Weird A/F swings 04 GTO LQ9

[orange15]

Hello All,

I am having a hell of a time trying to dial in my VE table. Granted, I am trying to tune in the street, but my issue is that for a given tune the A/F will seem to go lean overall as I am driving and logging. For example, my idle will be great at 14.7 for the beginning section of my log, but then will work its way up to 15.3ish and climb higher. My coolant temp stays within 5 degrees F during my drive, and I have the EQ set to 1.0 for the whole tune so I can focus on just the VE table for now. I am not sure if I am missing something simple that is messing with the VE table?

Here is the log and tune file.

5thVE tune.hpl
engine_swap_base_VEtune_sd_3_smooth_dropped idle.hpt

Any help would be awesome!

Thanks,
Bailey

[orange15]

Forgot to mention my mods.
Stock LQ9 bottom end, and stock heads
.66" dual springs
ls7 lifters
longtube headers
36lb/hr injectors
btr stage 1 twin turbo cam
7.425" pushrods
K&N CAI

[matty b]

Given your log, you are a good 10% rich from 3-4k rpm. Do you know how to do an AFR Error histogram? 100 * (Actual AFR - Commanded AFR) / Commanded AFR

[Doug Doty]

I am new here and still green !! but with that said when I study your scan I see something familiar to myself from last week working on my stock 6.0 chevy truck. I was having those wild swings in timing dropping to single digits or 15 to 25 degree swings as well and I could not stop them until i killed torque management completely on my 4l80e. Why I mention this is your lean peaks seem to coordinate themselves with the drastic timing retard events perfectly in one area I was studying. This would seem to me that 2-5 degrees of advance is not sufficient advance to get the same burn as the mid 20's timing is and I would guess that possibly the lean spikes are from a lean tune and the rich could be less efficiently burnt fuel from those low advance spikes. Mine cleaned up w/o the torque mgmt. and give me smooth graphs. My truck in stock form is using 8.5% additional fuel in the LTFT's and I assume this is in part to the E10 gas I have available these days -vs- factory tune for straight gas.

Also the only severely lean spots look like TPS pullbacks and is to be expected.

Once again I am too green to provide great info but just mostly wanting to bring your attention to the correlation between the timing and the afr events on the scan. Anxious to see what more experienced people think.

Doug...

[Fast4.7]

I am new here and still green !! but with that said when I study your scan I see something familiar to myself from last week working on my stock 6.0 chevy truck. I was having those wild swings in timing dropping to single digits or 15 to 25 degree swings as well and I could not stop them until i killed torque management completely on my 4l80e. Why I mention this is your lean peaks seem to coordinate themselves with the drastic timing retard events perfectly in one area I was studying. This would seem to me that 2-5 degrees of advance is not sufficient advance to get the same burn as the mid 20's timing is and I would guess that possibly the lean spikes are from a lean tune and the rich could be less efficiently burnt fuel from those low advance spikes. Mine cleaned up w/o the torque mgmt. and give me smooth graphs. My truck in stock form is using 8.5% additional fuel in the LTFT's and I assume this is in part to the E10 gas I have available these days -vs- factory tune for straight gas.

Also the only severely lean spots look like TPS pullbacks and is to be expected.

Once again I am too green to provide great info but just mostly wanting to bring your attention to the correlation between the timing and the afr events on the scan. Anxious to see what more experienced people think.

Doug...

Just FYI to compensate for E10, go to a commanded stoichiometric air/fuel ratio of 14.1:1. Often fuel trims will drop right in line on a stock tune and stock engine.

[orange15]

Given your log, you are a good 10% rich from 3-4k rpm. Do you know how to do an AFR Error histogram? 100 * (Actual AFR - Commanded AFR) / Commanded AFR

I have not had the chance to really hit 3-4k very well because I am trying to dial in my lower RPM VE table. I do know how to use the AFR error histogram and have been using it. My concern is that overall (All RPM and pressure levels) are swinging leaner as time goes on. This is what I do not understand. Say I am running 14.7 at idle, and 13.5 at 3-4k rpm, then it will swing to 15.3 at idle and 14.2 at 3-4k rpm later. Not sure what is causing these inconsistencies in the same driving log.

[matty b]

I have not had the chance to really hit 3-4k very well because I am trying to dial in my lower RPM VE table. I do know how to use the AFR error histogram and have been using it. My concern is that overall (All RPM and pressure levels) are swinging leaner as time goes on. This is what I do not understand. Say I am running 14.7 at idle, and 13.5 at 3-4k rpm, then it will swing to 15.3 at idle and 14.2 at 3-4k rpm later. Not sure what is causing these inconsistencies in the same driving log.

You really arent that far off Commanded in the lower RPM and start to richen up beyond commanded around 2k.

I'll give you a bit of advice to slow down your brain a bit, dont worry about mixtures in the load range youre in right now. Do your AFR error histo and tune the VE and MAF to get those errors to close to 0% if you can. I agree that you should change your Stoich AFR value to 14.1 as opposed to 14.68 as we just dont get pure gas anymore, but even then youre pretty close to commanded and thats just fine.

When you start getting into the 90-100kpa zones and AFR error is positive you got a problem, I dont see that happening the way youre set up. Just drive and log AFR error AND turn off your DFCO or youll have to make your histograms disregard lift off throttle.

[orange15]

Thanks for the info, Matty b. So - I didn't touch the tune, and the AFR was near perfect this morning on the way to work, even after the vehicle was at operating temp. Could it be something to do with the IAT that's making my car run lean? Does the AFR just very depending of the temperature of the air going in? Seems like the hotter it is outside, the leaner it runs.

[kingtal0n]

technically the hotter the air gets, the less dense the air is, the richer it should run.


However this is rarely the real world result. Often things such as:
1. rising temperature increases resistance of electrical circuits such as injector driver circuit, things as the alternator gets hot, voltage output seems to diminish
2. rising temperature means fans run longer and put more electrical load on the alternator which can lower voltage
3. "" does mean less dense air, and since the friction/power requirement of the engine is only barely reduced through thinning oils the engine still needs more air volume than when it is cold, which means more power is lost on the intake stroke and needs to be made up for somehow, which usually results in higher open throttle position (higher IACV counts) when fully warmed up compared to when the climate is cold.
4. "" will/may have a slight effect on wideband readings, depending where they are located. For temperature and possibly pressure reasons. Hotter exhaust gas is expanded more and there will be higher pressure and temp.
5. fuel specific gravity may change significantly if it warms up enough to alter a/f ratio slightly leaner, remember air, water, gasoline are all fluids with specific weight which is density times gravity.
6. Because of all the changing variables, the engine will typically 'run' and 'idle' in a different region of the VE map. It means when you tune the idle cold (say 550rpm and 45kpa) you might not be tuning the engine for when it is hot (550rpm and 52kpa) it could infringe into other cells when warm that it does not when cold
7. because the injector delay map is vs voltage, and voltage is affected by warming resistance of electrical system, the voltage injector delay map may not be compensating enough for electrical voltage drop at the injectors

8. I am sure there is more. The bottom line is there are many reasons why it could happen and many ways to compensate for it. And perhaps most importantly to know that you are not alone. You pretty much described an extremely common occurrence and the ideal method for dealing with it may vary, i.e. your ultimate solution may not be the same for every engine you visit.

[orange15]

King Talon, awesome writeup! Do you think it would maybe be effective to run the car for say an hour+ and then do my AFR error scanning so that I can cover my lean conditions? What are some strategies to nail this down? Just want to make this tune as solid as possible. I also plan to just do SD and not have any MAF because I am going to FI in the near future anyways
Thanks

[orange15]

I believe I found my answer. My IAT values are going up to 150* when the car is idling. Timing is the retarded from the IAT temps, which changes my manifold pressures. This then changes my AFR, because it is hard for me to replicate these values without abnormal idle conditions like this.

[kingtal0n]

I put this sensor
https://www.diyautotune.com/product/...-with-pigtail/

In an aluminum pipe right before the throttle body. I then put a heat shield around that. Its after an intercooler. Seems to work very well.

Usually iat will heat soak sitting in a manifold or similar positions where surrounding temperatures (my manifold runs 165*F for example) dramatically warm the sensor.

[orange15]

I put this sensor
https://www.diyautotune.com/product/...-with-pigtail/

In an aluminum pipe right before the throttle body. I then put a heat shield around that. Its after an intercooler. Seems to work very well.

Usually iat will heat soak sitting in a manifold or similar positions where surrounding temperatures (my manifold runs 165*F for example) dramatically warm the sensor.

Thanks for that info! Yep - similar here. I noticed that when I returned to idle, my IAC jumped to 160*F and I was pulling like 4* of timing, thus causing my MAP to change. My AFR error table looks much better now. Just a few more rich spots.

8th VE tune.hpl