auzivision

How does the EZK retard the ignition?

The ignition system looks pretty much like an electro-mechanical devise. With caps on the cams turning in distributors connecting spark plugs to coils I don’t get how an electronic module inside the cabin gets a chance to intervene.

To think it has the ability to retard specific cylinder varying amounts makes that even more intriguing. The spaces between the distributor contact points and the speed of the rotor appear to be fixed my mechanical means.

So where does the timing variation come from?

Leon Speed

I don't know, but would guess that the pulse to the coil is timed. If you look at the rotor, the contact point about one cm wide. The distributor point is about 0.5 cm. So in total about 2.5-3 cm variance on a circumference of, what, about 40-50 cm. That is appr. 360 degrees/40 = 9 degrees/cm * 3 = 27 degrees variance (ok very rough calculation, but hope to get the point accross )

FredR

Kurt,

The EZ system receives signals from the knock sensors and crank sensor that enable the computer to understand which cylinder is actually knocking and the severity of the knock and it winds back the advance electronically to suppress knock until it is at its full range [I seem to remember it is either 3 degrees total or 9 degrees total in 3 x 3 degree incremental steps- the ST2 enables you to select the step width]. I am not sure whether it applies this correction cylinder specific or to all cylinders at once but either way it should be no problem for the system to do this if it was designed that way as it is a simple logic step.

Remember, the cycling speed of the computer is very fast compared to the engine speed and in any event the engine wil not suffer if knocking occurred for two cycles instead of one [for instance]. The computers decide when the ignition timing signal triggers sending the trigger pulse to the final elements [the signal comes from those two modules sat on the front lip under the hood and they relay the signal to the coils. The coil output then pulses with HT voltage and this fires the plug connected to the cylinder position of the distributor that is in the connected position.

On the ST2 we can tell which cylinder [or cylinders] are knocking and the severity of the knock and when adjusting the ignition map we can retard specific cells by the margin we select until the knock is under control [i.e. onset of knock]- we perceive that point to be the optimal ignition advance for that cell. We cannot select specific advance values for each cylinder however.

Once the EZ detects a knock I am not sure whether it applies a global correction to all cylinders or a correction specific to that cylinder when it fires the ignition trigger signal but the computer knows which cylinder it is firing so it may well work like that. For sure john Speake will know the answer to that one.

The distributor and rotor only facilitate the ignition and where it happens, they have nothing to do with when it happens.

Regards

Fred R

auzivision

Thanks Aryan, I think you are on to something. The contact patch is bigger than the impulse width allowing some variation. Obviously, the control would have to be prior to the coil since everything beyond is fixed mechanically. I’ll have to ponder that some more.

auzivision

I understand the EZK takes a variety of inputs to calculate the retard and has the ability to retard only certain cylinders. In fact, trying to understand how it has the ability to retard only one cylinder and not another is what got me wondering about this subject in the first place.

Not to derail this topic, I’m not concerned with knock sensors, cam timing, hall sensors or even the EZK ignition maps, those seem to be fairly straight forward concepts. My question is once the EZK make up it’s mind to retard a certain cylinder... How does it actually do it?

Cosmo Kramer

It retards all 8 when it senses a knock, by changing the timing pulse to the coils. Newer OBD2 systems with individual coils on each cylinder can retard a cylinder individually, the EZK doesn't have that ability.

69gaugeman

Not that I know anything about this (I don't), but why could they not do each individual cylinder seeing as they know which one is knocking. Otherwise they would just put one knock sensor on and call it done?

auzivision

Maybe the EZF retards all 8 at the same time, but the EZK does have the ability to retard specific cylinders. Hence the hall sender is incorporated to know which one knocked and which one gets spanked with a retard.

The older cars don’t have knock sensors so I presume the timing is set by the EZF map and doesn’t automatically retard when knock is present. I really don’t know much about the earlier systems or how they work.

The sad thing is my dad designed ignition control modules and early electronic spark control modules in the 70’s, but we never got into details how these things worked… just a little bit about what they did, but then again I was only 10 at the time.

I thought I might pick his brain to see what he remembers. He was more of a transistor and IC guy, so he might not know or remember. Anyhow, I thought it might be fun to strike up a conversation on the topic and would love to be able to converse with confidence on the Bosch system.

jcorenman

As Aryan points out, the distributor contacts are wide enough to distribute the spark to the correct cylinder over a wide range of spark timing-- which is the case with all distributors, whether the advance is handled mechanically or electronically.

For the EZK, the amount of ignition advance is determined by RPM, load (from the MAF via the LH), temperature (from the Temp-II sensor), and throttle position (idle and WOT), and then modified by any knock-retard that may be active. (Mechanical distributors use RPM and load-- i.e. engine vacuum-- to determine advance).

The crank position sensor (CPS) and hall sensor tells the EZK the rotational angle of the engine, and when to fire the spark. The CPS indicates the rotational angle and the hall sensor indicates whether the cylinder is on the compression stroke (and should be fired) or on the exhaust stroke (and need not be fired).

The two knock sensors listen for small engine knock sounds, and along with the CPS and hall sensor allow the EZK to determine exactly which cylinder is trying to knock. If a knock is heard then the ignition timing for that cylinder is retarded 3 degrees; if the knocking for that cylinder continues then the timing is retarded further in 3-deg steps, up to a total of 9 degrees. With no knocking, this retard is removed at the rate (IIRC) of 1 deg per 3 seconds. As noted, these numbers can be changed with Sharktuner. Louie suggests smaller steps, e.g. 1 degree retard steps, and removing it quicker-- otherwise a retard that might be picked up at 3800-4000 RPM is still in effect at higher RPM when not needed.

If the hall signal is missing, or there is a knock-sensor fault, then the timing is retarded 6 degrees, and the knock-retard system disabled.

It's actually quite a nice system, even by today's standards.

Tom in Austin

Hi Kurt, I've always wondered about this same question. Assumed the width of the rotor tip allowed the spark to come from the coil at slightly different times, since, as you noticed, everything else is mechanically fixed to the movement of the cam and crank.

What I think is also true is there is no way to set the initial timing on an EZK engine, other than ensuring the cam sprocket is timed correctly to the camshaft, and the hall and crank sensors are functioning correctly.

jcorenman

Kurt, the EZF is a simpler ECU and does not have any knock-retard. It also doesn't have the hall-sensor signal, so it fires the spark for each cylinder on each revolution-- both compression and exhaust stroke. (This is a "wasted spark" method and is quite common. Since sparks are "free" there is no real waste).

The EZF also has access to engine vacuum, I don't know if it uses that instead of a MAF-derived load signal (from the LH) or in addition. In theory the MAF's load signal tells the EZK everything it needs to know about engine loading.

Cosmo Kramer

OK so I think I get it now, the width of the rotor tip gives enough "contact window" to allow for some changes in timing to the individual cylinders via the pulses from the coil. Interesting. I was thinking that this part of the system was fixed and hence a system wide retard in timing or at least one bank at a time.

PorKen

As I understand it, the EZF/EZK is always calculating when to fire the next + 1 spark, so it has plenty of time. The very next spark is already locked in. It sets an internal counter to fire sooner or later, depending on conditions.

Two cam driven distributors = no wasted spark.

Once the engine is warmed up, the EZF uses rpm, it's internal MAP (vacuum) sensor, and air temp as variables. I've found that air temp is a good predictor of the advance limit. (On my EZF chip, I rescaled, and recoded the air temp retard to be more reactive to running temps.)

LH 2.2 calculates and outputs an EZK type MAF x RPM load output (pin 24), but it is not used.

auzivision

I’ve never studied ignition systems nor timing before and only have a few data points that cement my incorrect assumptions.

Once when I was a kid, I remember a neighbor guy ‘timing’ an older Firebird. He had a timing gun and showed me how the spark would flash a light and he could move what I think was a distributor cap which adjusted the timing. Kind of analogues to adjusting a stroke light flashing a fan.

Second, somewhere along the line I’ve learned that current in a coil can not be instantaneously switched off. This is why if you if you turn off a ceiling fan with a light switch the current will ‘jump the gap’ and create a spark. For that matter a large incandescent light bulb will have enough inductive load to throw a spark.

So I incorrectly assumed that when the contacts inside the distributor ran out of room (or disconnected) that’s when the spark event occurred. That’s was obviously wrong now that I really think about it, because we wouldn’t want the spark under that cap.

So that leads to my next question. The coil in essences is really a transformer (a pair of coils magnetically coupled)… not just a simple inductor as I also incorrectly assumed? A low voltage signal is being switched on and off and the transformer (aka coil) is ‘stepping up’ the voltage... and since the current can’t be instantaneously interrupted; it will seek the least path of resistance back to ground. Which in our case happens to be the spark plug gap that is conveniently connected to the desired spark plug at the time.

Anyhow, I think I may have answered my very question. So the big wires to the coil provides the high voltage positive to the plugs via the distributor and the ground via the strap. The little wires are low level signals generated by the EZK be pulsed on and off in time sequence with the engine.

So in slow motion is this the proper sequence of events?

1) cap/rotor line up to create contact to spark plug
2) low level signal to coil is turned on
3) low level signal to coil is turned off
4) current spike jumps the gap at the plug
5) cap/rotor disconnect move to next cylinder

Please fill free to correct or ‘fill in the gaps’… so to speak.

jcorenman

Actually, pre-85 928's set timing the same way... And you can check timing on any 928 with a timing light, you just can't change it (without a Sharktuner) on the later models.

Correct, an inductor is basically an electrical "flywheel": current wants to keep flowing, and changing the current requires voltage to speed it up, and generates voltage as it slows down.

If you close a switch and apply voltage to an inductor, then current will start flowing and build up to some level which depends on the resistance of the circuit. If you then open that switch, current will continue flowing and will create a very high voltage until something starts conducting that current-- such as an arc across the switch, or an breakdown of insulation (or your fingers ).

Traditional ignition systems used an 8-lobed cam in the distributor (for an 8-cylinder engine), which closed and opened a set of contact points (breaker points) eight times per distributor revolution-- i.e. once per cylinder. As the breaker points closed, current built up in the coil, and when the points opened, the current stopped-- generating a high voltage.

The ignition coil is both an inductor and a transformer. What happens at the primary (the side that the points control) is reflected at the secondary (and vice versa), times a few thousand. So when the points open, the secondary voltage gets very large and at 5-10kV (depending on pressure and gap) the park plug starts arcing, and continues to arc until the current runs down.

In the meantime, the rotor is attached to the same shaft and has rotated to the appropriate contact, for whichever spark pug is to fire. There is another gap there, but small compared to the spark plug (and no combustion pressure). But the rotor has nothing to do with generating the spark, it only distributes it to the correct cylinder.

The 86+ 928's are the same, except the breaker points are replaced by electronic switches, and the rotors live on the ends of the cams instead of in a separate distributor.

Correct. The coil is both an inductor and a transformer; the inductance is what generates a high voltage when the contact is opened, and the transformer multiplies that change of voltage to be high enough to fire the plug.

Plus one more piece: The current (and voltage) through the coil is too high for the EZK to switch directly, so there is an amplifier between the EZK and the coil-- located at the front of the engine compartment on the driver's side. This amplifier takes the low-level signal from the EZK and uses that to close a electronic switch from the coil to ground, in the same way that a traditional breaker point connects one side of the coil to ground to charge it. Then, when the EZK wants to create the spark, it drops the low-level signal which opens the switch and dumps the coil into the spark plug.

Yep, that's it basically. Except of course the rotation of the rotor is continuous, and the spark is fired as the rotor is moving. And there is no literal contact between the rotor and the wire to the plug, just a very small gap. (There is literal contact at the center contact).