Engine Runs Rough After Starting From Heat Soak

[nnicholas]

If the car has been sitting overnight I can get in, start it up and drive as long as I want without any issues. However, after it's reached operating temperature if I go to restart the car after it's been sitting for 20-30 minutes, the car starts but runs really rough. I don't think it's related to an idle setting because I can drive it for several miles after restarting and the car feels like its running on only a few cylinders. If I let it sit overnight it will work fine again in the morning. Has anyone ever seen this issue?

I thought it was related to heat soaking of the IAT which was placed where the EVAP sol used to be, but I recently relocated it to after the intercooler and had the same issue.

-5.3L 7875 turbo
-IAT right after the intercooler
-IAT bias off
-SEP 1500cc injectors

[TheMechanic]

Check if you have a leakdown issue in the fuel system. Like an injector that is leaking. Next is a compression test when the symptom is occurring. Then go from there

[kingtal0n]

There is a heatsoak table which can be used to enrich fuel for these situations after a hot-refire. The IAT probably has nothing to do with it. More likely something with the heating of fuel, injector solenoids, low voltage, injector quality, heating of the map sensor, or some combination of these

[TheMechanic]

There is a heatsoak table which can be used to enrich fuel for these situations after a hot-refire. The IAT probably has nothing to do with it. More likely something with the heating of fuel, injector solenoids, low voltage, injector quality, heating of the map sensor, or some combination of these

Where is that table? ctrl+N?

[kingtal0n]

This is mine
hotsoak-enrichment.jpg

there is a timer on it and a decay value somewhere so you can dial it in perfect

[nnicholas]

Found an injector that was stuck open. I'm going to replace it and go from there. I might try out the heat soak table you mentioned also.

[nnicholas]

After fixing the injector issue I still have the same problem. Attached is a log file of me driving the car around from a cold start. The second log file is driving it around after 20 minutes of letting the engine heat soak.

There is a huge difference. EQ error is almost 20% lean across the board. Using the special functions I turned on and off closed loop control and even when closed loop kicks in and brings the EQ closer to 1.00 it still feels like its only running on a few cylinders.

Take a look at let me know what you think.

[kingtal0n]

I can't open your TUNE file. Something about remote tuning network , no access, not sure what it means but I would gladly look at your file if you can figure out why its locked to me

However I spent some time looking through your logs. COLD vs HOT. I notice COLD starts off with a warm engine, very good, you did the right thing there so we can compare warmed up engine from cold start vs warmed up after a heat soak start.

Coolant temp, air temp, idle pulse, cruise pulse/duty, load (iacv command vs idle a/f ratio), control voltage, MAP, all look extremely close in both logs. Meaning whatever is causing this issue has a large percentage chance of not being related to the tune file at all. But its tough to rule out too much of the file until I looked through it obviously.

My initial assessment is the fuel filters may be clogged or something to do with the fuel pump and fuel pressure is dropping. Please confirm fuel pressure between HOT and COLD.
With alcohol fuel the clogged filters is a common issue.

I would compare alternator voltage at the battery between HOT and COLD just to rule that out as a factor.

Its quite a swing from 14's to 17's between the logs on the gasoline afr scale. That is a major deviation. Make sure CTS bias is disabled just as a precaution.
Also the timing really jumps like crazy at idle, I would fix that IMO But it isn't necessary and nothing to do with your issue.

will have to think about it a bit more. Lets see about the fuel pressure first though.

[nnicholas]

I confirmed that the fuel pressure is the same between both tests. At idle it's sitting right about 41 psi. I also opened up the fuel filter and it looks brand new.

I've also added two new tune files. One with the tune lock on and one with it off. The original had the tune lock off so I'm not sure why that isn't working.

Thanks for the help

[kingtal0n]

remote.jpg

https://forum.hptuners.com/showthrea...u-not-part-off

You might be using a beta version of HPtuners? I only have the recent non beta installed on this computer

If the fuel pressure is the same and injector pulse is the same you would think the a/f ratio would be same at the same load and engine rpm. I'll think about it some more but is there some chance the O2 sensor could be wrong? Is it old? mounted too parallel to the ground per chance? Do you have narrowbands and if so does the a/f of the narrowband correlate with the wideband at lamba=1 (in hot AND cold)?

[nnicholas]

Okay I'll download the latest version and try again.

I was thinking if it could be an o2 sensor and I think it's unlikely because when I command closed loop my wideband also agrees that is running stoichiometric. If they narrowband or the wideband was reading incorrectly I would think they wouldn't agree. Definitely not ruling it out though.

After thinking about it I'm wondering if the turbo is locking up after coming to rest and heat soaking. Maybe something to do with thermal expansion. The added restriction on the cold side would decrease the intake gas density and cause it to run poorly or the increased back pressure on the hot side would mess with the airflow. It also explains why it feels like it's running on a few cylinders and won't build boost. I'll check it today.

[nnicholas]

Or there is a vacuum leak somewhere that only occurs when the engine is heat soaked (also possibly from a thermal expansion difference that doesn't occur until certain parts of the motor have heated or cooled). Maybe an intake manifold gasket or PCV issue?

I was also able to turn off the tuner lock. This file should work now. Let me know if you still have trouble.

[kingtal0n]

Or there is a vacuum leak somewhere that only occurs when the engine is heat soaked (also possibly from a thermal expansion difference that doesn't occur until certain parts of the motor have heated or cooled). Maybe an intake manifold gasket or PCV issue?

I was also able to turn off the tuner lock. This file should work now. Let me know if you still have trouble.

Ah- the pressure test! And PCV discussion.

yes please Pressure Test your hot engine, I make a video to help people do this critical test
https://www.youtube.com/watch?v=w1sb5Y1J068

Make sure PCV valve does not leak. Check hoses for leaks etc...

During boost do you notice a similar offset in the a/f ratio? A vacuum leak causing lean condition will disappear in boost. But then again the fueling magnitude between part throttle and boost is so great that I would consider the difference to be negligible anyways, any significant disparity would lead me to suspect a bad oxygen sensor.

Please use OEM pcv for the engine, you can test the crankcase pressure drop like I did here using a cheap $5 multi meter and a used 1-bar map
https://www.youtube.com/watch?v=0oRbfNPnHaI&t=1476s

[kingtal0n]

Looking through the file here are a couple notes

EGR - zero the table if you don't have EGR? Not sure what affect this table has but it is suspect
Closed loop - disable this completely until you sort this issues out, all of it, STFT, LTFT, make sure you disable all of them
Injector phase - Try much higher numbers than 5.85 in 'normal' table, perhaps 6.85 around that
Short pulse limit, min pulse, default pulse - I would zero these out until you sorted this issue out
Afterstart initial soaktime mult - set to 1.00 as precaution for now
Fuel to wall impact factor- how did you get these values? If you don't know then try stock values from other stock files , same goes for fuel boiling and impact factor this could be the issue or part of it
Trans pressure for low part throttle shifting seems low, did you dial these with a trans pressure gauge?
minimum TCC apply PWM is very low, the converter must be slipping? Doesn't seem right to me, generally I raise the minimum 80% 90% ranges
Spark overspeed and underspeed tables need smoothing desperately, unless you like the low idle quality
Idle spark advance in drive I would smooth these for a better transition to the main map and also smooth the region where it idles holding 0mph.



Thought- how hot do the injectors get where they are? How close is the turbo stuff to the injectors and fuel rail? Could the injectors be getting overheated? This will cause a slow response and possible low voltage at the injector.

Another thing you can try is straight 93 octane gasoline compare with flex fuel to see if the behavior is related to flex algorithms
Try misting the injectors for the hot start with 100% distilled water. Distilled water does not conduct electricity is safe to use on electrical sealed components like sealed weatherpack injector wires.
You can also try pointing an IR thermometer on the injectors after cold shutdown and vs hot startup to find the temp differences many parts.

I like the vacuum leak idea though. Could be simple as a pcv valve that leaks when it gets hot or something like a intake port leak, you can find during pressure test, pop the pcv valve, hope you are not using a chevrolet OEM pcv valve lol they can leak like crazy in boost causing oil leaks and so forth. I use toyota supra PCV valve twin turbo.

[nnicholas]

Thanks for the feedback. I'll run through the tune and make the changes you suggested.

I'll also try out some of these ideas and report back with what I find.

[nnicholas]

Where do you see that the trans pressure for low part throttle shifting is low? Do you have a file I could reference? The force motor looks to be about stock.

[kingtal0n]

Its impossible to say whether its low or high in reality without a trans pressure gauge. But compared to other files with 4l80e in turbo applications including mine it seems numerically low. It depends how the trans is built, what PR valve is in, what mods are done. For example lube to line will drop baseline pressure and require a force motor increase (smaller numbers) depending on the orifice drilled in the pump.

I'm just mentioning it because if you notice the trans feels like its slipping and haven't used a pressure gauge beware it wont take much slipping during a shift to create a huge mess of clutch materials in the pan.

My file 4l80e gen3
https://forum.hptuners.com/showthrea...l=1#post686802
but my transmission has some internal mods which influence pressure, for example
-Lube to line orifice
-direct drum bleed orifice
-enlarged fluid feeds near leaky shaft unions
-reverse feed blocked off for dual fed mod
-seal left on in direct drum

[nnicholas]

I did an HD2 kit with most of the parts. I pretty much copied Matt Happels trans settings since he did the same kit to one of his 4L80Es.

I found a massive vacuum leak today. My old (not currently used) IAT sensor, that's in the place of the EVAP solenoid, was leaking through the pins. I'm going to see if tomorrow, when the motor is cold, if the leak seals itself. Fingers crossed this is the issue.

I also measured the injector resistance when it was hot and acting up. They all pretty much ready 13.8 ohms +/- a couple points. One however measured 11.8 ohms. That seems like a pretty big difference (15%). I do have a spare 9th injector I can swap in to see if that solves the issue. If the injector was shorted internally when hot, I wonder if it would take out the whole bank.

[kingtal0n]

Thats really good find on the leak and injector. I always test injectors from the ECU plug so you can get the entire wire through the injector into the test. I hope that makes sense. Some injector clips are wonky and make variable contact which is why I started doing this way.

Make sure you still pressure test, it is absolutely critical in forced induction turbo applications. If there is a pressure leak (boost leak) it will rise EGT and EGP and cause high temperature to ruin the pistons and overspeed the turbo. Engine damage from boost leaks is real. Don't overlook that and the potential leaking PCV valve, if pcv valve leaks it will pressurize crankcase, cause oil leaking, engine seal damage, and unseat piston rings early at the end of power stroke causing ring flutter and oil inclusion to rings eventual cylinder wear and failure.

[nnicholas]

I did a pressure test today and found bunch of leaks. The intake manifold gasket seemed to be the worst offender. I sealed everything up and it drives better.

The car still goes off the scale lean when driving in the hot condition however. My latest theory is that the coils which are extremely close to the exhaust are overheating. I'm trying to rationalize how an erratic or nonexistent spark on one bank would make the AFR report lean.

It would make sense though because there is a decent amount of air flowing through the engine bay while driving but as soon at the car sits and heat soaks, all the heat goes directly into the coils. Then the car doesn't run well until they have had a chance to cool completely.

Attached is picture of the engine bay and the proximity to the turbo downpipe.