AEM series 2 cold start refining/tuning
07-18-2011, 08:33 PM
#1
AEM series 2 cold start refining/tuning
I'm tinkering with the cold start settings; it's not practical to pay the tuner to look at the car at 6am each morning. if someone could point me in the correct direction on what settings i should be using to get the car to 'start faster'.
I am under the impression that initial crank pulse and start extra vs temp are the adjustments that affect how well the car starts at each temperature. I think start extra is pretty much right for the "warm starts", perhaps cold needs some tweaking.
I started datalogging the startups at relevant temperatures, but i dont really see anything obvious that would give the car a hard time at 20 - 30 celcius; (40-60 celcius untested at this point). what should i be looking at? i don't see anything wrong once i am actually getting data from the running and warming up engine.
The fuel pump primes everytime like it should for 5 seconds. NO gas pedal input was needed to start the car so far. laptop was connected so i would know when the cam synced and the injection pulse was happening. I can get the car to start perfectly in one try at 65celcius "warm start" according to the attached datalog (i used this datalog to check if there was anything substantially different when starting at different temps). I am confident that once i get this dialed in the car will cold start just like stock.
When starting at 21 (overnight/morning) and 27 (leaving work after 9 hours) degrees celcius, the car always takes two tries at these temperatures. The 2nd attempt starts the car perfectly whether i cranked a lot or a little the first try. i have no ideas for this. On the first attempts, there's just no sign of life. If someone could check out these logs and make suggestions on which way i should go with this (more or less gas needed) that would be great.
Hard start from 27 celcius 7-18 1650 low fpulse.stf
Hard start from 21 celcius 7-18 0638 low med fpulse.stf
Easy start from 64 celcius 7-17 2130 default fpulse.stf
I am under the impression that initial crank pulse and start extra vs temp are the adjustments that affect how well the car starts at each temperature. I think start extra is pretty much right for the "warm starts", perhaps cold needs some tweaking.
I started datalogging the startups at relevant temperatures, but i dont really see anything obvious that would give the car a hard time at 20 - 30 celcius; (40-60 celcius untested at this point). what should i be looking at? i don't see anything wrong once i am actually getting data from the running and warming up engine.
The fuel pump primes everytime like it should for 5 seconds. NO gas pedal input was needed to start the car so far. laptop was connected so i would know when the cam synced and the injection pulse was happening. I can get the car to start perfectly in one try at 65celcius "warm start" according to the attached datalog (i used this datalog to check if there was anything substantially different when starting at different temps). I am confident that once i get this dialed in the car will cold start just like stock.
When starting at 21 (overnight/morning) and 27 (leaving work after 9 hours) degrees celcius, the car always takes two tries at these temperatures. The 2nd attempt starts the car perfectly whether i cranked a lot or a little the first try. i have no ideas for this. On the first attempts, there's just no sign of life. If someone could check out these logs and make suggestions on which way i should go with this (more or less gas needed) that would be great.
Hard start from 27 celcius 7-18 1650 low fpulse.stf
Hard start from 21 celcius 7-18 0638 low med fpulse.stf
Easy start from 64 celcius 7-17 2130 default fpulse.stf
AEM EMS S200 Basic Setup
Engine Start
Set Crank Injector Time Table to (0,0) and Ramp up Linearly across the Throttle Range andTry Starting Until the Correct Value is found by opening the throttle more each time **i found the right value; no gas needed**
Modify Initial Crank Pulse Table to Get Car to Start Faster **this is what im looking for**
Leave the last few breakpoints of the Crank Inkector Time Table to 0 as for Flood Clearing **got it**
Engine Start
Set Crank Injector Time Table to (0,0) and Ramp up Linearly across the Throttle Range andTry Starting Until the Correct Value is found by opening the throttle more each time **i found the right value; no gas needed**
Modify Initial Crank Pulse Table to Get Car to Start Faster **this is what im looking for**
Leave the last few breakpoints of the Crank Inkector Time Table to 0 as for Flood Clearing **got it**
07-26-2011, 08:33 AM
#4
Thanks a lot for the education. I have it pretty close to dialed in now, i was able to start the car cold in 1 try this morning. I should have it perfect in a couple of days. i never needed to use any throttle input to get the car started. The car runs perfectly fine even when cold once it gets past cranking (my tune is pretty dialed in). i let the car idle for 30 to 45 seconds or so after starting so that i can get a good datalog of the events.
I guess start rpm offset is the more relevant way to change the idle rather than using the base idle. both accomplish the same thing sort of...
I guess start rpm offset is the more relevant way to change the idle rather than using the base idle. both accomplish the same thing sort of...
07-28-2011, 07:10 AM
#6
i figured it out and now i know how to see the issue on the logs. my car starts strong now at 20 degrees celcius (i live in socal and the likelihood that cold start water temp is ever below 15c is very remote) and now i know what to change to get it to start anytime. ill post some logs and screenshots later. it wasn't the initial crank pulse, it wasnt start extra, and it wasn't idle rpm base or offset. it needed exactly 30% more gas (surprisingly i dont think this table is temperature compensated either); but i was approaching the wrong tables to give it more gas and wasnt even getting past the 'cranking stage'.
Try to keep in mind that you're not going to be able to fire the engine off and have it 14.5 afr in 5 seconds. Port injected piston engines are dirty. This is why the hybrids don't use the starter motor to fire off... If you want to "cheat some" put a scope on your car(when it's stock) and get some data on injector pulse time. This will make it a whole lot easier for guys that don't do this everyday. The next thing is "acceptable" the closer you get to having the engine clean and lean the more work it becomes. This is why our fathers and grandfathers love their carburetors, rich is easy and drives well... If the AEM has "stat synced" and it's still cranking it's up to you to figure it out. Typically when they don't have enough fuel they crank too long, add fuel and try again.
Trending Topics
04-14-2012, 09:06 PM
#10
i refined it further since, but i dont have any method to read to logs to get it past the point i illustrate above. i can cold start 95% as well as stock. All i can say is the car is probably overfueled during cold start and not enough ignition timing on the typical basemap.
it is definitely possible to get stock like starts with the series 2 with the right settings. it took a lot of trial and error to figure out what it likes.
there are a lot of variables. you have the left 3 or so columns of the fuel map, initial injector pulse, start injection, and battery compensation to deal with. remember when youre cranking the voltage probably drops to like 11v. the battery offset table will dump more gas during cranking.
the battery offset probably needs to be mellowed out a bit. start injection curve and crank timing is probably the next best place to start. the initial injector pulse i found to be sort of useless it's very minimal compared to original basemap.
it is definitely possible to get stock like starts with the series 2 with the right settings. it took a lot of trial and error to figure out what it likes.
there are a lot of variables. you have the left 3 or so columns of the fuel map, initial injector pulse, start injection, and battery compensation to deal with. remember when youre cranking the voltage probably drops to like 11v. the battery offset table will dump more gas during cranking.
the battery offset probably needs to be mellowed out a bit. start injection curve and crank timing is probably the next best place to start. the initial injector pulse i found to be sort of useless it's very minimal compared to original basemap.