Dwell time increasing top end pull?

[Supercharged111]

I've done some searching, and most people who've increased dwell haven't noticed any noteworthy gains. However, since my engine (blackbox L31) is apparently rev limited by the dwell values, it stands to reason that I'm losing a bit of top end pull to crappy dwell values. I've seen claims of people who installed the '02 Express van PCM claiming more top end pull. I looked at the dwell values of the 2 and the blackbox actually has a longer dwell time than the Express van's stock tune. Has anyone ever set the dwell on a single coil V8 before? Seeing how it cuts the engine off around 5600, I figure it's gotta be hampering performance as I near my 5000RPM shift point. The first one is my blackbox table and the 2nd one is from the Express van tune.

[swingtan]

Here's some maths for you....

At 5600 RPM, you have...

• 93 Revs / Second
• 46.33 "combustion cycles" / Sec
• 373.73 "Sparks" / Sec ( on a single coil V8 )
• One spark every 0.00268 seconds ( or every 2.68 mS )

So making some assumptions on how the ECM controls the coil charging...

If your dwell time is set to be longer than 2.7mS ( in a single coil V8 ), then at 5600 RPM, the ECM would effectivly kill the spark by re applying battery power to the coil. Even just below that, the ECM would be cutting the spark shortly after it fires which could prevent a good ignition of the intake charge. The "time between sparks" for the top table would look something like this....

RPM	Time

3700	4.05
3800	3.95
3900	3.85
4000	3.75
4100	3.66
4200	3.57
4300	3.49
4400	3.41
4500	3.33
4600	3.26
4700	3.19
4800	3.13
4900	3.06
5000	3.00
5100	2.94
5200	2.88
5300	2.83
5400	2.78
5500	2.73
5600	2.68
5700	2.63
5800	2.59
5900	2.54
6000	2.50

Note that the theoretical time between sparks is less than the dwell time settings from 3700 RPM. So from 3700 RPM the ECM is extinguishing the spark by applying voltage to the coil rather than the coil charge disipating.

There are a couple of thoughts here though...

1. Applying voltage to the coil before it has fully discharged, allows the recovery of some of the charge for the next spark event. This would help maintain maximum coil charge in the higher RPM areas, giving more spark voltage to initiate the spark.
2. Applying voltage to the coil will stop the rapid collapse of the magnetic field that is generating the spark voltage, so it will extinguish the spark. If this is done too early, the intake mixture may not ignite correctly, or at all.

So you need to find a happy medium. I'd start with the theoretical values and see how that goes. Then maybe move up and down by 0.5mS to see if it gets better or worse.

Just remember that too much dwell time will burn out or over heat a coil......

Simon.

[Supercharged111]

Geez, I thought the theoretical was an available 17ms at 6000RPM, I should search that reference out again. What I'm seeing is that my stock black box main dwell values are actually higher than your theoretical values. EFILive suggests reducing the minimum dwell settings by 1/3. It is in fact lighting a new spark before the old one is done depleting the coil, isn't it?

[Supercharged111]

OK I found it. I also found more along the way, I guess I'm not the first to notice this.

http://forum.efilive.com/showthread....ighlight=dwell

Here it was you who said 17ms.

FWIW, at 7000RPM, the time between spark events for a single coil is about 17mS ( 7000 RPM in a 4 stroke motor is 3500 sparks per min. for any given cylinder. This is 58.3 sparks per second or one spark every 17mS ).

http://forum.efilive.com/showthread....t=dwell&page=3

[joecar]

17ms @7000 rpm: that's for 8 separate coils (each coil has two revolutions to complete one charge/discharge cycle).

If you have a single coil/distributor, then it is shared among 8 cylinders, so you divide the two-revolution-cycle time by 8... and that's only the upper limit, you have to subtract a few milliseconds for spark duration (1.5-2.5 ms).

[Supercharged111]

He said single coil, so I interpreted that as my engine with a single coil. Regardless now it looks like the Express van has shorter, more ideal dwell values, unless I'm missing something? It seems counter-intuitive that the factory would give values that are too long in duration as that would shorten the life of the coil. I must be missing something.

[joecar]

Like Simon said, at 5600 rpm, two-rev-cycle time is 2*60*1000/5600 = 21.4 ms.

With single coil and 8 cylinders, 21.4/8 = 2.68 ms.

Spark duration guestimate = 1.5 ms...

This leaves 2.68 - 1.5 = 1.18 ms for coil dwell.

[joecar]

Do you have a distributor...? Or do you have 8 coil packs...?

[swingtan]

OK I found it. I also found more along the way, I guess I'm not the first to notice this.

http://forum.efilive.com/showthread....ighlight=dwell

Here it was you who said 17ms.



http://forum.efilive.com/showthread....t=dwell&page=3

You need to read the post in the context of the entire thread. We were talking about "single coil per cylinder" setups in that one, not "single coil per engine" as we are here. For comparison sake, you could divide the 17mS by 8 to get the spark repetition time @ 7000 RPM, for the single coil per engine setup.... 2.125 mS.

Simon

[Supercharged111]

I've got a distributor. The thing that has me perplexed the most is why the black box Vortec PCM's dwell values seem to be not only more than a theoretically ideal value, they're more than the amount of time between firing each cylinder. It has me wondering if there's not something I'm missing, like maybe there's another table that's not unlocked. I'd like to put the Express values into the black box, but I'm kind of scared to since they're so radically different. Both vehicles have the same L31 engine. Since the Express van has 5200RPM WOT shifts in performance mode, I figure there's got to be something about that stock tune I can steal for my truck because it seems like around 4600-4800 it really falls on its face and knowing that the stock dwell values are what actually causes the rev limit in the black box I really think it might be part of that. If I just steal the values from the 14V column in the Express tune and plug them into my black box is it going to work? The fact that the values are so far off from each other has me really second guessing that. Hell, I'd like to plug in the values that swingtan posted, does it strike anyone else as really odd that the stock values are too high?