[Joe Perez]

In preparation for removing the engine from my wrecked '04 FXT to transplant it into a new vehicle, I've been scoping out all of the critical timings and voltages on the motor as I'm still a bit on the fence as to whether I'll be retaining the stock ECU or going with a Megasquirt. That thread is here.

So a couple of days ago, I went in to establish the timing relationship between the crankwheel and the #1 ignition coil. Here's the scope trace I took of that measurement, at idle, with CKP in trace #1 and IGN 1 trigger in trace #2:

Note the timescale: 10ms/div. Which means that the dwell was approximately 8ms.

?!?

Needless to say, this struck me as quite odd. I'm much more accustomed to seeing dwell measurements in the 2.5-3ms range on modern COP ignition systems.

Last night, I measured it again, this time with a DC current probe as well, clamped to the +12 supply to the coil. Trace #2 is the current probe, which is scaled to 10mV=1A, and the scope vertical is 20mV/div. Timescale here is 1ms/div. Trace #1 is the trigger to the coil:

Ooooooo-kay. Now, at idle, I'm seeing about 4.8ms, which is more reasonable, but still a lot longer than I'd have expected, particularly given that the coil is clearly self-limiting at 9A after just under 3ms.


Searching around here hasn't turned up much on the subject of dwell, though a bit of googling produced links to a few other sites where times in the 2-3ms range were being tossed about. I've been poring through RomRaider, and I can't find any obvious table which specifies dwell settings.


I will add as an aside, that I've been leaving the battery on this vehicle disconnected except when I'm actively working on it, and that these measurements were taken after periods of running ranging from seconds to tens of minutes. I can't imagine that the ECU has any sort of auto-learning feature for coil dwell- can it? There's no feedback mechanism that I can see.


Does anybody have any thoughts on the matter which they'd care to share? This isn't exactly a show-stopper, but it's damned curious and I'd really like to understand whatever phenomenon I'm seeing here.

 

[thefoos]

You clearly understand some areas of engine dynamics. You have to remember here that you are dealing with a completely different animal when talking to the EJ25. Its not your standard square motor. Big bore, short stroke relatively speaking.

If you go MS, you can certainly just play around with the dwell. You aren't going to break anything by cutting the dwell in half.

Not knowing what proper dwell is supposed to be on a time base, I'm going to go back to simple statics and dynamic calculations, and use some standard assumptions. With a stroke of 79mm, and a rod length of 131.5mm (I think thats what it is, going off memory), knowing you want to start the spark at 15*BTDC, and I'm guessing continue the spark for somewhere between 10-30*, what sort of dwell does that give you at 700 rpm? I'm too lazy at the moment to do the calculation. Perhaps I'll tackle it tomorrow if you haven't...

But I would advise you to stick with the stock ECU. I'm not sure why you would want to bother with wiring in a standalone when the stock ECU is fully programmable, which you know, since you looked into Romraider.

 

[Joe Perez]

With a stroke of 79mm, and a rod length of 131.5mm (I think thats what it is, going off memory), knowing you want to start the spark at 15*BTDC, and I'm guessing continue the spark for somewhere between 10-30*, what sort of dwell does that give you at 700 rpm?

Dwell has no effect on the duration of the spark event. Well, almost no effect.

Dwell, as used generally in an automotive context, describes the period of time during which current conducts through the primary winding of the ignition coil. During this time, the coil can be considered to be "charging up" so to speak, as the inductance of the primary winding is overcome and a strong EM field is generated. On the scope traces above, the dwell period is that during which the ignition trigger signal is high.

At the end of the dwell period the trigger shuts off, opening the circuit on the primary winding. It is at this instant that the spark event occurs, as the field collapses into the secondary, inducing a voltage into it until the resistance of the air gap across the plug is overcome.

Now, we can control how much energy we put into the coil by adjusting the dwell time, at least, up to the point where the coil self-limits. On older engines with distributors and breaker points, this is precisely what you were doing when you set the point gap. A smaller gap means a longer dwell time. (Though of course, in those engines, the dwell time, as expressed in milliseconds, was not a constant, but varied inversely with RPM. On a computer-controlled engine, dwell time generally is a constant, with only minor variations to compensate for system voltage and the like.)

The more energy we allow the coil to absorb during the dwell period, the more energy it is able to deliver in the form of a spark when the discharge occurs. More energy does not necessarily mean more spark duration, or even a "hotter" spark per se. It's more a matter of there being a threshold point below which you are not going to get a reliable spark across a plug gap of a given size and with a given mixture and pressure in the chamber, and above which you are.

The key point here is that we can control when the spark happens, but we cannot control how long the spark lasts. At least, not on most OEM ignition systems. There are some aftermarket units which provide a multistrike ignition system, where the coil is fired several times for each combustion event, generally using a capacitive boost circuit on the primary (the company MSD takes its name from this concept: Multiple Spark Discharge) however these system are largely extinct in the modern world.


Forgive me if this seems elementary, I just got the feeling that we weren't quite on the same page here. I understand how ignition systems work, I was just really curious as to what software setting (if any) influences the dwell computation on the stock ECU. I can't recall ever seeing one emit such wildly different signals.

I'm not sure why you would want to bother with wiring in a standalone when the stock ECU is fully programmable, which you know, since you looked into Romraider.

Well, given the application (engine swap into a custom built vehicle), it would actually be a lot simpler from a wiring perspective for me to go with a standalone, especially something like a 'Squirt which I am intimately familiar with. I'd be able to eliminate a lot of things (FPC, DBW, VSS, MAF, etc) which aren't strictly necessary for proper engine operation, but which are none the less required to keep the stock ECU happy.

On the down side, I would lose AVCS control, as the MS3 does not yet have support for dual-variable cam drive while retaining closed-loop positioning control.