Sequential spark LS COP ignition system.

How is the ideal spark duration calculated? This indicates you have 2.5ms for a V8 at 6k. I'd like to the ideal ms range for a V8 at 7000? It looks like the default tune has the MAXIMUM SPARK DURATION set at 1ms. Which seems wrong. Trying to milk all I can out of the smaller square truck coils. (D581 I believe). I have my dwell set at 4ms in the upper RPM. I'm sure I can bump this too as I don't believe the square coils auto fire above "X" dwell setting like the heat sink D-585 coils do.

 

This would be interesting info to know, but with the caveat that we know increasing the dwell too much will burn out coils. Perhaps "optimum" would be a better goal than "ideal".

 

The optimal dwell is recorded on several you tube videos this is the actual spark duration time that you add to dwell. Which I believe I can raise to 2 at least for a little stronger spark.

 

Hit up Pantera EFI (Lance Nist).

 

Those values control coil charge time, not spark duration. As gametech said, too much dwell will just compromise the coils from overheating. What do you have that suggests a weak spark is an issue for you?

 

Running a twin 72mm 5.3 LS. Non-intercooled on E85 with a healthy amount of water/meth. variable from about 10gph to 30gph with PWM pump control.

Currently around 15gph the car drops MPH at the track. (washer fluid) I'd like to drop the water/meth volume until I see minimal power drop off at the track. Then adjust the coil settings to "optimum" and see if I can pick that lost power back up. Ideally I'd like to drop the ethanol content down to 30-50% and bump the water/meth up to enough to compensate. But I seem to be having spark issues at higher volumes. Plan to switch from washer fluid to 50/50 and see what that nets me.

I think the MS software it uses spark duration and dwell. With seq. ignition sparks only have to occur every two revolutions per coil. So there's something nutty like 20ms between individual coil firings. The square coils don't put out near as much as the D585's but they can operate at higher dwells. I had read that there is not much gain over 6MS but the coil actually peaks around 8. The coils ignitor limits dwell to 8ms total anyway.

Targeting 5.6 milliseconds of running dwell, MS says the nominal dwell parameters should be set to:

Maximum Dwell 6.1 milliseconds
Maximum Spark Duration 2.0 milliseconds
Acceleration Compensation 0.6 milliseconds

Which doesn't make sense to me.

Currently I am only running 4ms max dwell, and 1ms on spark duration. (I don't even see the acceleration compensation field) I'm wondering if the default settings aren't allowing my maximum spark potential with these coils. Weather that’s enough to make a noticeable difference is what I'd like to find out. I also use a table with RPM VS MAP. So I wouldn't be running the higher dwell settings at cruise/idle.

 

To be totally honest, maximum dwell settings are a 'set it and forget it' type of calibration like many other sensors and actuators on the engine. They are matched to the coils specifically for their operating environment. You can run less than the maximum dwell to allow a lower average operating temperature and to prevent the unnecessary power draw, but not more. Any more than the maximum recommended by the manufacturer while operated at steady state will overheat the coils. You can technically get away with running above max settings for short periods because temperature doesn't increase instantly, but at what duty cycle that is permissible or even possible then comes into question. What are the recommended settings for your coils from the manufacturer?

 

After a little more research the spark duration field is mainly used in single coil setup where there is much less time between coil firings. There is so much time between coil firings with individual coils that it’s not really relevant.

I read the square coils max out completely at 8ms but see very little improvement over 6ms. Also read these don’t “auto fire” at higher dwells. As mentioned above I can setup the dwell on an RPM VS MAP table So I don’t need to run peak dwell all the time and risk heating up the coils. I'm at 4ms now I'll bump it to 5.5 in the upper RPM boosted cells and see if I notice any improvement.

Appreciate the help!

 

Hi FF86, YOU are correct, the "terms" are incorrect.

I work with DIY and asked the same question. ( their coil supplier)

First question (spark duration) = The correct term IS Arc Duration in micro seconds.
To calculate for 6000 RPM = 10 ms for a crankshaft revolution.
Thus we know that 36 degrees of crankshaft rotation = 1 ms. (correct Arc Duration by Ricardo)
The Arc Duration IS dependent on Coil Energy, Spark Plug Gap, and Cylinder Pressure.
Second question Coil Dwell = Test of coil for Dwell Truncation.

Most LS coils (10 amp) will fill with electrons in 5-6 Ms at 14 volts.
WHEN this happens, the coil will fry as the copper turns (primary) winding BECOMES A DEAD SHORT AND/OR the IGBT will SHORT.

The best way is to "scope" the coil primary, you will see when the coil becomes FULL of electrons, the current graph ling will go vertical.

The term Maximum Spark Duration (my guess) is Anti-Dwell OR the required time the coil cannot by charged do to the coil arc duration, the time the spark occurs.

I could add more information ?

Lance

 

Do you have any information specifically regarding the 12558693 D581 (square truck coil)?

According to some old Mega Squirt research the LS1 coils were tested on a scope at 14v. At 5ms dwell they put out around 40 milliamps of power and the indicate these continued to make additional power up to 8ms, but the gains were negligible over 6ms. They don’t state the power made over 5ms. Or if it’s worthwhile to run the dwell over 5ms. They also mention the round heat sink D585 makes 120 milliamps at 4ms and 14v!

There is very little info out there on the D581 square truck coils. There also seems to be a huge difference in spark power on these aftermarket ECU’s using the GM coils. People are having all kinds of spark issues with the Holley ECUs even with the huge D585 coils and mild boost.

Yet you have guys like HellbentS10 running 10:1 LS 430 cube engines at 26psi on methanol with .024 gaps with the square truck coils on the BS3 with no issues. AFR’s in the 3.0 range making 2000+ hp.

Which makes me thing the settings for the coils may not be ideal.

 

5ms max.

Enjoy!

 

And, there are multipliers. Make sure you review those and do the math, so you don't go over what you want.

 

According to whom? Was that tested on a scope? At what voltage? Have you tested higher than 5ms?

I’m looking for actual data on the subject with these specific coils. Not monkey-see monkey-do internet BS people just repeat without having any sort of real information. (not saying that’s the case with your reply)

I’m using an aftermarket ECU. There are no modifiers I am aware of. Below is the LS1 dwell table pulled off a stock ECU. Though this doesn’t tell us much if there are 10 other modifier tables that act on this table. Which is why I wanted actual scope info on a coil tested at X voltage and Y dwell.

 

I'd be curious to see what primary coil input voltage on your application looks like on a scope. If the voltage is lower than it should be then the coil won't have enough current to build sufficient energy for the spark.

 

Hi FF86, I believe your MS report IS correct with the ANSWERS PROVIDED.

The Report:
5ms max output = OK + 40 mj ? seems low as our GEN-IV coil has 65mj with only a 19% increase in energy. (my report=58mj)
6ms max output is the same as 5ms because the coil flux is saturated = True
8ms test = yes as possible because of the FREQUENCY of their test is NOT stated. (it does take time to fry a coil, a low amount of heat could be absorbed before damage)
I run the GM OEM coil at 2.4ms at 100KPA, 2.1ms below (ramped), then a MAX of 4.88 ms dwell at WOT for a boost/NOS application

When you LOWER coil dwell at low loads, the coil winding/IGBT will be cooler allowing better energy output at HIGH LOADS for short periods.

Our "581" RFQ to Delphi was done in 2003, we lost by under $1.00 (6.56-5.86) the cost of a Rare Earth Magnet which is NOT fitted into the "M" coil.
At that time they provided the specifications, I could look ?

Holley = YES with my "tech" provided to Doug Flynn.
The short term answer was the fit the IGN-1A coil with success.
The current firmware is 4.? with many of the problems fixed AND a "BOOST" (MAP)/Coil Dwell object added.

The "585" = NO (120mj) my observation is closer to 85mj.
This is ONLY a 10 AMP coil, getting that energy/time doesn't make the math correct.

Lance

 

Thanks!

So 5ms in the big boost range is high as I’d want to go with the square 581 coils. I can ramp up to that easily on the dwell table. And the “spark duration” field is pretty much something to leave at 1ms and ignore? Or should I have my dwell set to 4 and my spark duration set to 1 for a combined dwell of 5?

Is there any evidence of the 581 coils becoming overly saturated anywhere nears 5 ms and “auto firing? Or as I understand it is that just the D-585 heat sink style coils?

The milliamp values were also taken from the MS MEGA MANUAL

http://www.megamanual.com/seq/coils.htm

 

The modifier table values are the multipliers in HP Tuners.

I have tested higher. No gain.

 

Hi FF86, WRONG = 4ms + Anti dwell of 1 = 4ms Coil Dwell, NOT FIVE.

IGNORE the 1 ms of Spark Duration.

I would use the provided tech.

Lance

 

theres a thread on yb where a bunch of reputable guys, including kurt urban, said they were going away from the d585s and were seeing better results from the square truck coils.
there might be some information in that thread that would help you.

 

Hi TS, I too agree and with my GEN-IV coil design (Delphi RFQ) fitted WITH a 'finned" IGBT Heat Sink and the placement of a Rare Earth magnet in the steel.
My GEN-IV coil when tested back/back to the "585" will outperform when tested in my CO-2 pressure chamber AND there is no dwell "limiter".

Most funny remarks = A Smart Coil is equipped with a current sense IGBT AND the ability to spark when the coil decides it is best ? (EARLY)
LET the ECU MAKE that decision as an EARLY spark is the same as stating the Ignition Timing is too Advanced !

Now for FF86, some MS history : The EFI332 project was a Time Based (BG) project when the 68332 processor was released.
John Meaney just graduated (1985) creating the DFI (HC-11) ECU for his Theses using some of the "open source" code. (Time Based)
This was sold to Accel (John not paid well) which became Accel DFI.
That "cold slap" to John caused him to go to TCI for their Transmission Controller requirement for Felpro leading to the Birth of the FelPro EMS.
Next Felpro (TCI) division was SOLD to CompCams creating the renamed FelPro brand to the the FAST XFI brand.
That end (John not paid well) caused John to work for Holley creating the Commander 950. (Time Based)
TODAY John's BS-3 (BS-2 sold to Vick) is a very well accepted Drag Race EMS. (Degree Based)

MY WORK in early 1980 was a Degree Based (60-2) Ignition System.
Then the use of the popular HC-11 microprocessor combined with our HiREC chip created the first DIS Engine Management System, DEGREE BASED. (1985)

THUS the reason for the outdated MS manual "tech" that concerns a Time Based ECU, NOT a Degree Based DIS ECU with COP/CNP as in a MS-3.
There IS much good tech in the MS Manual, just know much was created in the 1980's by many who had little modern knowledge.

The testing of a coil at 12VDC with Dwell Comp at 12VDC IS NOT a SAE method.
That standard is 14 VDC with a known "pico" load.

I hope this history helps you understand the difference between Modern DIS and early EFI.

Lance