Maggied G8 GT Cold Startup Stalling Issue

[Aaron407]

My car sat outside all day today, and the coolant temperature showed as -2 C (28 F) on my Aeroforce. I had my laptop and tuner along to log a scan, but it decided not to cooperate and wouldn't boot into windows. So, after sitting in the car for five minutes while the computer tried to boot up and the windows fogged up, I fired it up, and it still had the same issue after all of the changes noted above (with a few small adjustments to smooth things out and not affect other scenarios). It was fighting down around 100 RPM to stay alive until I tapped the throttle to give it more life, after which it ran just fine. Man, I wish I could have grabbed a scan as I'm thinking that there's really not much else to work towards without one to shed some light on things. Hang in there with me, I'm sure I'll need some help interpreting...

EDIT: Attached is the newest tune that I was running today (hopefully the forum attached the correct one).
EDIT #2: I've also attached a config file that I was going to use for scanning the startup. If there's anything that I can/should strip out of it or change, please let me know.

[gmtech16450yz]

My car sat outside all day today, and the coolant temperature showed as -2 C (28 F) on my Aeroforce. I had my laptop and tuner along to log a scan, but it decided not to cooperate and wouldn't boot into windows. So, after sitting in the car for five minutes while the computer tried to boot up and the windows fogged up, I fired it up, and it still had the same issue after all of the changes noted above (with a few small adjustments to smooth things out and not affect other scenarios). It was fighting down around 100 RPM to stay alive until I tapped the throttle to give it more life, after which it ran just fine. Man, I wish I could have grabbed a scan as I'm thinking that there's really not much else to work towards without one to shed some light on things. Hang in there with me, I'm sure I'll need some help interpreting...

EDIT: Attached is the newest tune that I was running today (hopefully the forum attached the correct one).
EDIT #2: I've also attached a config file that I was going to use for scanning the startup. If there's anything that I can/should strip out of it or change, please let me know.

Don't be discouraged, I still think you're on the right track!

And yeah, having startup logs would definitely help, but you've said something a couple times that to me is a MAJOR clue. Look at what I highlighted above. It will stay running if you push the gas pedal a little right? THAT'S IT. What is that clue telling you? It evidently needs more air. That's what you're doing when you open the throttle more and it runs. That tells me that the fuel pressure, cranking fuel, startup fuel and ignition timing SHOULD be ok.

So here's what I would do at this point. Just for kicks, do a compare between the tune where you made my changes and the one before it. It should show you the changes you made to the startup airflow tables. DOUBLE those changes. Just select all the values shown in the "show differences" and multiply by 2. Save that tune as "doubled changes" or something like that, then load it in the car. Sooner or later you're gonna get to the point where it races on startup. If that happens, no problem, just shut it off.

If you're not familiar with the showing differences stuff, just let me know and I'll walk you through it. Try that and see what it does.



Edit- try using this config file. There are a few things missing from yours.

General LS config37 12-6-14.cfg

[Aaron407]

I actually managed to find an old scan of this startup issue (I thought I hadn't saved it). Note, however, that this was early in the tuning stages and I had been playing with transient fuel, which might have contributed to some of the ridiculously high IPWs. Anyway, what I noticed is that it actually looks fine for airflow, and never seems to have a distinct dip. It looks more like the fueling might be the culprit, though. It looks like it fell on its face after the IPW went to the injector minimum, then they ramped up without much luck. Anyway, maybe this will give a bit more insight for the time being. I'm hesitant to add more air since, when it's just a few degrees warmer, it fires up hard without any issues. I'd like to get an up-to-date scan before many more changes, unless you can suggest anything based on the the one attached.

Now I'm going to head out and find out which PID in VCM Scanner is causing the scan speed to drop like a rock...

[gmtech16450yz]

Ok, I'm running out the door right now but I did see a couple things in that log that definitely don't look right. Commanded AFR goes hella lean right after startup, right when it dies. Look at that in the tune. Also, why is the alternator startup delay so long? It would be nice to have 14v sooner. Gotta go, I'll look more later. And you're right, get a good log before doing anything else. Kinda hard flying blind on something like this!

[Aaron407]

Thanks. I really wonder if it has something to do with what it's determining as an intake valve temperature and how that's affecting the commanded AFR selection when it moves to the open loop equivalence ratio table. Commanding a mid to high 12s is, to my knowledge, fairly lean for a cold startup, but it was apparently appropriate when it was NA. I'm going to make sure I catch IVT on the next scan so I can track down which cells it's referencing on the OL EQ ratio table. I also noticed a strange transition in that table in a colder area where it was commanding an EQ ratio of 1.15 with 1.3 on each side of it. I just cleaned that up and might richen part of the table, depending on the results of the next proper scan.

In terms of the alternator startup delay, it's still stock. I assume the delay in place is to allow the RPM to better stabilize before charging, but I'm not sure what is a reasonable time. Comparing to a CTS-V, I see that its delay is 0.6 seconds versus the 3.9 seconds for a G8. I might look at reducing that...

[gmtech16450yz]

Hey Aaron, it's 11:30pm and I finally got a chance to compare your tune to a stock '09 G8 file. I'm trying to help you here so don't take what I'm gonna say the wrong way. I'll try to be as delicate as I can! lol.

Sooooooo, I made an assumption in my first reply to your request for help. From this statement you mentioned, I assumed the rest of the tune was solid and this was simply a cold start issue.

I've tuned it extensively for performance and driveability, it starts/runs absolutely beautifully

My assumption was wrong, this is way more than just a cold start issue. I'm REALLY struggling here on how to say this because you've been so appreciative of my help and you seem like a smart guy. What you just said in your above post was very insiteful, and I was impressed with what you said about the alternator startup delay. You compared the delay to another car, came up with a theory on why they would make the delay so long, and thought about shortening it. All that was excellent! Really.

So here's the hard part. By looking at your tune changes from a stock G8 tune, I'm afraid you know just enough to be dangerous! Sorry, I didn't know how else to say it. The good news is you DO seem like a smart guy and I still think you can handle this and learn from your mistakes. After all, with 90% of the guys on here being afraid to share screenshots of their tunes, how would you ever know how to do it correctly? THIS is one of the reasons I strongly encourage posting what you're doing for others to see. Otherwise you have no idea if you're doing it right or wrong. That means me too, that's also how I learn. It's all a learning process, and even for all the years I've been doing this, I'm STILL learning new things everyday.

So here's one of the major things I see in your tune. Smoothing. When you're making changes in table values, THEY HAVE TO MAKE SENSE. In other words, they have to make sense in relation to the other cells around them. There is NO WAY your ign timing table could be making your engine very happy. I'll post some screen shots so you can see what I'm talking about...


This is a stock high octane ign timing table...





This is yours...






This is the differences between a stock file and your file...








I just grabbed one of the open loop eq tables and also did a comparison. This is the differences between a stock file and yours...







So here's my advice. I think you should start over with a stock tune and carefully re-tune everything. There obviously are some tables you can carry over if you know for sure they've been done right, but honestly I feel that most of this tune needs to be scrapped and started fresh. Again, I'm sorry if I was harsh with this, I'm really trying here. If you want, I can go into detail about making smooth tune changes, or someone else can chime in and help. Your car will be WAAAAAYYYYY happier in the end, and you'll be a better tuner. If anyone else out here wants to comment or help out, please jump in. Let me know what you want to do and I'll help as much as I can. I am hella super busy with my own projects and my job, and I've already been ignoring a bunch of other guys that have been begging for help, but I'll do what I can to help you too.

[Aaron407]

First, I will note that I do appreciate the help you have given. I'm going to try to take a breath and not take offence, but I do need to give you some background as there's more to the story.

At face value, the timing table is ugly. Oh, how I know that. The cell transitions are not smooth, but there's a reason for it. When the car was stock, it had up to 8 degrees of KR, even when running 91 (they really shouldn't claim that it's suitable for 87). I began tuning the timing from the table provided by Magnuson, which was fairly close to stock and generally smooth, and found that there was still a lot of KR resulting from their provided tune. After nailing down the AFR, I spent many iterations of timing adjustments. When scanning for KR, I found that a smooth timing curve, while getting the most power out of it, isn't really possible for my car in particular. Other cars that I've tuned haven't had issues, but this one is fairly unique. There are some cells where the timing has to be 4 or more degrees less than the cell beside it to kill off the KR (unless I wanted to blindly throw fuel at it, which I don't). It was found through many scans that these cells just don't seem to like timing, and consistently caused knock, while the ones beside them were fine. My general method was to scan for KR and reduce timing in the cell by half the measured amount, while taking into account the decay period to avoid unnecessarily reducing timing in cells that didn't cause KR. Pretty typical stuff. There were a few high airmass, low RPM spots where I also had to tweak the TCC release to avoid bogging the engine since it was also causing KR and large timing reductions, but generally just in the manual mode. I also adjusted my shift points to avoid bogging the engine in these area so the low timing cells are rarely hit.

In the end, that's the timing table that came out, and it gives me no more than 0.5 degrees of KR. Do I wish it was smooth? Absolutely. However, the only way to make it smooth would be by reducing all of the cells around the troubled ones simply to make it look visually smooth. In driving, there are (surprisingly) no noticeable shifts in power when the timing jumps around, so I've left it as is. The idea of reducing timing and power where there are no issues was just something I couldn't convince myself to do when there were no apparently driveability issues. As I said, other cars that I've tuned have allowed fairly smooth timing table transitions, but this one is an oddity. If you're aware of any detrimental effects other than "feel" for having higher timing in a cell rather than lower to better match a cell beside it, I'm all ears, but I'm also avoiding unnecessary timing reductions so I don't add extra heat to the cats.

As for the open loop equivalence ratio table, are you sure you're looking at the correct stock file? The table in my tune is almost identical to the stock table, with the only cells changed being the ones that were originally <1 changed to 1 so I don't run leaner than stoich. The stock table is in and of itself rather odd with some of the transitions, and is where I'm looking at this point to target the startup issue in conjunction with the IVT vs. MAP modifier table. That, and I wonder if the Magnuson IAT sensor is reading correctly.

Here's a comparison of mine to stock:

OL EQ Compare.png

Overall, I appreciate the sincerity. However, I'm not blindly changing timing cells as I know that's a recipe for disaster. My changes have been as a result of measured KR, which I feel is the best way to tune rather than visually. Thanks again for your help.

[Aaron407]

I think I've tracked down the problem. There are a couple of cells in the stock open loop equivalence table that commanded it to run a lot leaner than in the cells around it, and they happened to coincide with the trouble temperatures I've had (right around the freezing point). I've balanced these cells out and it hasn't had any issues, neither at the temperatures where it was problematic before nor at colder temperatures.

Stock OL EQ Ratio.png

I'm hoping that turns out to have fixed the issue, and the other changes that I made (thanks gmtech!) have allowed it to start even more confidently. I also did a bit of selective smoothing to my timing table since it sounds like it's not a good idea to have major differences in adjacent cells. I did, however, put in some logic in the change to not increase timing during the smoothing operation in order to not add timing to cells where I had reductions due to KR. I haven't noticed a difference in how it drives yet, but I'll take the word of those who know better that it will be good for the motor in the end.