the energy trade starts; the inductor starts charging the capacitor. As it does, the inductor loses energy; the lost energy appears on the capacitor as a voltage. There are no losses in this model, so the capacitor voltage starts out as an undamped sinewave that will have a peak amplitude of about 350 V. The secondary's peak amplitude will be 100 times that, or 35,000 v. But hold on! We never get to the first peak of that sinewave. The rising voltage across the capacitor puts a rising voltage across the secondary, and the spark initiates. What happens? Say the arc initiates at 10 kV and rapidly falls to 1 kV sustaining. During that time, the primary voltage must fall from 100 V to 10 V. That means the arc rips a lot of charge out of the capacitor; that charge does not flow through the magnetizing inductance -- it goes straight across the ideal transformer to the arc. Also during that fall, the magnetizing inductance is spitting out its continuous current. That current would have gone to the capacitor, but now it is redirected to the arc. The primary stablilizes with 10 V across it, the capacitor current is zero, and now all of the arc current is provided by the magnetizing inductance. It falls linearly until the arc extinguishes. At that point, most of the original stored energy is gone. What little is left in the capacitor and the magnetizing inductance now resonates.

Here's the wonderful link the Elektrik Fanne provided.

• http://www.worldphaco.net/uploads/CAPACITIVE_DISCHARGE_IGNITION_vs_MAGNETIC_DISCHARGE_IGNITION..pdf

Look at figure 9. Top trace shows a wild damped sinewave oscillation at switch off, but it is at a much higher frequency than the 2 kHz primary resonance. (Page 10 explains this ringing as leakage inductance.) You can see the 2 kHz primary damped oscillation after the arc extinguishes; it was supressed during the spark; note the constant primary voltage during the later stages of the arc. Middle trace shows some initial high frequency ringing (above the primary/secondary resonances) but then the linear disfluxing of the inductance.

I've never disputed that an LCR makes a damped sinewave. What is the purpose of making a damped sinewave (or even an undamped sinewave)?

You need to be careful here. The coupling of the solenoid induction coil is much different from the typical 1960s ignition coil. The Ford Model T used the induction coil design, but more modern designs have a near-complete magnetic path; there's an ignition coil cut away above. See worldphaco.net page 10 that states, "This is because of the tight coupling of the primary tuned circuit to the secondary tuned circuit." Consequently, I won't concede the lightly coupled issue (but it's also not that important). Taylor-Jones studied induction coils; his more involved oscilation theory made use of the poor induction coil coupling by tuning the primary resonance to twice the secondary resonance. He used a coupled primary peak to boost the secondary output to over 0.25 MV. The ignition system does not need to play that game.

Glrx (talk) 20:21, 2 April 2016 (UTC)[reply]

Thanks once again for your response.

This ideal (or otherwise) transformer is becoming a distraction. I had never intended to embark on a discussion of this but to merely illustrate a point (even if I did not do it particularly well). I suggest that we leave it.

Elektrik Fanne's reference that you linked to is interesting because it makes reference to "the primary winding's tuning capacitor", a clear reference to a resonant circuit. Also, "This electric field energy, if not dissipated another way [through the secondary to spark] then returns to become magnetic field energy again in an alternating or oscillating way because the capacitances and inductance form an oscillatory or resonant circuit", stated as a fact. Later on when discussing all of the circuit elements of the ignition coil, it says, "[after discussing the coil capacitances] There is also the capacitance, typically around 0.22uF (sic) of the contat breaker capacitor which "tunes" the coil primary winding", again not only implying a resonance, but there follows a table of data for an actual coil listing the resonant frequencies of the various parts. In fact the ignition coil model shown shows two very obvious parallel resonant circuits (plus the not so obvious one involving the contact breaker capacitor.

The reference also supports my point that older coils are built where the windings are built on "an iron bar", not unlike an induction coil. It also mentions that the iron bar design has a higher leakage inductance (or poorer coupling - though does not quantify it). It also discusses later these flux lines not linking the two coils. This was also discussed by 85.255.232.25 elsewhere, but he failed to mention the evolution from the bar type coil (which he seems to be discussing) and the more modern closed magnetic path type coils.

Figure 9 shows these resonances in action. It shows a much higher frequency one when the spark is burning, followed by a lower frequency one when the spark is extinguished.

From all of this and my knowledge and experience, I have to declare that your reference, "The theory of the induction coil" is an unreliable reference because it is incorrect to claim that, "the only use of the primary condensor is to check the formation of an arc at the interupter". It does 'check the arc', but it also is part of a resonant circuit so it is not the only use. You set the precident on declaring references unreliable because they are wrong when you dismissed a reference from Elekric Fanne over at Talk:Induction coil, which was actually correct on the point he was making. And in your response you agreed that the L, C and R form a 'tank circuit'.

If you are not going to accept any of this, then I can only suggest at this point that all we can do is agree to disagree. Let us see if any of the other contributors come back.

BTW, when I searched for 'An engineer's guide to vehicle electrical systems', and turned up a raft of unrelated audio recordings, although I did notice at the time that they were recordings of interviews on totally unrelated subjects, I did not notice before I saved my last post, that down in the details the first three recordings on the list showed the interviewer as one 'David Williamson'. Strange, considering 'Williamson' was not part of my search ctriteria. The database engine was clearly doing something screwy but I do not suppose that, given the unrelated subject matter, that it is the same D Williamson, though it does seem a very odd coincidence. 86.149.141.166 (talk) 16:38, 4 April 2016 (UTC)[reply]

Thanks again. We are not far apart.

The ideal transformer is a reasonable thing to use. Models of real transformers use it. EF's link uses an ideal transformer in figure 6. Casting it aside does not make sense.

Taylor-Jones is an excellent reference for induction coils. Taylor-Jones describes the Rayleigh theory and the purpose of the capacitor. Taylor-Jones, when explaining his oscillation theory, describes why choosing resonant frequencies is important to the design, but resonance is not important in the Rayleigh theory. Distinguish the notions of purpose and artifact/side-effect. The purpose of the capacitor is to suppress the break arc. A side-effect of the capacitor is a resonance, but there is no design purpose to that resonance. That's what TJ means with only use. (The capacitor also limits the peak voltage, but that is not an issue if the spark gap is present.) Worldphaco is a limited reference; it gives some wonderful component values and pictures, but it does not address the contact breakdown physics at all (its focus is more modern ignition systems). The contact breakdown voltage with separation x, the contact opening speed (dx/dt), and the voltage across the points (dv/dt) are all related. Understand that relationship, and the capacitor choice is clear. EF's sources do not address that relationship.

EF's link may refer to a "tuning capacitor", but it does not offer any purpose of the tuning or describe how the tuning improves the circuit. Why does the coil have to be tuned at all? Yes, put a capacitor and an inductor together and you get a tuned circuit. But that does not mean the purpose of putting the capacitor in the circuit was to create a tuned circuit. The article models the resonances, but that does not mean they have a purpose. The linked article only mentions "breakdown" once, and that is late in the game about the Boyer adaptation.

Please look at the coil construction link supplied at the "discussed elsewhere" link you referred to:

• http://firetrucksandequipment.tpub.com/TM-5-4210-230-14P-1/css/TM-5-4210-230-14P-1_543.htm

The text I said with it was:

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."

Notice the return path laminations on the interior walls of the case. The magnetic path is broken in two places for the convenience of insulating the HV; the magnetic core is not just a simple bar.

(Maybe I should have responded to the "discussed elsewhere" diff's new text, but my position had been stated. Yes, editors may make picture of objects without running afoul of WP:OR, but an editor may not take a picture of a dog (induction coil) with a broken leg (break arc) and conclude that all dogs have broken legs. I didn't disagree (and never have) with the damped sinewave, but there is a problem with the coupling argument in the 2nd and 3rd paragraphs.)

Re figure 9. I'm an electrical engineer, too. I understand the origin of the HF ringing at the start of the spark. I understand why the LF ringing is not present during the spark. I understand why there is a positive peak at B.

Re the agree to disagree. The only significant issue separating us is the assignment of purpose to the capacitor. We agree on contact protection. We disagree on resonance. No source has explained a purpose for a resonance. All the energy has already been stuffed into the coil; we don't need a resonance to build up any energy (such as in shattering a wine glass).

Glrx (talk) 19:41, 4 April 2016 (UTC)[reply]

As I suggested, let's see what others views are. 86.149.141.166 (talk) 10:48, 6 April 2016 (UTC)[reply]

As far as I am concerned, there is no shortage of references that claim that the capacitor forms a resonant circuit with the primary and I added a raft of them before they were deleted by someone else. --Elektrik Fanne 13:25, 7 April 2016 (UTC)[reply]

I concur. However, I would go one step further and point out: that because a reference does not mention the resonant circuit that does not make it a cite that it is not a resonant circuit. It only proves that the author either chose not to mention it or was unaware of it (which cannot be determined). 86.149.141.166 (talk) 12:49, 9 April 2016 (UTC)[reply]

• Huh? There is no doubt that a capacitor and an inductor form a resonant circuit. That is not the issue.

What is the purpose of the capacitor?

If you believe it is to form a resonance, then where is a reference that explains why making such a resonant circuit is a reasonable thing to do?

One of Elektrik Fanne's deleted refs shows that the L_m C resonance is killed during the spark. If the resonance is important, then why does the resonance cease during the spark?

Glrx (talk) 17:16, 13 April 2016 (UTC)[reply]

Answered above. But as you obviosly missed it: it is because the resonance is damped by having the energy coupled out of it. — Preceding unsigned comment added by Elektrik Fanne (talk • contribs) 16:45, 20 April 2016 (UTC)[reply]

I didn't miss it. Your claim is the capacitor is added to create a resonance; that suggests there should be some purpose to the resonance. I posed two questions about that purpose: (1) why is the resonance important in the first place, and (2) if the resonance is important, then why is the resonance not manifest during the spark.

Saying the resonance is damped during the spark answers neither question. Nothing you said goes to the ultimate purpose of the capacitor.

The first question is to focus on the purpose of the resonance. (It's not the resonance that is important, but it is keeping the contact voltage controlled to avoid a short arc breakdown. That requires keeping dV/dt low while the contacts separate at dx/dt.)

The second question is to point out that the capacitor's purpose is finished once the spark ignites. The arc voltage settles to a quasi-constant voltage so the effect of the capacitance disappears (is shorted out). If we ignore the second-order leakage inductance ringing while the initial breakdown voltage settles to the sustaining voltage, there's no visible resonance during the spark, so the capacitance has no effect during most of the spark. Resonance cannot be important during the spark, so any design purpose it had must have been before the spark -- when the cap was suppressing the break arc. That's why Taylor-Jones explains the capacitor's only purpose is arc suppression.

Glrx (talk) 17:31, 20 April 2016 (UTC)[reply]

As has been pointed out, there are no shortages of references that state that the capacitor and coil primary form a resonant circuit. That should end the matter. Your opinion is precisely that - your opinion - and WRONG. Taylor-Jones is therefore an unreliable source because he is wrong to claim that arc suppression is the capacitor's only function - for reasons already discussed. --Elektrik Fanne 15:41, 22 April 2016 (UTC)[reply]

Just think what is the complete path taken by the current produced in the secondary winding. (It is not via the coil casing). James Turner, Birmingham UK — Preceding unsigned comment added by 86.25.123.52 (talk) 19:39, 16 September 2016 (UTC)[reply]

So how exactly do you believe: that the secondary current gets back to the other end of the secondary coil? --Elektrik Fanne 17:07, 16 October 2016 (UTC)[reply]

You overlook the car industry's constant strive to design things to be as cheap and simple to make as is possible. Connecting the earthy end of the secondary coil to the vehicle chassis would be the electrically correct way to do it. However, car part suppliers have discovered that if the secondary coil is connected to the earthy end of the primary, the connections of coil can be made before it is inserted into its casing (usually plastic moulding these days). It also works because the capacitor offers a low impedance to the relatively small and fast changing secondary current.

The driver of your average family runabout is not going to notice the difference. On the other hand, high performance engines may well notice the difference the extra impedance makes and consequently earth the secondary coil to the engine casing. This is why circuit diagrams can be found which show either version of he circuit. 148.252.128.92 (talk) 16:33, 28 October 2016 (UTC)[reply]

• Note. Elektrik Fanne has been blocked indefinitely as a sockpuppet. Glrx (talk) 20:26, 17 December 2016 (UTC)[reply]

The references were removed before because they are blogs or do not support the claim. Supplying many poor quality references does not improve the source for a claim. Books that cannot be found in WorldCat or books whose focus is incidental are also inappropriate.

The purpose of the capacitor is to limit dv/dt across the switch contacts while they open. That's in Ott, cited above. The voltage must be kept below about 10 V (V_A) to prevent an arc occurring while the contacts separate at microscopic distances. Consider 10 V across a distance of 1 micrometer is a field of 10 MV/m. Furthermore, the voltage must be held to less than 300 V while the contacts are separated at the minimum of Paschen's law. If the voltage is too high, then the air breaks down, there's a glow discharge that allows some current to flow, that current then increases into a low-voltage arc discharge. The arc discharge would limit the primary voltage to a small value; the secondary voltage would also be limited by the transformer action. The simple way to achieve that goal is to have 300 V * N exceed the spark plug breakdown voltage.

Now, the addition of capacitor also happens to cause a resonance, but that is not a design goal in the automobile ignition system. A resonance is exploited in exotic induction coils, but that is not the case here. That is also mentioned in the above discussion. That resonance is prevalent if there is no spark (that's what the Harvard lab notes talk about; also phaco fig. 8). In functioning ignition systems, the resonances that are seen are due to leakage inductance. (phaco p. 15.) The resonances are on top of the magnetizing inductance; phaco points out that the resonance never interrupts the spark current. (phaco p. 15.) The resonance is not a design goal; it is not a design goal.

The article's statement ("As a result of this resonance the duration of the spark is sustained and so implements a good flame front in the air/fuel mixture") is unsourced and just wrong. The spark is maintained by the magnetizing inductance; the linear decay of is clearly visible during the later portions of the spark in phaco figure 9, middle trace, spark current from E to F. If the "resonance" is sustaining the spark, then why is the spark still going when the resonance has decayed away?

Glrx (talk) 18:12, 24 October 2016 (UTC)[reply]

You are the only person who seems determined that resonance has nothing to do with it. As such you are also keen to dismiss two references that discuss the subject as 'incidental' plus others that specifically mention the point (though large numbers of references saying the same thing are pointless). The resonance angle is too well referenced, to be ignored. The fact that you cannot find the books does not preclude them. WP:AGF says that you have to accept them, especially as they appear to specifically discuss the subject. I cannot locate them either (but they seem to be old books, so this may be why), but in the absence of any indications to the contrary, I have to accept them. 148.252.128.92 (talk) 16:33, 28 October 2016 (UTC)[reply]

STOP PRESS: Although my local library does not have copies of either book, they have just phoned me to say that they have located a copy of the Williamson book and are getting it for me (so at least it exists). 148.252.128.92 (talk) 16:47, 28 October 2016 (UTC)[reply]

@Glrx: Thank you for your patience during these prolonged and sometimes repetitive discussions. If nobody is able to satisfy the WP:BURDEN of providing citations that directly support the disputed material, I support its removal. Burninthruthesky (talk) 16:29, 31 October 2016 (UTC)[reply]

Rejecting a citation because you have not seen it is a violation of WP:SOURCEACCESS in the Verifiablity policy. The AGF policy does require that if one editor asserts that an offline source supports a fact, we must all accept it unless we hve seen it with our own eyes. Editors might agree that there are multiple other sources that contradict the offline source, but those other sources would have to be cited by you. If you have nothing to contradict Williamson and the Emerson citations, then you don't have any reason to reject it, and WP:BURDEN has been met. You can't tag a source as "fails verification" if you haven't seen it yourself.

Besides citing additional sources, if you can't find the two books in your local library, go to Wikipedia:WikiProject Resource Exchange/Resource Request and reuest another editor find it and check it for you, or ask their advice on tracking the book down. There's lots of assistance for settling something like this. --Dennis Bratland (talk) 17:18, 31 October 2016 (UTC)[reply]

@Dennis Bratland: Have you read and understood the material above? There are citations that contradict the claim.

See this delta whose before contains a cite about the single purpose of the capacitor and whose after is the vague claim about resonance.

The references that I have used cite to the purpose of the capacitor. The break arc must be suppressed; that is a French invention; Taylor states the purpose. Ott discusses contact arcing in both the resistive and inductive cases. Ott discusses a hybrid contact breakdown model the combines both short-arc and gas breakdown. He draws a graph that shows the breakdown characteristic as a function of contact separation distance from submicron to mm distances. He then redraws the x-axis to be time (integrate contact opening speed to get contact separation). He then describes using a capacitor to prevent short arc breakdown. The switch, before it opened, was feeding a constant current (I) through an inductor. Putting a capacitor across the switch limits the rate the voltage can rise (dV/dt) to the current divided by the capacitance (I/C). Use a large enough capacitance, and the contact voltage will stay below the breakdown portion of the curve.

Under normal operating conditions, that's all that the capacitor needs to do. Nothing is said about resonance. There can be another purpose of the capacitor: limiting the peak voltage across the contacts if the spark plugs are disconnected. In that scenario, one notices that the energy stored in the magnetizing inductance (0.5 L I²) will be shifted to the capacitance (0.5 C V²). Consequently, choose C to be large enough to store at least all the inductor's energy at the maximum desired V; conservation of energy does the rest. The issue is not resonance: there's no compelling reason to choose a particular resonant frequency. It's just recognizing that the inductor and capacitor will transfer energy back and forth. And it is irrelevant to the normal operation of the ignition system because the spark plugs must breakdown before reaching that maximum voltage -- otherwise they will never fire.

Leakage inductance introduces a much more subtle issue. The worldphaco source is good on that point but would be impenetrable to most readers.

Both Williamson and Emerson were cited without specifying publishers or dates. There were direct requests for that information. That is not a SOURCEACCESS issue but rather a source identification issue. I don't care about an isbn; I have plenty of books that don't have an isbn. Williamson is quoted, but the quotation shows Williamson does not understand what is going on. Williamson claims the energy "discharges into the condenser"; the energy is discharged into the spark gap. Emerson's quotation only states a resonant circuit is formed. Nobody disputes there's a resonant circuit. Hobbyist can see an L and a C and conclude there's a resonant circuit. The issue is what is the purpose of the capacitor. The reliable sources do not say it is to form a resonant circuit. It's to prevent an arc on the primary side so the spark will happen on the secondary side.

Glrx (talk) 18:50, 31 October 2016 (UTC)[reply]

Have it your way. There are easier ways to resolve this kind of content dispute, and resources are available to help you. You should not say a source fails verification based only on a short quote. If you haven't seen the book, you don't know. I've learned to not use quotes at all in my citations; without the complete context, quotes merely fuel debate rather than put it to rest. --Dennis Bratland (talk) 19:27, 31 October 2016 (UTC)[reply]

The AGF policy requires no such thing. Direct quotations of the relevant sources were duly volunteered in accordance with WP:VER. On that basis, the {{Fails verification}} tag in the article is absolutely correct. Burninthruthesky (talk) 10:22, 1 November 2016 (UTC)[reply]

What are you on about? The sources given directly discuss the resonance. Thus the 'failed verification' is not appropriate given that it has been verified (see next para).

I picked up the Williamson reference from my library this morning. The quoted passage appears om page 55 (not 54) in this copy, but it is the third edition so may have moved in a revision. This third edition was published in 1953 and appears to be a college text book written for the days when people actually repaired the electrical parts in your car rather than just swapped the entire assembly for a new one.

It is lost in the article history who originally wrote the paragraph, but given the discussion and that the paragraph is not attempting to document the modus operandi, has anyone considered some compromise wording? Since the resonance is double referenced, It can be mentioned along with what it does (cited by Williamson and one other ref). The point about sustaining the spark is unreferenced and can be deleted until a reference is found. The point about minimising arcing at the contact breaker is mentioned in Williamson, so I have added that as a cite, but deleted the one point not mentioned. 185.69.145.139 (talk) 17:01, 2 November 2016 (UTC)[reply]

Who is the publisher for Williamson? I want to be able to find it in WorldCat or similar index. The publisher suggests the stature of the work. Something published by Cambridge University Press / McGraw-Hill / John Wiley carries more weight than Haynes / Howard W. Sams.

You state that the book is about auto repair. "An engineer's guide to vehicle electrical systems" implies a reference work about designing electrical systems. A book about repairing such systems is much different: an auto repair guide must know how to fix systems but need not know the details about how they work. The quotation shows that Williamson is either careless in his explanation or confused.

If this "resonance" is important, there must be some reference that explains why the resonance is important. Thank you for deleting the sentences about the resonance sustaining the spark and implementing a good flame front because it failed your verification. But where is the design rationale behind the resonance?

Where do you see a compromise? Do a capacitor and an inductor form a resonant circuit? Yes. Is that the purpose of the capacitor across the points? No. Furthermore, which capacitance and inductance are we talking about? The secondary has a lot of stray capacitance: EF's worldphaco ref says 50 pF; refer that to the primary and it becomes 85² × 50 pF = -0.3 μF -- which is comparable to the 0.22 μF capacitor that is across the points. Also, which inductance are we talking about? If there's no spark, then the magnetizing inductance dominates; if there is a spark, it essentially shorts out the magnetizing inductance and leakage inductances are more prevalent. See worldphaco around page 11, "expected resonances"; one resonance is 2 kHz; another is 7 kHz. Does Williamson address leakage inductance?

The paragraph you edited is full of technical nonsense. WP cannot rely on sources such as MGAguru's personal website. I do not see Williamson as a reliable source for the design of ignition systems; the given quotation indicts his reliability.

I've provided sourced content that disagrees with the WP text. Taylor-Jones is old but reliable; T-J built a 250 kV induction coil back in the day, so he should know a thing or two about using inductors and transformers to generate sparks. Ott is a recognized source on contacts, Ott comes from Bell Telephone Labs, and the publisher is John Wiley.

This subject is not simple.

Glrx (talk) 19:31, 2 November 2016 (UTC)[reply]

That was partly my point for the way I CEed the section. Williamson and Emerson and others all mention the fact that the coil and 'condenser' (capacitor) are in parallel and thus form a parallel resonant circuit. Thus the point is cited (and indeed true) and can be included in the article. There is a cite already in the article for a claim about energy repeatedly being transferred to the secondary until exhausted that has been there for some time (so that stayed). But if the source is unreliable (you suggest it is WP:SELFPUBLISHed), then delete the claim and allow someone else to find a source.

For the two works (and many other sources) even to mention the resonance implies that it is important. However, Williamson does not expand the overall effect it has on the circuit operation and only leaves the implication (I have no idea what Emerson says beyond the quote). Therefore without any references as to that effect, the claim about sustaining the spark and extending its duration has to go unless a supporting reference can be found.

Unfortunately, having no further need for Williamson, I returned it to the library and failed to note who the publisher was before I did so. As I said: the book was clearly aimed at motor engineers perhaps a cut above your average garage mechanic. It did have quite a bit of technical detail on how the more complex parts of a vehicle's electrical system (battery, dynamo, starter motor, charging regulator etc. etc.) actually worked but without going into the sort of theory that a proper electrical engineering textbook might. It looked as though it was written by a lecturer on the subject as a course text book.

The library informed me that the only place they have found Emerson is the copy that is in the British Library, which they (or anyone else) are unable to borrow. Going there is just impracticable and I don't have the time.

The compromise was including all that was cited and removing that which was not. A capacitor and a coil in parallel (including capacitance and inductance elsewhere in the circuit but appears in the primary circuit) unconditionally forms a resonant circuit. This would remain true even if you were to find a way of totally 100% damping the oscillation somehow (not possible in practice). The resonant part of the circuit must still produce some resonance as there has to be something present for the damping to damp. If the oscillation were totally eliminated, the damping would have no effect allowing some oscillation to occur.

As I see it: the contention here is if this resonance is important to the operation of this specific circuit? This is what is not adequately supported by reference. Opinion seems divided. You seem to believe it doesn't, others seem to think it does. I have no idea, but it does somehow seem logical. A cite either way is required. Yoy state that you have provided such a cite. Where? What cite have you provided that specifically states that resonance has no part to play in the operation of the circuit? 148.252.129.151 (talk) 14:45, 8 November 2016 (UTC)[reply]

Thanks for the comments.

You won't find references saying that "resonance has no part to play". That's not how relationships are described. Just like you won't find a reference that says elephants do not get to vote in US elections. You should be able to find references that say what the purpose of a component is. I don't want a reference that simply says the "sky is blue" (there is a resonance); I want a reference that describes the sky is blue due to short wavelength (e.g., blue) light being scattered and long wavelength (e.g., red) light not being scattered.

Here's my ref that got deleted:

Under the Rayleigh theory, "the only use of the primary condensor is to check the formation of an arc at the interrupter". (Taylor-Jones 1921, p. 17) harv error: multiple targets (2×): CITEREFTaylor-Jones1921 (help) (emphasis added)

• Taylor-Jones, E. (1921). The Theory of the Induction Coil. Pitman & Sons. {{cite book}}: Invalid |ref=harv (help)

Taylor-Jones does not say the purpose (or the secondary purpose) of the capacitor in the Rayleigh is to form a resonance. Taylor-Jones does explain that the capacitor is a disadvantage in the Rayleigh theory. Lord Rayleigh got a much better spark when he did not use the capacitor but rather used a rifle bullet to interrupt the primary current (create a large physical gap and blow out any arc while you're at it). Rayleigh and others understood that the energy storage characteristics of inductors and capacitors will set a maximum voltage, but there is no preference for choosing a particular resonant frequency (other than lower capacitance allows higher voltages for the same energy). If you cannot use rifle bullets, then you must use a capacitor. Furthermore, Taylor-Jones knows to discuss when resonance is important because he has a much more exotic induction coil theory that resonates the primary to 3 times the secondary resonance to achieve an in-phase voltage addition at the secondary (Taylor-Jones 1921, p. 38) harv error: multiple targets (2×): CITEREFTaylor-Jones1921 (help); T-J shows how to exploit resonance to maximize output voltage. (That is not a concern in an automobile ignition system; it is not seeking a world record spark.)

Ott describes the purpose of the capacitor is to keep the voltage across the contacts below their breakdown voltage. "if a discrete capacitor is placed in parallel with [the contacts], the peak voltage and the initial rate of rise of contact voltage can be reduced to the point where no arcing occurs." (Ott pp. 184-185.) The capacitor is not going for a resonance but rather reducing the rate the voltage rises across the contacts and the maximum voltage. The purpose is not to create resonance but rather to suppress arcing.

Ott also describes that once an arc starts, it is a lot harder to stop it. A DC arc can be stopped by reducing the voltage below the arcing voltage (about 12 to 18 volts) or the current below the minimum current (about 400 mA to 700 mA but possibly 40 mA). (Ott p. 176.) That's will be difficult to do with a magnetizing inductance fluxed to 4 A.

Glrx (talk) 19:11, 8 November 2016 (UTC)[reply]

We have already established that Taylor-Jones is an unreliable source. That he does not mention resonance proves nothing. That many sources do must mean it is important otherwise, why would they bother? This has been discussed to death and I see no point in just repeating the points that you are reluctant to accept. I have not read your attempted technical explanations they are unreferenced (read: snow job). --Elektrik Fanne 14:07, 11 November 2016 (UTC)[reply]

• Note. Elektrik Fanne has been blocked indefinitely as a sockpuppet. Glrx (talk) 20:24, 17 December 2016 (UTC)[reply]

It has been proposed to merge Ignition switch into this article (Ignition system).

• Oppose. Ignition system deals with igniting the fuel-air mixture in the cylinder. The ignition switch (aka starter switch) has other duties such as powering all electrical accessories and temporarily engaging the solenoid/starter. Glrx (talk) 19:52, 5 November 2016 (UTC)[reply]

This article needs a lot of work. I started fixing some stuff in one section, but then realized the material was incorrectly covered in the previous section. The Model T magneto is an alternator in today's terminology. The Model T didn't need to use rectifiers to make DC (which could recharge a storage battery) because the voltage peaks were synchronized for powering the induction coils. Glrx (talk) 18:33, 28 August 2017 (UTC)[reply]