GTVEVO

iTrader: (22)

Ok first of all this topic has been raised several times that when people move to the AEM the ignition seems to really go down hill compared to the stock ECU. So after some time and some thoughts I started trying to figure out why. As I explain and look for answers please feel free to pipe in your thoughts as I share mine but lets keep it clean.

So the first thing I have done is look at the AEM base map, I noticed right away that the coil dwell vs RPM table seems very low compared to the DSM. Its such a similar setup electronics and all why so different. I do know that the coils and igniter set up is different on the EVO as they are combined in the coil but in all honesty the EVO coils should operate better but under the AEM they do not. Running the DSM coil on the EVO is much much better using the DSM coil settings.

The next thing I noticed between the two is the Dwell Factor along with the Dwell Max and Min settings. After doing some research and understanding these values I don't understand why the Max setting on the EVO base map is .5 Teeth. Allow a max charge of .5 Teeth means the coils is only allowed to charge 25% of its potential 100% duty cycle. Wow if I am right what a waste, I know that you can't run the coils at 100% duty but 50%-60% duty should be very possible without issue I would think. Now to reach this duty you have to figure out what the current charge time and duty the car is running at certain RPM's.

Next I calculated out a comparison of charge time between the DSM and EVO coil dwell base file settings.

DSM at 8000rpm
5.779685 usec


EVO at 8000rpm
1.822555 usec

So max power is 1/3 of what the DSM coil can produce? I wouldn't think so.

Big difference, no wonder the DSM coil runs so much better on the base settings but still why are they like this? I dont see any reason the EVO coil couldn't run a similar charge. Even if you figure up the DSM settings, it is only set to run a max of 50% of its duty cycle since its Max Dwell is set to 1. I think both might have room for allot of improvement.

This is just what I am thinking, next will be testing so stay tuned.

Spaceball 1

iTrader: (8)

I will help you with the testing if need any. I have a 35r car that runs only C16 and high boost on the stock ign. Just let me know if you would like me to test some things to help you out. I have not even begun to look into the dwell settings yet as i have not had any blow out issues yet...

EVO8_PR

iTrader: (5)

Humm, has anyone else noticed or tried this?

GTVEVO

iTrader: (22)

Ok guys I did some testing today, first please know that I have never been able to run my stock coils after I installed my GT3076 after about 22psi on pump gas and even at that the car would try to misfire unless I gapped the plugs way down to like .018 and they would only last a few days before they started crapping out it seemed.

Today I run a complete AMS part car (race long block) and GT37R (PT67 PTrim) and all supporting mods. After my ignition troubles on the GT3076 kit I built my own 1g ignition conversion and it worked very well but is seeming to get sketchy around 32psi but I know that DSM coils have been used way past this point so when I started seeing this happen again I feel there is still something not right with the AEM coil charge settings from the base maps (DSM 1g and EVO) the best I can tell but the DSM 1g settings aren't as bad as the EVO settings.

My testing today was just that a test to see if I could run the stock Coil packs with e85 and Meth spraying with my new settings that are still very consertitive at the best of my knowledge. With my changes I went from an ign charge of 1.8 -1.9 uSec to a New charge of 4.2 - 4.3 uSec and the Results are Very Very good the best I can tell. This is with the Max duty of the coil being set at 1 or 50% which is very low compaired to what I believe it could do. I was able to due multiple pulls and I stopped at 31psi with e85 (Dual Pumps) and Meth because my 1150's are about 90% +/- right there. I couldn't of ran close to that with the stock EVO 8 Settings.

I started out with the stock settings in the wizard and what I change to on my first test is posted below. Please be aware this is just for testing and coil packs can burn up if pushed too far but I feel this is a pretty safe test as I am going to continue to run it and push the coils to see what I can get out of them.

So here are the settings I went to

Dwell vs Batt Volts = Same as stock EVO VIII Coil settings in the Wizard

Dwell vs RPM = 105 across the board, this will give you a steady charge time and not allow it to drop off until you hit your max charge time/ charge duty cycle set by the Dwell Max

Coil Dwell Factor = 35 (Same as Stock EVO VIII Coils settings in the Wizard)

Dwell Max = 1 (This means a max charge of 1 tooth regardless of tables or 50% since each rev is 2 teeth and we are working with a wasted spark)

Dwell Min = .1 (This means the min charge of .1 tooth regardless of tables so 5% duty)


Guys I will keep you posted with other settings I test because I believe we are on to something.

GTVEVO

iTrader: (22)

Spaceball, If you are running high boost and not seeing any ignition issues you must be running a higher Dwell Max, if you don't mind lettings us know what you are running on that variable? I don't want anyone to disclose their tuning or feel they are being asked too but I think it could help us learn if it is different than the .5 that the stock EVO cal comes with since I feel it is so low and not allowing the coil to charge effectively at high rpm.

GTVEVO

iTrader: (22)

On next find, Has anyone looked at the difference in the base cal on between the EVO 8 and EVO 9. The EVO 9 is running the Coils at 200% Max Dwell so I calced up the charge time on the EVO 9 Base cal and its around 3.33 uSecs (13volts at 7200 RPM). This is still way better than the stock EVO 8 Cal file and once again I don't see any reason that they shouldn't be the same.

Keep in mind the settings I am testing are allowing for more charge and have greatly improved the coils/spark. But it looks like maybe even AEM decided to do some testing and try to improve the EVO 9 base cal. Still 200% Max Dwell and 0% Min Dwell doesn't make since from a safety perspective to me. I don't think any coil could live Dwelling that high and you never want to coil to not have a min so you can block turning it off if possible via Dwell settings. Just my .02.

I think this just identifies that AEM's base map for the EVO 8 was simply a base map as they have stated. Even though I figured the Coil Dwell Wizard would set the coils up properly just as the Map sensor Wizard would but I don't think this was the case.

Spaceball 1

iTrader: (8)

Actually i running the AEM 1313 MAP settings for the Evo 8 on the coils. NGK 8's at .020". I do have a little blow out if i roll into a high gear, high load but it clears up after a second. It also could be comp surge.

Spaceball 1

iTrader: (8)

I have not looked into this until now and found some interesting things.

Ok i have a ton of differnet calibrations from a ton of different cars. Some from well know shops and some from the lesser known shops. Most are the exact same as the original AEM in the dwell settings. There is one however that is exactly what you are running for the Max dwell and min dwell. They are however not changing the dwell vs batter volts. They are changing the dwell vs. RPM, I do not want to post the info though. GTVEVO if you are interested PM me.

GTVEVO

iTrader: (22)

Thanks for the info I appreciate it, I believe that I am right with what's been going on in regards to the poor ignition performance on the 8. I did some additional testing today and moved my Max Dwell up to 1.25 just to verify that with setting at 1 in higher rpms I was getting capped by it as my charge time would move down about .1-.2 at about 8k. With the 1.25 it actually goes up due to the Dwell vs BattVolts moving down in due to battery voltage. This causing the coil to dwell just a bit longer to have the same charge. I am still running the stock EVO 8 Dwell vs Batt table and the Dwell vs RPM still set at 105 across the board and Dwell factor of 35. I am continuing to run these settings with no issues and nothing but good results.

Spaceball it sounds like you would be another good test to run the settings I posted in my 2nd post to see if your ignition issues clear up any. If you do decide to test run them please watch your logs to make sure everything stays looking good to be on the safe side.

SwiftEVO

iTrader: (32)

Love stuff like this thanks man!

Spaceball 1

iTrader: (8)

GTVEVO I did a couple pulls today with my new settings. I did not go to 105 across the board but more dwell in the higher rpms than the standard AEM map about 75. I did not log any pulls but the car came alive and sounds different at high rpms.

I missed my dyno time today cause i ending up running to Norwalk to watch bushur and Jamie run his 8 second supra. When i got back i was too tired to drag out the laptop and get some street logs going. I will try to get some logs this week or weekend.

I can feel a difference and i really didn't think i was even having a problem. Maybe it was the cold air but the car just sounded more aggressive in the higher rpms also...

GTVEVO

iTrader: (22)

If you have raised your max dwell to above the .5 you would indeed being seeing more charge time. Going from around 1.82 at the stock 40's setting to around 3.17 at 75 depending on your batt voltage. I am still running my max dwell at 1.25 which should still be very safe but to keep me from getting capped at higher rpm. When I started experimenting my car also sounded much smoother and crisp at the higher rpms as I ramped things up so I know it has to be doing some good. My plugs even seem to look cleaner as e85 was seeming to foul them out quicker than pump fuel. Thanks for testing!

Spaceball 1

iTrader: (8)

Check my math at

1000 rpms it takes 60ms to do one revolution
8000 rpms it takes 7.5ms to do one revolution

with that then since we have wasted spark each coil fires once every 60ms for 1k rpms and once every 7.5ms at 8000 rpms since there is 2 revolutions to do a complete cycle.

So that dwelling 2 teeth is 100% duty on the coil since there is a wasted spark.
There is 8 points being "seen" during a complete cycle of all cyclinders, but each coil must fire 2 times per revolution due to the wasted sparks. GTVEVO are you off? or me?

You can't really dwell much longer than your 1.25 max cause that is 62.5% duty cycle?

I am just kinda speaking aloud and you really got me going on this subject.

With that Ignition charge time = 2*dwell factor*dwell batt* dwell rpm

i get 3.15ms charge time at 8000 rpms which is no where near to the 7.5 ms charge time of the possible 50% duty cycle possible due to the max setting of my 1 tooth max.

Please someone check my work

Spaceball 1

iTrader: (8)

If am correct we can get a way hotter spark in the cylinder than i or GTVEVO are even currently running.

Maybe we can completely eliminate this COP stuff.

Anyone know how long the discharge takes?

GTVEVO

iTrader: (22)

Ok I was thinking that the EVO was 4 teeth per rev but I think you are right its 4 teeth per cycle (720 deg) so at 1.25 I would be getting close to my cap of 62.5%. You are correct but I am still getting much more charge and better ignition performance and more charge time. I still don't understand how you came up with 18.75 ms? I will go back later and correct my findings but I don't think I would want to run much more than the Max Dwell of 1.25 without possibly damaging the coils. I think running in the 60% range is probably closer in comparison to the stock ECU performance.

At 2*35*.6055*105 = 4450.425/1000 = 4.45 and change. This is what I calced and see in my logs with in a tenth or so.

When I figure up 62.5% duty at 8k I get the figures below.
8000rpm / 60 sec = .00075 or 7.5 then .00075 * .625 = .0046875 or 4.6874 uSecs of Charge time max at the 1.25 Max Dwell Setting.

Good find and thanks for bringing in the light, its much appreciated.

With Spaceball findings upon Teeth being for 720 deg or 2 revs I have corrected my calcs above and in previous posts so this thread stays as accurate as possible.