So I've been trying to diagnose a rough idle, and as I was pondering the distributor and ICM I realized I don't really understand the actual mechanism of action the computer uses to control the spark timing.

There's a mechanical linkage between the distributor and cam via the meshed gears, and as far as I can tell the rotor has no way of being adjusted back and forth, so what's going on that allows the computer to advance timing so much?

The only thing I could think of was that because the base of the distributor is a helical gear and because the distributor shaft has some up and down play, that the computer was somehow physically moving the distributor shaft up and down to adjust timing. Is that what's happening?

I could see the computer maybe sending the spark either the very instant the rotor gets close to the cap terminals vs sending the spark at the very last moment as the rotor moves away from the terminal, but is that enough to give 20+ degrees of timing advance?

 

The signal to create a spark is generated (by the pickup coil on carbed or TBI, by a crank sensor on certain other vehicles) WAY in advance of the actual timing desired. The computer receives that signal, delays awhile (like holding that signal in computer memory) before sending that signal on to the ignition module which grounds and un-grounds the ignition coil to create sparks. The amount of delay is variable based on the computer's spark-advance tables. At this RPM, under this much load, in this gear, the delay is this much. At a different RPM, or a different load, the delay is shorter...or longer.

The shorter the delay, the more-advanced the ignition timing is.

The spark tables programmed into the computer set the expected timing advance, but signal from the knock sensor can further delay the spark, leading to additional retard to cure "pinging".

It's all computer/electronic magic.

Yes, the spark jumps the rotor-to-cap gap with the rotor ahead of, aligned-with, or behind the cap terminal depending on how much the spark is advanced. The geometry is set-up to provide a reasonably-centered alignment most of the time. On the Vortec trucks, this is adjusted by means of the "distributor offset" function on a scan tool. Otherwise it's just "baked-into" the way the parts are made.

 

That's such an interesting integration of late stage mechanical tech and early stage digital tech.


So from the time that the rotor tip is close enough to the terminal to jump the gap, to the time that the rotor is just far enough away to not jump the gap, that 1/4" difference is enough to get 20-30* of advance?

The rotor must spin faster than the engine, right? So one revolution of the motor is equal to say four revolutions of the rotor, so that even a quarter inch rotation would be a pretty small degree of overall engine rotation? Does the rotor make multiple passes by a given cap terminal before the computer tells it to send a spark?

 

Schurkey summed it up well.

As far as the rotor to engine speed, it is the rotor is slower then the engine. For every two engine revolutions, the rotor makes one revolution. Recall that the cam spins at half the engine speed. Takes two full engine revolutions to fire all eight cylinders (4-stroke and all).

As for rotor tip to cap tip alignment. The ECM is only allowed to program so much retard and so much advance into the distributor. This is to prevent cross-firing.

HTH's,

RBob.

 

VERY interesting!

 

I cut a hole in a cap some years ago to watch just what you're looking into now. There are no moving advance parts, just electronic timing and at low rpm or starting rotor is right on the distributor cap terminal. At full advance it's about 1/2 inch away and the spark really lights up in there. That 1/2" in the small cap with the half speed of the dist. does give a lot of advance. Amazes me that the AM radio works without a bunch of static too.

 

Wow! I wouldn't have guessed the spark could reliably jump a gap that large, but I guess the voltage is pretty damn high.
Fascinating.

Also sidenote, I fixed my idle. Turned out the distributor cap was cracked and the timing needed to get redone. I also got my EBL flash II going this weekend and wow it's never run better.

 

The rotor makes a 360-degree path to fire all eight cylinders. (As said, the crank turns 720 degrees in this time)

360 / 8 = 45. There's 45 degrees of physical separation between the spark plug terminals on the cap. But that represents 90 crankshaft degrees, and crank degrees are what we're talking about when dealing with ignition advance. Again, 2:1 ratio of crank to distributor degrees.

So 30 degrees of ignition advance is 15 degrees of rotor-tip-to-distributor cap terminal, and hopefully set up so that at typical "running" advance, the spark path is minimized. Given the width of the rotor conductor and the distributor cap conductor, there isn't much misalignment at all. I have seen some distributor rotors that have an extended-width conductor to assure minimum firing voltage between rotor tip and distributor cap conductor.

Note the width of the rotor tip on this Ford distributor rotor: