Forgive my ignorance, A/M ignition

[sv51macross]

A thread about a MSD ignition module on RX7club got me thinking; I have little understanding of how these units work and how they are integrated with the engine.
ECU sends signal to coil, coil fires, the uber-high voltage goes through the plug wires and/or directly into the plug where the voltage sparks across the gap and grounds through the block.
But where do these ignition modules figure in, and what do they do? I take it they aren't coils themselves, but controllers? And a faq on AEM's site claims that their twin-fire module was designed with rotaries in mind...?
Might someone please educate me on the purpose of these A/M modules? What do they do that upgrading the coils/wires/plugs [and ensuring proper connections/grounds] doesn't?

[NoDOHC]

I don't know about AEM A/M modules, but the MSD is typically a capacitive discharge ignition system. It drives the ignition coil at a higher voltage, but with comparable energy. This results in a high spark driving voltage but very short duration spark. At low speeds, CDI (Capacitive Discharge Ignition) systems will typically deliver multiple sparks in case a single, short-duration spark fails to ignite the charge.

CDI used to be the best way to get high energy ignition. Modern Inductive ignition systems are very comparable and are probably superior in boosted applications. The CDI system can usually be easily retrofitted into the system or even piggy-backed on the factory ignition system. High-energy inductive ignition coils (like LS2 Coils, etc.) will require a revamping of the entire ignition system and probably an aftermarket ECU.

[sv51macross]

So then, a CDI module would go between the ECU and the factory coils? And this works on stock wiring? [well, as you mention, it just increases the voltage. I assume amperage goes down].

[RETed]

Let's go back to basics...

The (stock) ECU fires a low voltage signal to the ignitor.
The ignitor fires a higher voltage signal to the coil.
The coil - using a design of a step-up transformer - steps that signal up to voltage levels exceeding 10,000 volts.

There's some reasons for such a complicated design.
1) You don't want too high voltages running through the delicate electronics of the (stock) ECU. High voltage (pulse) emit EMI. EMI wreaks havoc on low voltage signals - which is what you mostly see internally in the (stock) ECU. Also, high(er) voltage signals require high(er) power amplier circuits that that use more juice - having lower power requirements allow the (stock) ECU to run cooler with less power consumption.
2) Having the higher power components outboard of the (stock) ECU allows you to replace them a lot easier and faster. Higher power components will tend to have less reliability and longevity. Would you like to keep replacing the (stock) ECU if the on-board (high power) ignitor circuit died?

The stock ignition system is called an inductive ignition.
This is typical of most OEM ignition systems.
Inductive ignitions systems have the advantage due to the fact that 1) they are very reliable, and 2) give good combustion burn due to the relatively long spark dwell.
Downside of an inductive ignition system is that it has harder time firing at higher RPM's (and combustion loads).

Let's look at higher RPM's...
As RPM's rise, the window of opportunity of firing the spark gets smaller and smaller.
This is due to the fact that the rotor is spinning faster and faster, and the ignition system has to fire within a narrower time window.
The inductive ignition system has to "charge" or "rise time" of the circuit before it can effectively fire a spark.
If the coil does not get the minimum voltage charge, it cannot fire (or discharge) a spark.
This is called a "misfire".

This is where capacitive (discharge) ignitions come in...
MSD (multiple spark discharge) built their company on this concept.
By using (a bank of) capacitors, you effectively eliminate the problem of charge time at high revs.
Capacitors can build up voltage in very short time - way faster than purely inductive systems.
Now, this is a double edged sword...
Because the capacitor charges so quickly, it also discharges very quickly...
A CDI (capacitive discharge ignition) box disadvantage is that it can only fire a very short duration, high powered spark.
This is where the inductive ignition has an advantage - the longer duration spark is better for combustion.
This is where MSD comes in - multiple spark discharge...
By firing a bunch of sparks at low RPM's, it tries to compensate for the disadvantage of the capacitive ignition.

Most CDI boxes can input a signal from ignitors, intercept that signal - step up the voltage, and then output a much higher voltage to the coil.
Typical CDI boxes can step up the voltage up to 300...400+ voltages at output!
Inductive ignitions have very hard time getting voltages that high.

Now, CDI technology sounds real good, so what's the downside?
If this box fails - and it will - it will knock your entire ignition system out.
Due to the high voltage it runs at, it will eat spark plugs.
This can be somewhat compensated by running colder plugs.
Also, the fact that the CDI can fire large spark plug gaps, the spark plugs will last longer - relatively - because an inductive ignition system would probably misfire due to the wider gap.

Now with that out of the way, let's take a look at the stock RX-7 ignition system...
Some of you might already know this, but it's worth repeating.
I'm ignoring pre-FC3S ignition systems that all 1st gen's use - twin coil, distributor systems.
This applies to direct fire systems that the FC and FD (and possibly RX-8?).
This stock ignitions systems use a twin tower coil to fire the leadings and a pair of single tower coils to fire the trailings.
This system allows for a wastespark system which effectively fires three spark events per (rotor) chamber.
That 3rd spark event is also known as a far trailing spark.
This extra spark makes for more complete combustion.

What does this have to do with CDI?
A stock ignition system will have the characteristic brrrrrrrrr....brr.....brrrrrrr idle.
Anyone with an unmolested stock ignition system can relate to this.
That small hiccup is due to the twin tower leading misfiring cause the stock ignition system cannot fire reliably.
Remember that due to the wastespark system, that twin tower coil is firing into one freshly charged combustion chamber filled with fresh air + fuel + pressure *and* one highly ionized charge combustion chamber that has just fired and mostly filled with exhaust gases.
So which one do you think coil will want to fire easier?
1) high pressured and filled with air + fuel?
2) less pressurized and filled with highly ionized burning gases?
Yes, #2.
The twin tower coil will tend to fire easier in the wastespark.
This is where the misfire occurs.
Mazda knew about this, but it doesn't really matter cause it doesn't really affect performance.
Slap on a CDI box on the leadings, and this "problem" disappears.

Any comments or corrections welcome...


-Ted

[sv51macross]

RET, thanks a bunch, perfect answer for my question (and alot I didn't know.) Very much appreciated.

[NoDOHC]

Wow ReTED, you should make that a sticky. I thought I was a little sparse in my description, but you have put me completely to shame!

Kudos to you!

The only question that I have is:

I thought that the stock coils fired using wasted spark. This is usually configured so that the current must pass through both plugs. It seems strange that a wasted spark system could possibly fire on only one of the plugs (are you sure about the cause of this symptom)? I am using the stock ignition and I only get misfires if I lean the mix up too much. Otherwise the engine purrs as smooth as glass. I think it might be more of an ECU issue.

[Pete_89T2]

I thought that the stock coils fired using wasted spark. This is usually configured so that the current must pass through both plugs. It seems strange that a wasted spark system could possibly fire on only one of the plugs (are you sure about the cause of this symptom)? I am using the stock ignition and I only get misfires if I lean the mix up too much. Otherwise the engine purrs as smooth as glass. I think it might be more of an ECU issue.

The leadings coils fire wasted spark, the trailings do not. The leading coil is essentially one coil with its output wired in parallel with 2 towers. So for every spark event triggered on the input side of the leading coil, in theory you get a simultaneous spark firing at both leading plugs.

[RETed]

It's just a lot of crap floating in my head...
Trying to type everything coherently is the big challenge. :)

Like Pete_89T2 said, the twin tower leading coil is just like a regular single tower coil except it has two towers...outputs...in parallel.
As long as the coil can discharge it's charge potential, it doesn't matter if it fires out one or both towers.
In an ideal situation, it fires two (identical?) sparks out the twin towers.
In reality, the wastespark system doesn't allow for balance loads across that twin tower leading coil.

One spark is firing in a pressurized, air + fuel combustion chamber - more resistance to firing a spark across the gap.
The other spark is firing in a highly charged, ionized (exhuast) gas, decreasing pressure combustion chamber which is a whole lot easier to fire a spark across the gap due to 1) less pressure and 2) ionized burnt particles of carbon.

It's this imbalance that causes the misfire of the twin tower coil.



-Ted

[NoDOHC]

That explains it, I thought it was a typical wasted spark system. I wonder why Mazda did it that way?

Typical wasted spark has:
http://i594.photobucket.com/albums/tt25/NoDOHC/WastedSpark.png

Obviously, current can.t flow through one plug without flowing through the other one too.

The advantage is that the spark voltage can be twice as high (for bridging the dielectric strength of the gap) with the same 'system' voltage. It also helps to eliminate the shock hazard, as the spark needs a path to ground through the other plug to zap you.

What I understand from 'in parallel', is that Mazda connected the wires like this:
http://i594.photobucket.com/albums/tt25/NoDOHC/ProposedMazda.png

The current could follow either path, but as ReTED said, it will always follow the easy one. I find this very hard to believe, it would never fire the plug in the high pressure chamber.

There is an easy way to test this (I am traveling or I would).

Disconnect both lead plug wires at the coil and have a friend crank the engine.

If the spark jumps between the two posts, it is wasted spark, if it jumps from one of the posts to ground, it is wired in parallel (which makes absolutely no sense why they would do that).

[RETed]

Good point, but wouldn't the system be more reliable with parallel circuits?
If, for some reason, you get a misfire, a series twin tower coil would prevent both sides from firing.
If you get a misfire, in a parallel system, at least one side would end up firing...

Side note...
KDR used to sell this ridiculous system which eliminated the trailing spark plugs and fired everything on just the leadings.
If the ignition system were in series, I'd be scared to run it like this...


-Ted

[NoDOHC]

Correct you are, but there is only one side that matters anyway (the other side is the 'wasted' spark) so a misfire is just that, a misfire.

The air has a very high resistance until the electric field intensity exceeds the breakdown level (based on plug gap, AFR, compression ratio, charging efficiency, boost, etc.) no appreciable current will flow until the air between the electrodes is ionized. Once this happens, the resistance drops to a very low level, effectively shorting the coils out (voltage drops very low) The actual spark is then a current through the now-ionized air gap.

The problem with parallel circuits is that the dielectric strength of the spent chamber (just completing the exhaust stroke) is a lot lower than that of the compressed chamber with air and fuel in it. This means that the spark voltage will build until the field intensity exceeds the breakdown level of the lower dielectric, then that ionized gas in the air gap of the spark plug will short-circuit the coil (spark voltage goes low) while current flows through the air gap. This means that the charged chamber will never get enough voltage to the spark plug to exceed the higher dielectric strength and the charged chamber will never fire.

Wasted spark is very reliable (typically). I don't like it because it requires twice the effort (ignition events) out of each coil (while it only requires half as many coils) the charge time for the coil is half as long as it would otherwise be, this means there is not enough charge time to get a reasonable ignition energy to the air/fuel charge in the chamber at high revs.

[t_g_farrell]

The twin towers are a bean counters solution to reduce the number of coils
required for the ignition. They should have used a single coil for each leading
plug or for each plug for that matter.

Anyway, I run one TFI coil per leading plug and it runs very smooth, no trailing.
Until now I wasn't sure why but RETeds' explanation was an ah-ah moment for
me once I read it. Makes sense to me.

[My5ABaby]

Perhaps I'm confused, but the best system for a rotary is ____?

[Pete_89T2]

The problem with parallel circuits is that the dielectric strength of the spent chamber (just completing the exhaust stroke) is a lot lower than that of the compressed chamber with air and fuel in it.

What data supports this? If this were true, then the odds of the the stock FC ignition system ever firing the leading plug of the *correct* rotor (i.e., the one with the compressed fuel/air charge) would be slim to none.

This means that the spark voltage will build until the field intensity exceeds the breakdown level of the lower dielectric, then that ionized gas in the air gap of the spark plug will short-circuit the coil (spark voltage goes low) while current flows through the air gap. This means that the charged chamber will never get enough voltage to the spark plug to exceed the higher dielectric strength and the charged chamber will never fire.

This explanation is correct, but I suspect you've got the dielectric characteristics reversed - it must be much easier to jump a spark across the gap when the medium is a compressed fuel/air charge than exhaust gases & blowby, assuming the full regime of variables as Mazda designed, otherwise our leading plugs would always misfire.

[RETed]

Circuit is not as simple as we thought...

Did a quick test of resistance on the terminals.
If the parallel circuit was as simple as above, measuring resistance across the towers should be nil - not the case.
In fact, the towers are isolated from the two coil inputs + and - (no surprise there), and there is no resistance between either tower to coil ground - kind of a surprise.
The coil outputs are isolated from every part of the coil body including from each other.

Twin, independent coil circuits wound within the coil body itself?


-Ted

[Pete_89T2]

Circuit is not as simple as we thought...

Did a quick test of resistance on the terminals.
If the parallel circuit was as simple as above, measuring resistance across the towers should be nil - not the case.
In fact, the towers are isolated from the two coil inputs + and - (no surprise there), and there is no resistance between either tower to coil ground - kind of a surprise.
The coil outputs are isolated from every part of the coil body including from each other.

Twin, independent coil circuits wound within the coil body itself?


-Ted

Ok, now it's starting to make sense. By inference of Ted's measurements, the tower outputs can't be wired in parallel. I can't draw & post the circuit, so hopefully this description will make sense to everyone if you reference NoDOHC's diagram and use your imagination.

If you added a 2nd distinct coil winding to the secondary side of the transformer, and connect one leg from each of these secondary windings to a common ground through a diode (i.e., allows DC to flow in only one direction, connect "-" side of the diode to ground, "+" side to coil winding) and the remaining coil leads become your L1 & L2 tower terminals, Ted's measurements make perfect sense. In effect, we have two secondary coils driven by a common primary coil.

[NoDOHC]

You may not be able to see resistance, a lot of coils are capacitively coupled. It is simplest to observe the direction that the spark takes while cranking.


This explanation is correct, but I suspect you've got the dielectric characteristics reversed - it must be much easier to jump a spark across the gap when the medium is a compressed fuel/air charge than exhaust gases & blowby, assuming the full regime of variables as Mazda designed, otherwise our leading plugs would always misfire.


Actually, you can research the dielectric strength characteristics of air easily. Higher pressure increases the density of the air, which increases the dielectric strength of the air (more molecules to force to ionize).

Breakdown Voltage (http://en.wikipedia.org/wiki/Breakdown_voltage)

Anyway, I am back in town today, maybe I will test this myself tonight.