508 Hard Starting | AEM
8 posts / 0 new
Last post
kromberg
Offline
Last seen: 2 weeks 4 days ago
Joined: 03/03/2018 - 14:54
508 Hard Starting

I have a 508 with the LS 24x hardness and the car is having a hard time for cold and warm starting.   It is needs to be cranked for quite a while( 5+ seconds ) before it will fire up.   I am looking for some help how I can improve it. 

 

Files attached.

Start.doc

 

Start01.itssn

Start01.itlog

 

Start02.itssn

Start02.itlog

alexui90
Offline
Last seen: 1 year 1 week ago
Joined: 04/30/2015 - 16:08
Post your log and session

Post your log and session with all tabs layout so i can see all the parameters

alexui90
Offline
Last seen: 1 year 1 week ago
Joined: 04/30/2015 - 16:08
Post your log and session

Post your log and session with all tabs layout so i can see all the parameters

kromberg
Offline
Last seen: 2 weeks 4 days ago
Joined: 03/03/2018 - 14:54
I think the files are

I think the files are attached to the first post.

AEM_SB
Offline
Last seen: 2 weeks 3 hours ago
Joined: 06/07/2014 - 14:29
Your files were attached (on

Your files were attached (on our server somewhere) but not inserted (viewable in your post). I edited your post to insert them, and I'll see if we can have one of our guys look at the logs later tonight or tomorrow.

kromberg
Offline
Last seen: 2 weeks 4 days ago
Joined: 03/03/2018 - 14:54
Hey AEM_SB, any chance anyone

Hey AEM_SB, any chance anyone has had a chance to look at the logs and sessions for and suggestions?

AEM_SB
Offline
Last seen: 2 weeks 3 hours ago
Joined: 06/07/2014 - 14:29
The ECU needs to see 1-2

The ECU needs to see 1-2 complete engine cycles (2-4 crank revolutions) to see the crank & cam signal pattern it needs before firing. Anything you can do to spin the engine more quickly on the starter (fully charged battery, and good wires and grounds for the starter) will help it complete those 2-4 crank revolutions in less time. In the log Start_01 it looks like the ECU has turned SyncState on (which means it will firing injectors and coils) for about 3+ seconds before the engine starts. That suggests the cranking fuel tables need some adjustment.

 

That data looks OK in terms of the crank & cam signal inputs, and there aren't any rev limiters or cuts active.  Starting an engine is relatively simple: it needs air, the right amount of fuel, compression and a correctly-timed spark event to light the mixture. You can be pretty confident that the injectors are delivering fuel and the spark plugs are firing at approximately TDC (check using a timing light), because the engine start up eventually. Double-checking engine compression is not a bad idea if you have a compression tester nearby. You may simply need to make some adjustments to how much fuel the ECU is commanding. Here's a copy/paste from a previous email I sent to another customer regarding cold start tuning, hope it helps.

 
-------
 

If the engine is getting enough air to idle without stalling when cold, then the next thing to adjust is how much fuel the ECU injects when cranking.   

 
Below 500 RPM the ECU is calculating fuel using just two tables from the 'Start' page in the layout: CrankVE_Table and LambdaCrankingTable. For the most part, it's best to leave CrankVE_Table alone and focus only on LambdaCrankingTable. LambdaCrankingTable lets you set the desired air/fuel ratio to inject depending on the coolant temperature. The numbers in the table are usually going to be richer than normal running air/fuel ratios, because engines tend to need more fuel when cranking than when they are already running.
 
So the first thing to do is decide which coolant temperatures already work well for starting, and make sure to leave those parts of the table alone. In general, there will be some range of 'enough fuel' that will let the engine start in less than 3 or 4 seconds of cranking. If the engine is taking a long time to start, less fuel (larger numbers in the LambdaCrankingTable) will usually make it take even longer to start. That said, it's still wise to test with less fuel just to confirm the engine takes longer to start. Too much fuel could flood the engine and get the spark plugs wet, which will prevent it from staring properly later even if you get the mixture correct. So I would try one cold start with less fuel (assuming coolant temps are about 70 degrees F try 14-15 AFR Gasoline instead of 10-11 AFR Gasoline)... then once you've confirmed less fuel doesn't help, try adding fuel by using richer values in that table (12 AFR, 11 AFR, 10, AFR, 9, 8, 7, 6, 5, 4, 3...) until it starts.
 
 
The trick with cold starts is you only get one or two of them per day... as soon as the engine fires up and idles for a few seconds it will usually re-start too easily compared to a true cold start. Then you need to wait multiple hours for everything to cool back down and dry out.  Because of this, it helps if you can wait for the ECU's onboard O2 sensors to warm up before cranking the engine (warm sensors should show about 18 or 20  AFR if the engine has been off for a few hours). It think it also helps record datalogs so you can go back and see what the O2 sensors indicated right after the engine started. It also lets you double-check the time between the battery volts dropping (due to starter motor) and the ECU injecting fuel (due to getting signals from the crank & cam sensors). Comparing the Lambda readings after a good start against the Lambda readings after a slow start might be helpful.

 

I like to set ignition timing near zero degrees and that should be close enough to start. Too much ignition advance when cranking will result in the engine 'kicking back', remember the piston is moving very slowly so the combustion event might actually try to spin the engine backwards if you fire the coil at 30 degrees of advance at 100 RPM.  Leave the CrankVE at 75,  and sweep the LambdaCrankingTable values starting at lean (18 AFR) and decrease by about 1 or 2 AFR every second of cranking. Sweeping through the fuel values like that should start the engine.  Don't overheat the starter, give it a rest after 5-6 seconds of cranking.

 

Hope that helps,

SB

kromberg
Offline
Last seen: 2 weeks 4 days ago
Joined: 03/03/2018 - 14:54
Thanks SB, that advice has

Thanks SB, that advice has made a huge difference.   I changed the CrankVE_Table table back to the original values and have adjusted the LambdaCrankingTable.   Cold starts are night and day different, but still need a little tweeking.   At least now I know what needs to be done.   Again, thanks for the help.