Toyota Repository

[SlowNStock]

I think that I've made some improvements to this table in terms of smoothness. This would bring us to the matter of Fuel Trims though. That table was a result of trying to get LTFTs that were near zero. Is that not the goal that it seems to be made out to be? I've never seen fuel trims from a stock engine, so I don't know what kind of percentage swings are typical. How much correction is normally left to the ECU? I know that it has limitations. I understood that the intent was to minimize that swing.

We do want to minimize fuel trims, but we don't want to achieve that by creating air or fuel models that aren't representative of the underlying, physical system. In this case, the airflow model isn't changed from baseline (TRD/Magnuson), but the fuel model is (due to injectors), so our global corrections should take place on the fuel side.

The particular problem with simply chasing the lowest fuel trims possible on a Tundra is that we don't have the full transient models defined; the transient model has both airflow and fuel components. You can see this transient airflow model kicking in if you drop the polling interval low and log MAF against Absolute Load - on large transients you'll see logged Absolute Load deviate from Absolute Load as calculated via MAF and Engine RPM, which is due to the transient airflow model. The fuel model is the bigger issue, because different injectors have different transient characterization; any change in spray pattern/density affects wall wetting (fuel deposited onto intake surfaces), both impact factor (how much fuel hits intake surfaces) and tau (how long it takes that fuel to evaporate). Unless we are doing all of our tuning in steady state or filtering out all transients, we can wind up baking that fueling correction into our airflow model.

The PCM has a pretty wide range of corrections, but for our purposes, anything +/-5% or closer is generally acceptable, and +/-2-3% is about as good as it gets in a practical sense.

In this file, I've reset the MAF back to the TRD baseline, and carried your average correction (~+12.5%) into the injector constant. Keep in mind that equates to decreasing the injector constant, which tells the PCM that the injector is larger, reducing the injected fuel accordingly.

Wynnded 2011 Tundra SC'd ECT IAT LTFT Testing2 EDITS.hpt

Per the following quote, I also threw in a healthy amount extra into the IAT airflow table - your hardest start in those was with an IAT of ~130F and ECT of ~150F, so some extra airflow at high IAT should help. You can experiment with pulling the ECT airflow back down based on the extra IAT airflow.

This morning it seems that I was motivated enough to swap around some fuel injectors. 1 <---> 3, and 5 <---> 7.

I drove it around some with this tune: Attachment 133141
And the first log: Attachment 133139
The occasional misfires that were seen at Cyl 7 now appear to be at Cyl 5, and a single miss at Cyl 6. What is considered acceptable and what warrants replacing the injector? It has occurred to me that considering that there weren't misses at other cylinders, that that would be enough to warrant replacement, but I'm just speculating.

The vehicle sat for 90 minutes, and here's that log on the same tune: Attachment 133140
This one was still a really sooty start, but much smoother than in the past. I attribute this to no misses immediately after starting. There are again occasional misses at Cyl 5. Subsequent restarts on this tune were also quite sooty.

I changed this tune by bumping up the Startup Airflow ECT considerably: Attachment 133138
And the log: Attachment 133142
Here, as has been mentioned, there are post writing misses upon start. Then there are the now familiar spotty misses at Cyl 5. The restarts were much less sooty.

Then there's a log after just sitting for a few minutes, no tune changes: Attachment 133143
Note that there are misses at other cylinders, but none at Cyl 5. This observation is the basis for my "acceptable amount" question.

An occasional misfire, especially on different cylinders and particular conditions (startup, decel, etc.), aren't typically much cause for concern, there's no cylinder other jumps out as problematic in those logs, other than cylinder 5, as you're aware.

Typically the concern comes in when we start to see regular, repeated misfires on a specific cylinder, particularly in conjunction with other drivability issues (significant Bank 1 vs Bank 2 fuel trim mismatch, noticeable impact to output, and so on).

[Wynnded]

We do want to minimize fuel trims, but we don't want to achieve that by creating air or fuel models that aren't representative of the underlying, physical system. In this case, the airflow model isn't changed from baseline (TRD/Magnuson), but the fuel model is (due to injectors), so our global corrections should take place on the fuel side.

The particular problem with simply chasing the lowest fuel trims possible on a Tundra is that we don't have the full transient models defined; the transient model has both airflow and fuel components. You can see this transient airflow model kicking in if you drop the polling interval low and log MAF against Absolute Load - on large transients you'll see logged Absolute Load deviate from Absolute Load as calculated via MAF and Engine RPM, which is due to the transient airflow model. The fuel model is the bigger issue, because different injectors have different transient characterization; any change in spray pattern/density affects wall wetting (fuel deposited onto intake surfaces), both impact factor (how much fuel hits intake surfaces) and tau (how long it takes that fuel to evaporate). Unless we are doing all of our tuning in steady state or filtering out all transients, we can wind up baking that fueling correction into our airflow model.

The PCM has a pretty wide range of corrections, but for our purposes, anything +/-5% or closer is generally acceptable, and +/-2-3% is about as good as it gets in a practical sense.

In this file, I've reset the MAF back to the TRD baseline, and carried your average correction (~+12.5%) into the injector constant. Keep in mind that equates to decreasing the injector constant, which tells the PCM that the injector is larger, reducing the injected fuel accordingly.

Yes, this makes sense. Thank you for the clarification.

Per the following quote, I also threw in a healthy amount extra into the IAT airflow table - your hardest start in those was with an IAT of ~130F and ECT of ~150F, so some extra airflow at high IAT should help. You can experiment with pulling the ECT airflow back down based on the extra IAT airflow.

I'll experiment with this today. I had avoided bumping up the IAT table as initially I had though that it was getting enough airflow based on the high ECT airflow. Now I've realized that after the engine has been run up to full operating temp then shut off, with the engine compartment heat soaking and the IAT rising, the hotter air is less dense and the necessary flow needs to increase as well. Unless of course I'm thinking about that incorrectly.

An occasional misfire, especially on different cylinders and particular conditions (startup, decel, etc.), aren't typically much cause for concern, there's no cylinder other jumps out as problematic in those logs, other than cylinder 5, as you're aware.

Typically the concern comes in when we start to see regular, repeated misfires on a specific cylinder, particularly in conjunction with other drivability issues (significant Bank 1 vs Bank 2 fuel trim mismatch, noticeable impact to output, and so on).

I have noticed a mismatch between the fuel trims, I don't know what qualifies as significant though. I have ordered an injector to replace the one in cylinder 5.

Thanks again for the lucid explanations, taking the time to look at my tunes and all the logs, as well as answering all my questions and for coaching me through my tuning dilemma!

[Wynnded]

In this file, I've reset the MAF back to the TRD baseline, and carried your average correction (~+12.5%) into the injector constant. Keep in mind that equates to decreasing the injector constant, which tells the PCM that the injector is larger, reducing the injected fuel accordingly.

With the IC that low and the TRD MAF reset, the ECU can't compensate with fuel trims to reach stoichiometric. If memory serves, I've toyed with this a little and found that when gradually increasing the IC from about the same starting point, it needed to be very close to the injector suppliers recommended value to attain ~14.7AFR. This is where us uncouth tuners would then wreak havoc on the transfer table. The lowered IC also appeared to exacerbate misfires at Cyl 5.

Log:23-06-12 09-59-43.hpl

It's probably prudent to pause here until I can replace that injector. If it's leaking at all, even if it wasn't detected when I had the system pressure tested, it's just to paramount to ignore.

I'm still very intrigued as to how to correctly lower fuel trims without making the transfer table graph look like a scribble. When I've looked at my LTFTs using an untouched transfer table, they've invariably been particularly lean on the low end, and just about equally rich in the upper end.

If you're interested, here's a log of a warm start using the original scaled IC, my original crude transfer table edits, and your Startup Airflow ECT/IAT edits. It might be irrelevant with that suspect Inj 5.

[SlowNStock]

With the IC that low and the TRD MAF reset, the ECU can't compensate with fuel trims to reach stoichiometric. If memory serves, I've toyed with this a little and found that when gradually increasing the IC from about the same starting point, it needed to be very close to the injector suppliers recommended value to attain ~14.7AFR. This is where us uncouth tuners would then wreak havoc on the transfer table. The lowered IC also appeared to exacerbate misfires at Cyl 5.

Log:23-06-12 09-59-43.hpl

It's probably prudent to pause here until I can replace that injector. If it's leaking at all, even if it wasn't detected when I had the system pressure tested, it's just to paramount to ignore.

I'm still very intrigued as to how to correctly lower fuel trims without making the transfer table graph look like a scribble. When I've looked at my LTFTs using an untouched transfer table, they've invariably been particularly lean on the low end, and just about equally rich in the upper end.

If you're interested, here's a log of a warm start using the original scaled IC, my original crude transfer table edits, and your Startup Airflow ECT/IAT edits. It might be irrelevant with that suspect Inj 5.

I had the modified and baseline files flipped on my end when I was comparing, so the injector constant should have increased by 12.5% (to 0.078947) since I was carrying a -12.5% correction over, not vice versa - mea culpa for not taking a second look before posting.

Depending on the actual trims you're seeing low vs high, they may be fine as-is.

[Wynnded]

I had the modified and baseline files flipped on my end when I was comparing, so the injector constant should have increased by 12.5% (to 0.078947) since I was carrying a -12.5% correction over, not vice versa - mea culpa for not taking a second look before posting.

Depending on the actual trims you're seeing low vs high, they may be fine as-is.

Ahh, I think we've all done that. I know I surely have. No worries! See, I neglected to include the warm start log that I mentioned:

23-06-12 15-33-22.hpl

[Wynnded]

I received the replacement injector this morning and changed it out. That has eliminated the misfires at Cyl 5, outside of the few that also occur at other cylinders at startup but do not reoccur afterward. I've adjusted the IC a little bit to try to balance the lowRPM leaner with the higherRPM richer LTFTs maintaining the TRD transfer table. The adding fuel to subtracting fuel transition happens at about 2k RPMs which is the engine speed at 75mph, perfect for long Interstate drives.

I have another obligation, but when that's complete I'm going to check the fuel pressure again as the start issue still happens after the engine has been warmed up and then sits for ~15+ minutes. I might be able to fiddle around a little tomorrow, but that would be it for about 10 days.

[Wynnded]

I don't know if this tells us anything useful, but here we go.

Beginning from an overnight sit and cooldown, the truck will always start eagerly on the first turn of the key.

Today I started it and let it warm up until the ECT read 86F. Then I shut it off, waited for less than 30 seconds and restarted it. Fine.
23-06-17 12-44-45.hpl

I shut if off again then let it sit for 15 minutes. The restart hesitated slightly but was passable. Then I drove it a short distance until the ECT read 158F. I shut it off, then restarted it. Fine.
23-06-17 13-10-23.hpl

Next I let it sit for 15 minutes and restarted it. This required two attempts before it started. Then I drove it to bring it up to full operating temperature.
23-06-17 13-40-12.hpl

I shut it off and waited for about one minute and restarted it. Fine. Drove back home.
23-06-17 13-56-09.hpl

Upon arriving I shut it off and let it sit for 15 minutes. It just started, but just barely before the Auto Cranking timed out (I think I read that it cranks for 10 seconds).
23-06-17 14-12-43.hpl

If nothing else, this just seems to indicate that the issue isn't simply a general starting issue, but a starting issue more specific to a warmed engine restarting after sitting for 15 minutes or more. How long it needs to sit until a normal restart occurs will likely take much longer to try to determine.

This may sound outlandish, but it's a question that's evolved in my head. What if there are one or more injectors that slowly leak for an undetermined amount of time after having been active, but not so much as to create an immediate restarting issue? Could they slowly leak enough that after 15 minutes there's enough fuel to cause a starting problem? After another undetermined amount of time could they stop leaking allowing the fuel in the cylinders to evaporate enough returning to a more ideal starting mixture "the next day"? This would suggest that the injectors are more likely the issue than the tune, but then I remember that there are others who are experiencing this and they're using injectors from a different manufacturer and of a different flow rating.

Update:
Five hours later it took three (3) tries before the truck started.
23-06-17 19-22-39.hpl

[Wynnded]

I finally got fed up and changed back to the TRD injectors. Starting is perfect for all starting conditions now. With the TRD injectors, I experimented with some of the tables that I thought might affect starting. For example, for the Startup Airflow ECT table, I filled it with a value of 12g/s and it still started fine. I changed the Offset table from stock values to the ones suggested for the 750cc injectors. Staring was unaffected. I'm going to contact the seller and see what might be done.

[Wynnded]

I called the place that I bought the injectors from. Their take on it being an injector issue was that it would be unlikely on a grand scale but they'd be willing to take my money if I insisted on sending them in for testing.

Soooo....
I reinstalled the 750s again this morning to have another go at things. I used the same tune as with the TRD injectors changing only the IC. Then I went back to what @SlowNStock said about increasing Startup Airflow values in increments. As I went up, I got less and less sooty emissions from the tailpipe on immediate hot starts all the way up to the limits of the table, 20L/s. However, with each progressive change, starting remained problematic the longer it sat, starting after about 15 minutes.

I made logs of both iterations of the tune in an effort to try to spy any differences in fuel and airflow during cranking, but nothing that seemed exceptional. A couple milliseconds difference either way with the IPW, and similar variations of mass airflow. As always, I'm open to suggestions.

[Wynnded]

Up till now, the first start of the day has been in the morning when the ambient temperature was in the 50s and the truck started fine. Today I waited until midafternoon and the temp was in the low 80s and the truck started as though it had been run earlier in the day and cranked through the first attempt and started on the second. Further starts seconds later were successful. One hour later the truck cranked through the first attempt and started on the second, further starts were successful. An additional hour later yielded the same results. Here are the logs for each hour. In the tune the Airflow ECT values appear extraordinarily high, but I can literally use any value within the permitted table range and get the same starting behavior. Next I will see what happens at 60 or 70F after sitting overnight.

23-07-10 16-17-19.hpl
23-07-10 17-04-05.hpl
23-07-10 18-07-37.hpl
Wynnded V4.hpt

*In the logs there is an anomaly between some restarts where the MAF rises, the ECT, and/or IAT drop. I suspect that this is the SAIS bypass module and will remove it again. It was reinstalled while running the TRD injectors to see if it had any influence but it didn't.

[MetalMann]

I?ve just purchased a tundra! Yesterday! I?ll be making some adjustments very soon. Tuned 3 NA 5.7?s pretty well and also have an Avalon? which I?m reading and playing with through ktag. I?ll be wanting to learn, just as you on how these two can be read and compared, HPtuners and .bin 😉 then possibly learn Avalon 2GR-FE things from there. Is there anything I can send you to help you on this? My plan was to makrr we a change through HP, then read in .bin to see the changes.

[Wynnded]

@SlowNStock
So, with all the back and forth and your very generous explanations and patience, it appears that there is something faulty with my injectors. @Snivilous was kind enough to loan me a new set of 720cc injectors and all starts since then have been flawless, even using the stock Startup Airflow table. Tomorrow I will be reaching out to the company that I purchased the 750cc injectors from to see what they can do for me.

[Wynnded]

Still experiencing perfect starts. I finally got the problematic injectors in the mail to be tested. Now we wait to hear back.

In the interim, I noticed a different anomaly and am curious to see if anyone else has encountered the same. When decelerating where the gear is high and the RPMs are low, the ECU oscillates between OL/CL and RPMs hunt. This happens at the bottom of an exit ramp, or when decelerating off throttle approaching a traffic light. I read somewhere that someone else was experiencing something similar and they though that maybe the ECU was anticipating a downshift, but I'm unable to locate their posts on the matter. Here is a log (start at the 2 minute mark) and my current tune.

[SlowNStock]

Still experiencing perfect starts. I finally got the problematic injectors in the mail to be tested. Now we wait to hear back.

In the interim, I noticed a different anomaly and am curious to see if anyone else has encountered the same. When decelerating where the gear is high and the RPMs are low, the ECU oscillates between OL/CL and RPMs hunt. This happens at the bottom of an exit ramp, or when decelerating off throttle approaching a traffic light. I read somewhere that someone else was experiencing something similar and they though that maybe the ECU was anticipating a downshift, but I'm unable to locate their posts on the matter. Here is a log (start at the 2 minute mark) and my current tune.

Glad to hear you were able to get startup fixed via injectors!

What you're seeing is the PCM entering flex lockup to go into DFCO - as it locks up, it slows the truck just enough to drop out of flex lockup, then the truck gains enough speed with it unlocked to enter flex lockup again, and the cycle repeats.

Takes a specific set of conditions where the truck can gain enough speed with the converter unlocked but lose enough speed to keep the converter from staying locked, while in a gear with flex lockup enabled (3-5). 5th is generally the most problematic, simply because of how it falls with regard to engine braking and vehicle speed.

There are a couple options to remedy:

1) Force a downshift sooner (5-4 in this case).
2) Increase flex lockup hysteresis (the difference in speeds and/or TPS between flex lock and unlock).
3) Disable flex lockup, either for the gear or that RPM range.

[Wynnded]

Glad to hear you were able to get startup fixed via injectors!

What you're seeing is the PCM entering flex lockup to go into DFCO - as it locks up, it slows the truck just enough to drop out of flex lockup, then the truck gains enough speed with it unlocked to enter flex lockup again, and the cycle repeats.

Takes a specific set of conditions where the truck can gain enough speed with the converter unlocked but lose enough speed to keep the converter from staying locked, while in a gear with flex lockup enabled (3-5). 5th is generally the most problematic, simply because of how it falls with regard to engine braking and vehicle speed.

There are a couple options to remedy:

1) Force a downshift sooner (5-4 in this case).
2) Increase flex lockup hysteresis (the difference in speeds and/or TPS between flex lock and unlock).
3) Disable flex lockup, either for the gear or that RPM range.

I'm glad as well, many thanks for your patience.

I never would've even guessed that flex apply/release would've been where this is happening, having never really read anything about it. I will get into this on the weekend!

[Wynnded]

This hunting appears to happening in 6th gear as well, where Flex-Lock is not implemented. Unless of course I'm not understanding. I'm intrigued that this hunting has started when I've not made changes to any transmission tables in a very long time.

23-09-02 14-09-41.hpl

[Wynnded]

Well, I'm pretty confident that I don't know what it is that I'm doing. But I'm trying.

In an effort to try to force an earlier 5-4 downshift, I went to Transmission > Shift Scheduling > On-Power > Normal/Normal Fault/Cold/Cold Fault/Tow-Haul/Tow-Haul Fault and changed the cells in the 0% and 13% columns to the same values that are in the 25% column just to see if that made a difference. This would be changing the 875 rpm Output Shaft speed to 1600 rpm in all of those tables.

To increase hysteresis, I went to Transmission > Torque Converter > Flex Apply/Release > Output Speed vs. Throttle > D/S6 and to the 5th App and 5th Rel tables. In the Apply 5th Gear table, I tested increasing the 0% column value by 5, 10 and 15%. Not decerning an improvement, I reset the table back to stock. In the Release 5th Gear table I tested lowering the same column the same percentages, but was not able to notice an improvement. I was not able to be sure in what way to change TPS hysteresis, so I didn't try anything.

Trying to disable Flex-Lock for that gear, in Transmission > Torque Converter > Flex Apply/Release > Output Speed vs. Throttle > D/S6 I changed all the values in the Apply and Release 5th Gear tables to the same values that are in the Apply/Release 6th Gear tables as that gear is not suppose to utilize Flex-Lock. Unfortunately there was no change in behavior.

Interestingly, the oscillation appears in 3rd through 6th gears. I'm rather mystified.

[Wynnded]

@SlowNStock
Curious to know if it appears that I tried to implement your suggestions at least somewhat correctly.
Thanks!

Well, I'm pretty confident that I don't know what it is that I'm doing. But I'm trying.

In an effort to try to force an earlier 5-4 downshift, I went to Transmission > Shift Scheduling > On-Power > Normal/Normal Fault/Cold/Cold Fault/Tow-Haul/Tow-Haul Fault and changed the cells in the 0% and 13% columns to the same values that are in the 25% column just to see if that made a difference. This would be changing the 875 rpm Output Shaft speed to 1600 rpm in all of those tables.

To increase hysteresis, I went to Transmission > Torque Converter > Flex Apply/Release > Output Speed vs. Throttle > D/S6 and to the 5th App and 5th Rel tables. In the Apply 5th Gear table, I tested increasing the 0% column value by 5, 10 and 15%. Not decerning an improvement, I reset the table back to stock. In the Release 5th Gear table I tested lowering the same column the same percentages, but was not able to notice an improvement. I was not able to be sure in what way to change TPS hysteresis, so I didn't try anything.

Trying to disable Flex-Lock for that gear, in Transmission > Torque Converter > Flex Apply/Release > Output Speed vs. Throttle > D/S6 I changed all the values in the Apply and Release 5th Gear tables to the same values that are in the Apply/Release 6th Gear tables as that gear is not suppose to utilize Flex-Lock. Unfortunately there was no change in behavior.

Interestingly, the oscillation appears in 3rd through 6th gears. I'm rather mystified.

[Wynnded]

I tested briefly yesterday, this is happening with my reference "stock TRD tune" as downloaded from HPT. I took the intake apart and it's spotless, as is the throttle body. I thought that maybe something may have snuck in. Checked all the vacuum hoses too. Nutty stuff...

[SlowNStock]

@SlowNStock
Curious to know if it appears that I tried to implement your suggestions at least somewhat correctly.
Thanks!

Throw up a copy of this tune, and I'll try to take a look later this week.