Throttle "blip" or flare when I release the throttle while cruising - 6.2 Raptor

[trick76cj5]

I have been working through the SD tables and Torque Models, and while they aren't perfect, they are getting better. But, i've had this issue show up, while driving, when i release the trottle, i get a slight throttle flare before the rpms start to drop. I can hear it very clearly in the exhaust not, it is a very short "blip", but its there. Here is a screen grab that shows it very clearly:

throttle flare.jpg

Here is the full log file (this is my drive to work in the morning) and my tune file if anyone wants to take a look:

315am.hpl
VE_Mod24.hpt

Any direction you can provide to help me resolve this would be greatly appreciated!

[Pistol_91]

I believe this is your issue. I have encountered similar scenarios before. Your cam angles change dramatically before going to optimal stability back to 0. Your distance table is putting the car through the ringer with all the mapped points crushed all into one area where normal cruising and driving would be happening.
log.png
tune.png

[trick76cj5]

Thanks for taking the time to review my log and tune file!

Those tables, in my tune, are stock (other than fixing an error in the OE table at a MP that i don't use . . . not even sure it was necessary to change, but i did).

VCT_Mine.jpg

vct stock.jpg

I'm not disagreeing with your assessment, I'm just not sure what I would adjust to help fix this. I wouldn't imagine the mods i've done would have in impact with how the distance maps and VCT timing would operate. I have soooo much more to learn about how these things work!

Thanks!

[Pistol_91]

You only have MP 0-4 enabled so making changes to 10 wouldn't matter. Are all your distance tables stock? Including optimal stability? What changes did you make and when did you notice the issue pop up?

[trick76cj5]

The distance tables are all stock including Optimal Stability.

I noticed this when i changed my shift strategy to hold gears a little longer and not bog down in town. In increased RPM has highlighted this behavior. I can make it do it on the interstate, but its much less noticable and I have to be looking for it.

So, i'm not exactly sure when this started happening.

I most recently have been working through the Speed Density and Torque Modeling settings. The ETC Torque and Brake Torque match pretty well now, which i was hoping was maybe an issue. The Speed Density is pretty good, but I'm still working to make it better, but that doesn't' seem to have made any improvements either.

It looks like the Throttle is being held open a short time after the throttle pedal goes to 0, this makes me think its related to torque demand, but I can't figure out what the source would be.

[engineermike]

The issue occurs when it transitions from MP4 to MP0, which it must do to get from cruise to idle cam timing.

First of all, why are you changing the SD model? It's pretty challenging to get right and not super-critical to get perfect on a MAF setup. The main purpose the SD model exists in a MAF setup is to determine manifold pressure because it needs manifold pressure to know how far to open the throttle blade since throttle blade opening is a function of throttle body dP. The SD model will not affect fuel trims in a MAF setup. Unless you're logging actual MAP and comparing to calculated MAP, you won't be able to fix the SD model.

That said, I believe this particular issue is caused by your SD model. Some of your channels log way too slow, but I did the best I could to determine what's happening.

As you're letting off the throttle, the scheduled torque request is between 129 and 60 ftlb. This channel needs to be sped up, btw. Lets assume it's a constant 100 ftlb at 2000 rpm to get an idea of what's going on. So for a 100 ftlb scheduled torque, the load required will be between .22 and .19 in MP4 and MP0 respectively. Let's call it .2 load. Everything looks good so far.

Now, according to your SD model as it is, the manifold pressure at .2 load would be:

In MP4 at .2 load, the MAP is calculated to be 3.2 psia.
In MP3 at .2 load, the MAP is calculated to be 4.0 psia.
In MP2 at .2 load, the MAP is calculated to be 4.3 psia.
In MP1 at .2 load, the MAP is calculated to be 4.3 psia.
In MP0 at .2 load, the MAP is calculated to be 3.2 psia.

So in order to get achieve desired airflow in MP2 and 1, it has to open the throttle more because it thinks the vacuum isn't as deep and the dP is less across the throttle. So as it passes through MP3-2-1 it opens the throttle more. Now, there is a feedback loop on the airflow but it appears to be disabled in your cal (engine\airflow\electronic throttle\feedback control) but even if it wasn't, the feedback loop wouldn't work fast enough to hide this problem.

Conclusion: The SD model isn't right and I'm not sure why and how you're tuning it anyway.

Side note is that if the truck were a SD system with a MAP sensor, it would be much more important and easier to tune since you use the SD model to fix fuel trims. But even this is very difficult and time consuming.

Edit: Since your ETC feedback loop is disabled, then the SD model would actually be used to calibrate desired vs actual MAF assuming your TB model is correct. Is that what you're fixing? If so then then hats off to ya, but your desired Airmass channel is logging way too slow to do that.

[trick76cj5]

Engineermike, thanks for the response. I'm pretty sure I need to re-read this about 100 more times before I can digest everything! Thanks for all this information.

Also, I am still very much learning both HP tuners as well as how Ford applies their logic, I tuned the ford diesels up to 2008 and then stopped, so I'm about a decade behind the curve and trying really hard to catch up.

First off, I didn't realize my ETC Feedback loop had been disabled, is this the "Tip In Mgmt" switch? I don't have a "feedback control" listed as an option.

Second, I was following a few of murfie's posts talking about how the SD is good to have set right as it covers the vehicle performance in the transients. Using his approach,

• I wrote math that would use the "calculated" MAP and air intake temp to generate an air density, i could then calculate the amount of air consumed per intake event.
• Using the MAF sensor with a known flow, i could calculate the amount of air actually consumed per intake event.
• These 2 numbers are the final result of the volumetric efficiency calculation, so the error between these 2 numbers should represent the error in my Speed Density Chart since the speed density calculator appears to be a "standard" VE table with MAP on the Y and RPM on the X and the VE listed at their intersection.
• I would take that error and multiply the matching sell in the SD table by that percent error to bring them into alignment.

Maybe I completely butchered his approach, maybe my refresh rate is too slow which you pointed out, or maybe this actually has 0 impact on how the truck runs and I'm only creating problems for myself . I will revert back to the stock SD values and see if that helps clear things up.

I was reading the desired airmass, as it is the the inverse in the Torque Model tables, and then using the spreadsheet murfie provide the values for the torque tables. But now i think I'm seeing that desired airmass is not reading fast enough which is why i probably have some errors in this modeling too. I'll see if my data changes significantly with an updated refresh rate.

Thanks for all the help.

I will do some more research on the speed density tables to understand how i make the transition from one table to the other better.

I will turn my Tip In Mgmt back on which may help.

I will update the refresh rate of my Desired Air PID

If anyone has any additional thoughts or can point me to some posts that may help I am here to learn.

[engineermike]

Regarding the feedback loop, if you look at engine\airflow\electronic throttle\feedback control you'll see Proportional, Integral, and Derivative gains that are all set to 0. This is why the airflow feedback loop is disabled.

Regarding your math, the problem is that the SD calculator output table is load not VE. VE isn't actually used on these as far as I know. The only way to really fix the SD model is by comparing measured map to calculated map but there is no measured map so you have to get that data some other way. It's a science project for sure.

[Pistol_91]

SD is load like mike said. I have tried configuring SD 100 different ways and the easiest and best way I found was to use the populated loads and smooth them out. Look at a focus or EcoBoost's 3D SD mapped points in the calculator on their tune and you will get a good idea. Here are how my tables look on my car and it drives great, no hiccups.
sd.png

[SiriusC1024]

What mods have been done to warrant adjustments to the SD tables? Cam angles haven't been changed, either.

Inferred MAP is used to predict airmass for throttle transitions. If the target is wrong the results will be wrong. Those SD tables aren't correct.

Knocking a lot, too. I see you adjusted the borderline tables quite a bit.

[trick76cj5]

Regarding the feedback loop, if you look at engine\airflow\electronic throttle\feedback control you'll see Proportional, Integral, and Derivative gains that are all set to 0. This is why the airflow feedback loop is disabled.

Regarding your math, the problem is that the SD calculator output table is load not VE. VE isn't actually used on these as far as I know. The only way to really fix the SD model is by comparing measured map to calculated map but there is no measured map so you have to get that data some other way. It's a science project for sure.

Thanks for the clarification, these were all set to 0 on my stock tune.

What mods have been done to warrant adjustments to the SD tables? Cam angles haven't been changed, either.

Inferred MAP is used to predict airmass for throttle transitions. If the target is wrong the results will be wrong. Those SD tables aren't correct.

Knocking a lot, too. I see you adjusted the borderline tables quite a bit.

First off Sirius, i wanted to thank you as I am using the tune you provided some time ago on another post for my baseline. Sorry i've screwed it up a little though

Its a used motor i bought, but its been bored 0.020 over with 11:1 pistons, so that should change the thermal efficiency of the motor. I've also added headers which should impact Volumetric Efficiency, maybe its not enough to worry about though?

I think what you are seeing as knocking is actually negative knock retard, or timing being added back in because the borderline knock table has been modified to be conservative, especially with my Compression Ratio. Sorry, my graphs are all over the place as I'm still learning how to set the graphs up well, what to monitor, and clearly that some of my math is not good.

I will bring the SD back to stock and test run it tomorrow, i drove it home from work with the Throttle Tip In Mgmt turned on and it was terrible lol.

Thanks again everyone for the help! Its so nice to have a resource in figuring out these newer computers.

[murfie]

Engineermike, thanks for the response. I'm pretty sure I need to re-read this about 100 more times before I can digest everything! Thanks for all this information.

Also, I am still very much learning both HP tuners as well as how Ford applies their logic, I tuned the ford diesels up to 2008 and then stopped, so I'm about a decade behind the curve and trying really hard to catch up.

First off, I didn't realize my ETC Feedback loop had been disabled, is this the "Tip In Mgmt" switch? I don't have a "feedback control" listed as an option.

Second, I was following a few of murfie's posts talking about how the SD is good to have set right as it covers the vehicle performance in the transients. Using his approach,

• I wrote math that would use the "calculated" MAP and air intake temp to generate an air density, i could then calculate the amount of air consumed per intake event.
• Using the MAF sensor with a known flow, i could calculate the amount of air actually consumed per intake event.
• These 2 numbers are the final result of the volumetric efficiency calculation, so the error between these 2 numbers should represent the error in my Speed Density Chart since the speed density calculator appears to be a "standard" VE table with MAP on the Y and RPM on the X and the VE listed at their intersection.
• I would take that error and multiply the matching sell in the SD table by that percent error to bring them into alignment.

Maybe I completely butchered his approach, maybe my refresh rate is too slow which you pointed out, or maybe this actually has 0 impact on how the truck runs and I'm only creating problems for myself . I will revert back to the stock SD values and see if that helps clear things up.

I was reading the desired airmass, as it is the the inverse in the Torque Model tables, and then using the spreadsheet murfie provide the values for the torque tables. But now i think I'm seeing that desired airmass is not reading fast enough which is why i probably have some errors in this modeling too. I'll see if my data changes significantly with an updated refresh rate.

Thanks for all the help.

I will do some more research on the speed density tables to understand how i make the transition from one table to the other better.

I will turn my Tip In Mgmt back on which may help.

I will update the refresh rate of my Desired Air PID

If anyone has any additional thoughts or can point me to some posts that may help I am here to learn.

I have tried to explain the VE part of the SD as some only understand it that way. I dont personally do it like that unless you are just attempting modification with out a MAP sensor and are doing some explorartion and want a independant rationality check target.
I plot the MAP sensor on Y and Lbm /cylinder event on X(from airload, MAF is limited to 65535 g/sec integer limit) 220 in the image, and regress a quadratic curve, 204 in the image, through the plotted points to get the coeffecients out of that equation at a given MP and RPM just like in the Ford patent. Commanding fixed cam angles makes it easier.
This picture shows it, but the plotted points dont really fall too much out side the line, in reality its more scattered, and not always single curve looking. You can also plot as many or as few points as you like so polling rate isnt an issue. Regression goes over peoples heads most of the time I try to explain it but excel makes it very simple.

In this image above line 202 is less than 100% VE, below it is more than 100% VE, so people get confused when looking at these plots. Its how Ford tracks Blowthrough and relates to the Exh MAP in the equation. If its not tracking blowthrough it stays on line 202, and doesnt switch to line 210, calculating MAP or viseversa. Which is the same as saying the VE is 100% and the air density in the cylinder is equal to the air density in the manifold. Which in most boosted application is true, with the occasional exception of extremely inefficent FI setups.

US20130111900A1-20130509-D00002.png

[SiriusC1024]

Thanks for the clarification, these were all set to 0 on my stock tune.




First off Sirius, i wanted to thank you as I am using the tune you provided some time ago on another post for my baseline. Sorry i've screwed it up a little though

Its a used motor i bought, but its been bored 0.020 over with 11:1 pistons, so that should change the thermal efficiency of the motor. I've also added headers which should impact Volumetric Efficiency, maybe its not enough to worry about though?

I think what you are seeing as knocking is actually negative knock retard, or timing being added back in because the borderline knock table has been modified to be conservative, especially with my Compression Ratio. Sorry, my graphs are all over the place as I'm still learning how to set the graphs up well, what to monitor, and clearly that some of my math is not good.

I will bring the SD back to stock and test run it tomorrow, i drove it home from work with the Throttle Tip In Mgmt turned on and it was terrible lol.

Thanks again everyone for the help! Its so nice to have a resource in figuring out these newer computers.

You're welcome.

Put the torque tables back to stock, too. You should only change things as needed.

[trick76cj5]

murfie, thanks for the clarification. I'm going to do some more research on this to better understand Ford's patent and how this all works together, this is an amazingly complex system. Thanks for sharing your understanding of the system.

Sirius, thanks, i reverted those to stock as well.

So, the question is, how do i know if my Speed Density or Torque Model need addressed?

My ETC torque and Brake Tq didn't match, is this a problem?

Again, thanks so much for your help and patience!

[SiriusC1024]

First, do what you did with returning SD and TQ back to stock. Same for the rest (except maybe my modified tune settings which work really well if I may say so). Then do the same for the spark tables. You are most definitely getting knock. Remember what you were taught about double negatives, now

I see you adjusted O2 transport delay. You are correct to do so, but there are procedures that take away the guess work.
https://forum.hptuners.com/showthrea...-Time-Constant

After that you can get MAF right like you've been doing.

Finally if you notice driveability problems, particularly with throttle oscillation and hard shifting, you can start playing with relevant settings.

[SiriusC1024]

So, the question is, how do i know if my Speed Density or Torque Model need addressed?

There's a sticky on the non-eco subforum about SD. Lot of good info.

[engineermike]

Thanks for the clarification, these were all set to 0 on my stock tune.

Yes so the oem didn’t have feedback airflow control enabled. Disabling this is one trick I use to make tuning the tb and sd models possible on newer maf setups.

To go back to what I said earlier, assuming your tb model is correct, the easiest most direct way to tune your sd model is to use it to make desired maf match actual maf. I can explain more if needed.

And as Murfie said, this will be much simpler if you lock into/isolate single mapped points. You need to get load, desired maf, maf, and calc map data for several different loads at constant speed. This is nearly impossible if your mapped point isn’t constant.

[SiriusC1024]

engineermike is technically correct, but I'd say read that sticky. See under What about greater than 100% VE? where it talks about Aircharge Multiplier? Ramming effect from a tuned intake is going to be really similar to vacuum effect of longtubes. Both are dependent on RPM and air charge.
aircharge mult.png

Way easier, and again, only if you have to.

[trick76cj5]

First, do what you did with returning SD and TQ back to stock. Same for the rest (except maybe my modified tune settings which work really well if I may say so). Then do the same for the spark tables. You are most definitely getting knock. Remember what you were taught about double negatives, now

I see you adjusted O2 transport delay. You are correct to do so, but there are procedures that take away the guess work.
https://forum.hptuners.com/showthrea...-Time-Constant

After that you can get MAF right like you've been doing.

Finally if you notice driveability problems, particularly with throttle oscillation and hard shifting, you can start playing with relevant settings.

Just so i'm reading things right, positive knock retard means timing is being pulled due to knock. Negative knock retard means timing is being added because there is lack of knock. Is this correct?

I have read that the 6.2's, even stock, have a lot of knock retard. I'm not a fan of having much, but i tend to get some knock retard at tip in or at gear changes and its usually only a few degrees. Most of the time, at that same rpm and load its adding timing back in. So i've been struggling with how to manage this . . . this is actually what lead me down the path of SD and TM as i was assuming this was an error during transients that was causing the knock and i was hoping to fix that as generally my timing appeared pretty conservative.

That being said, i just did a compare to stock and looked at how much i had changed timing, timing is negative compared to stock in the middle of the rpm/load, but i am adding more timing at higher rpm after all my data logging, so maybe I lost track up there.

Anyway, sorry to ramble. My real questions were 2 fold: 1) did i understand knock retard correctly, and 2) is it not normal to have a little knock retard on these motors

I will go back to your tune and just adjust the transport delay and start over . . . hopefully with a little more clarity now . . .and a clear understanding of what to look for with knock as my compression ratio is pretty high as these motors tend to be pretty knock sensitive.

[engineermike]

I play it very conservatively with knock, probably overly so. But unlike ford even, I try to tune such that I never have a knock event. People get hung up on adding vs subtracting knock timing that generally leads people astray. If knock advance is good then just set borderline low and you’ll always have knock advance. It’ll be a terd but hey, you’re getting timing added so it must be good right?

I set mine up to where knock advance is limited to 2 deg, it rises to 2 and rides there without every having to back off due to knock detection. This can leave 2 or more deg on the table in high humidity but it’s pretty safe.