Let's talk about spark duaration
 

This is not about dwell time, but spark duration. I was going to post this in the MS section, but this might be a topic that benefits anyone that can adjust this parameter.

I started thinking about this today while thinking about the aftermarket ignition on my 69 charger. the BIG deal about these aftermarket ignitions (besides being electronic) is that they supply a hotter spark, and a longer spark duration.

A turbo application seems like a situation that might benefit from longer spark duration. Longer duration should promote better fuel burning efficiency.

1. Does this minimize the peak spark intensity because the coil has to spread the energy over the entire duration?

2. Is the coil capable of supplying peak spark energy over an extended spark duration?

3. Is the current .7ms spark duration optimal? If the answer to question 1. is yes, then maybe a shorter hotter duration would be more beneficial.

Let's see if this makes sense. It takes 2 crankshaft rotations for 1 piston to run through a complete cycle. At 6000rpm I see 1 cycle takes 20ms. The time between the compression stroke and the exhaust stroke would be 5ms or 4.3ms at 7k rpm. If we are running 36 degrees of advance at 6k rpm and wanted the spark to burn for 20 degrees past TDC, then we would need a spark duration of 1.55ms and ~ 1.32ms at 7k rpm.

BUT

While in boost: Maybe we are only running 15 degrees of advance so we need a spark duration of .8ms to ensure a burn past 20 degrees.

However, there is nothing really limiting you to 20 degrees past TDC.

As long as the coils are capable. You could tune your spark duration based upon the engine rpm and advance that you want to optimize.

1.5ms would also be appropriate for 4000rpm and 18 degrees advance burning to 20 degrees past TDC. (No point to this last one, just picked some numbers to run)

It divides by two.

+1

shit, your right, I'll fix my numbers after dinner. Keep moving forward with the concept. Ok, fixed, fortunately I the rest of the numbers are ok because it still winds up being 10ms per 360 degrees.

No.

That's not how it works. With inductive discharge the voltage at the plugs will rise to whatever is required to initiate the spark, as long as the coil is capable of reaching that voltage. Mixtures outside 12:1 and 15:1 require more voltage. Greater MAP and compression ratio require more voltage. Poor atomization of fuel, requires more voltage.

And then the voltage drops to whatever is required to maintain the arc. During this time, the current drops linearly as the current in the coil depletes. The larger the secondary inductance of the coils, the longer this takes. The larger the initial current at the beginning of the arc, the longer it takes. The initial current is a function of the primary current at the end of the dwell period, and the turns ratio of the coil (it's like a transformer).

Do we have any idea on the time frame that this takes? Does the MS spark duration setting really do anything? It sounds like once the coil starts discharging, it goes until discharged and nothing is really going to stop it.

there may be something to this. purely anecdotal, but a few years ago I had a burned up exhaust valve. I knew it, and had a new head to go on with stainless valves and such, but it was off getting resurfaced when my emissions were due. there is a loophole here in Maryland that if you spend $xx on emissions related parts you get a two year pass. well those goofy capacity discharge nology hotwires were two bucks over the dollar amount to get the pass so I ordered them, installed them so the referee could verify the new parts and then retested. needless to say I still didn't pass, but it was damn close to my surprise. the most surprising thing was how much they cleaned up every measurement. so I can't speak to power as I never did a before/after dyno run with them, but if it burns more efficiently with them on there is probably a small power advantage as well. I still run them to this day and have never had any spark related miss fires. even when the wideband reads pig rich at 14lbs of boost it still lights off no problem. however if there was a way to make a spark that intense for a longer duration I'd be all for it.

djp, the discharge ends when the energy's all gone, or if the next dwell cycle begins.

The time is on the order of a one to a few ms.

http://www.picoauto.com/waveforms/im..._secondary.png

I don't know what the MS duration setting is for.

Maximum Spark Duration (ms) (max_spk_dur) is the amount of time MS2-Extra, MegaSquirt-II (or MicroSquirt) tries to wait before starting another charging cycle. MS-II tries to fit the dwell time + max. spark duration into the time between sparks. If it can fit them, then the full dwell time is used. If the time is too short (because the rpm is high) to fit both the dwell and the max.spark duration, then both are shortened proportionally.
If the max spark duration is too long, however, then the actual dwell value gets shortened at a lower rpm, limiting the spark energy at higher rpms. Typical values are between 1.5 and 2.0 milliseconds.

There is MSextra documentation for running COPs now, and they use a spark duration of .5ms. It looks liek shorter is better. The last thing you want is the dwell time being shortened at high rpm.

yeah the way I understood it was the time between dwelling.

Putting the duration to 1ms seemed to help in some areas, but at higher rpm it didn't feel right. I lowered it to .5ms per the msextra documentation and it feels better all over teh rpm range. Better than .7ms, I can't tell.

Brain, would you try .5ms and see if you notice a difference. It seems like the shorter the better based on the msextra definition.

Adjusting the spark duration value in software does not directly affect the duration of the spark event.

The only way to 100% absolutely set your dwell correctly is to use an oscilloscope. One of things you can look for on the scope is the duration of the spark event. You take the measurement with the scope and input the data into the max spark duration box. This value adds a "dead time" between dwell events to compensate for the duration of the ark.

As there is only so much time between spark events (dictated by engine RPM), arbitrarily increasing the duration will decrease your available possible dwell time. Decreasing the duration will increase available possible dwell time. But changing this value in software does not have any direct relationship with actually increasing or decreasing the spark burn duration. The spark duration is not a direct function by the engine management. The MegaSquirt is telling your coil when to charge and when to discharge (fire), but does not directly control the duration of the ark--that's up to the ignition system itself and the other factors Jason detailed.