High revs, dizzy and 'spark duration'

[jsmcortina]

Just thought I'd post this for info.
I was on the dyno helping tune a Westfield with a 2.0 litre Vauxhall "red top" this morning and we encountered a high rpm miss.
(From the factory, the engine has a single coil, locked distributor and a 60-2 on the crank. The distributor also provides cam sync.)

Changing fuel had no effect.
Increasing the dwell (from 3.0 to 3.5ms) made it worse.

I spotted that the "spark duration" was set to 0.7ms. Bumping this to 1.1ms cured the miss.

Doing the maths afterwards, at 8500rpm there is only 3.5ms of time available, so the dwell is already being curtailed. Increasing the dwell was eating into the required spark discharge time.

After this change all ran well and I see no need to change it. If the customer ever has need to rev higher, going to a wasted spark coilpack will be a step to resolve this dwell issue.

James

[elaw]

Interesting... I thought spark duration had priority over dwell?

I would have thought that at 3.5 ms/trigger, it would have subtracted the 0.7 ms spark duration and made the dwell 2.8 ms, regardless of whether you set it higher.

[jsmcortina]

Interesting... I thought spark duration had priority over dwell?

I would have thought that at 3.5 ms/trigger, it would have subtracted the 0.7 ms spark duration and made the dwell 2.8 ms, regardless of whether you set it higher.

The code balances between the dwell and discharge time. (It used not to, but a user suggested that it should be changed.)

James

[slow_hemi6]

+1 for discharge time being what has been entered and not being varied at all. Only vary dwell.

[billr]

Are we voting?

+2 on keeping spark time constant, only reducing dwell if needed. It doesn't really matter to me, using wasted-spark on an engine that will only turn 6500, but varying only one (the spark time) seems more intuitive.

Edit: Does everyone understand that when James bumped spark duration up from .7 msec to 1.1, he isn't getting 1.1 msec at high rpm? It probably is reduced back to about the .7 actually needed as both dwell and spark are "pro-rated" to fit into the available time.

[jsmcortina]

This code function has been in place since May 2012.
There is some logic though - if the dwell time got curtailed, then the time required to discharge the coil will likely be less too.

James

[slow_hemi6]

The coil I was using took about 10 times longer to charge than to discharge. Discharge was about 0.35mS with an optimal dwell around 3mS. To me it is apparent any reduction in discharge will be fractional of the change in dwell.

[billr]

"There is some logic though - if the dwell time got curtailed, then the time required to discharge the coil will likely be less too."

Yes, that is true. I simply hadn't thought about that As said, it is a moot point for me, so I'll retract my "vote".

Is this discussing only pertaining to "standard dwell", does "fixed dwell" result in only the spark duration being trimmed, if necessary?

[jsmcortina]

I don't understand your question! Only "standard dwell" is actually dwell.

James

[billr]

Alas, my Windows skills are so primitive I don't even know how to do a screen capture! When I follow the path below in TS V2.6.0), I see four options for "Dwell Type": Standard Dwell, Fixed Duty, Time after Spark, and Charge at Trigger.

path = main menu=>ignition settings=>ignition options/wheel decoder=>dwell type

I have always used "Standard Dwell", but the others are not grayed-out, I think they are available picks.

[slow_hemi6]

The "others" rely on an external "smart" device/module to perform the dwell.

[PhoB]

I tested this some years ago, it cleaned the missfires on dyno, but blown headgasket on street another day. Maybe detonation, maybe wrong setup of this non well described settings.

So i would like to ask, is there any way, HOW TO SET/MEASURE CORRECT SPARK DURATION SETTINGS ? Never find any kind of info, from my several year megasquirt experience.

Now im in progress with 6cyl dizzy engine with turbo conversion. To max out this oldschool ignition, must be set correct, dwell can be measured, but spark duration...so let talk about it, please.

[jsmcortina]

Spark duration depends on the coil, so it is outside of Megasquirt control. I know that the AEM logic coils (same as DIYautotune coils) are stated as having a long discharge over 1ms.

James

[Matt Cramer]

The easiest way to measure spark duration, if you want an actual measurement, is to use an oscilloscope with a clamp on spark plug wire voltage probe. Duration is from when the voltage spike starts to when it drops back to zero.

If your coil has a maximum duty cycle, set the discharge time so that your dwell divided by the total dwell plus discharge time equals the maximum duty cycle.

Other than that, your best bet is to set it empirically. An incorrect value can cause spark blowout or a coil running hot, but not other problems.

[PhoB]

So, if im understanding right. Spark duration value, is forcing time reserved for for spark = coil discharge. On wastespark at reasonable RPM and COP at very high rpm, it can be set to overkill 2ms for example, to be sure, that spark will be made it this time window and we are just wasting a some time before coil will start charge - but as i notieced, there is "plenty" of time in these systems, and will be wasted anyway.

At dizzy we must be careful with this settings, not to waste=steal time for coil charge. But still is more important to make full spark instead full charge = according first post in this thread.

Now, im enclosing screens with test measuring of "spark duration" on friends opel C20XE, running stock Motronic M2.8 ecu(dual coil + bosch 200 module). You can clearly see, on idle is spark duration just around 0.04ms. Wrong style of measuring? Or it is due measuring on idle(and motronic) - but theoreticaly if dwell is constant = sparkduration/coildischarge is still the same no mater of rpm/load ?