2019 Bullit with Edelbrock 2650 Trials and Tribulations

[evolmotorsprt]

I've tuned quite a few Mustangs but this is my first coyote PD blower car. I've read the tuning school book, the coyote cookbook, and have scoured the forum and internet for good sources of data. I've spent a lot of time breaking down how the car interprets inputs and the flow of commands through the computer. I've also spent a lot of time in excel building worksheets and looking at other's spreadsheets. I'm just stumped because I feel like I can't achieve positive drivability results. BTW, I would like to thank Murfie for contributing as much to this forum as he does.

The car drives great at light throttle and full throttle, but terrible mid throttle. I find this to be a common issue with most threads about this topic. I've been able to accumulate several gen 3 coyote Whipple and Roush tunes on this forum, and have tried copying the majority of the Whipple file (SD, Spark, Torque Tables, etc.) into the Bullit calibration as a starting point. I then got the MAF in line, made recommended tweaks from the tuning school and coyote cookbook and took the car for a drive.

The throttle tends to overshoot and undershoot, oscillates at times, and just isn't acceptable on any level. To fix this I've tried tuning the predicted throttle angle, the torque tables, and speed density independently as well as combined. My results have either made no difference or have made the car drive worse. I've tried locking mapped points to eliminate variables, smoothing spark, slowing down the reaction of the throttle blade, etc, etc, etc...

So what am I missing? I want to understand what I'm doing incorrectly and how to fix it. I feel like I've been banging my head against the wall for a couple weeks now. I have a load bearing dyno all day Monday so I'd like to have a plan.

I'm considering flashing the entire Roush 2650 strategy to the car as a starting point. Is there really a difference between copying the parameters from the Roush file to the Bullit file, and paying for more credits so I can flash the entire Roush strategy?

I've attached some of the Roush/Whipple tunes I'm referencing along with what I currently have in the car. Any help would be appreciated. Thanks.

[Auto-Hentzschel]

I understand, you are not happy with the result! Bullitt OS is a bitch! It don't like boost. What Edelbrock kit did you installed? do you using the 102mm Throttle body? Don't waste credits for Whipple or Roush OS. It work with no different to the stock OS.

[evolmotorsprt]

I understand, you are not happy with the result! Bullitt OS is a bitch! It don't like boost. What Edelbrock kit did you installed? do you using the 102mm Throttle body? Don't waste credits for Whipple or Roush OS. It work with no different to the stock OS.

Yeah, this went from being a learning experience to an all consuming obsession. I hate not being able to figure things out.

It's a stage 2 Edelbrock 2650 GT kit on a Bullit using the OE Bullit Throttle Body. It also has port FIC injectors recommended by Edelbrock with good data. Stock manifolds and exhaust.

I have the HP Tuners Prokit so I may hook up a MAP, use only 6 mapped points and try to get the SD tables perfect on the dyno. But I'd hate to spend all of that time on it if I'm going in the wrong direction.

[evolmotorsprt]

So I'm sitting here tonight second guessing myself. Could someone please confirm that I have this command flow correct?

Pedal > Desired Torque > Torque Table > Inverse Torque > Satisfy Min/Max Load Tables (Fuel Econ, Optimal Stability, etc.) > Distance Table > Mapped Point

Mapped Point now assigns: Speed Density, Torque, Spark > Then, using the sensor inputs and mapped point data the computer comes up with a solution to generate the desired torque and using the TB model commands a throttle position. > The computer then calculates the torque generated based on airflow and spark to verify the engine is making the desired output as defined in the Torque Tables.

Is this correct?

If so that would mean the error I see in the throttle is a result of the generated torque and theoretical torque not agreeing. So I push the pedal, the computer overshoots a bit trying to make the desired torque quickly, boost hits, computer goes "shit too much power," and freaks out trying to match the torque tables. That would mean the solution would be fixing the torque tables and adjusting the PID controller for the throttle body to ease transitions.

Am I on the right track?

I can also see how the DD table might need to be adjusted for more torque at a given throttle position.

[Auto-Hentzschel]

the ST II Edelbrock kit comes with 103mm Throttle body and own aircleaner box. why you don't use it? The FIC injectors was part of the Kit?

[evolmotorsprt]

the ST II Edelbrock kit comes with 103mm Throttle body and own aircleaner box. why you don't use it? The FIC injectors was part of the Kit?

Edelbrock doesn?t make a Bullit specific kit. This was originally a GT stage 1 kit. The kit had an SCT that was supposed to be used for a base file, but Edelbrock didn?t have a Cal for the bullit. They suggested the owner upgrade to the stage 2 kit and they would give us a base file. It was piecemealed together by Edelbrock and they told us what to buy to make it work. They then reneged and told us that they still couldn?t support a bullit, and that?s where it?s at. So, at this point it has the stage 2 air box, stock throttle body, injectors which are just FIC matched Bosch injectors, BAP, and the stage 1 2650 kit. It should be a rocking setup if I can get it working.

When you tuned the Bullit, what changes ended up moving you in the right direction?

[LastPlace]

A couple pointers, use only known good injector data (sounds like you did), use known good throttle body data and build your torque to load tables using a dyno.

Do not follow the tuning school process of gutting the throttle body model.

When you get that ?surge? from the bypas closing and the supercharger kicking in, that is normal.

it can not be tuned out

Roush and Whipple calibrations have Ford oem bypass logic that handle how the bypass closes to keep that ?surge? from happening and make the transition from no boost to boost less of a light switch feeling.

It can not be tuned out you can only work around it.

some use a pill in the bypass line, some move the vacuum line to a spot right after the throttle body before the rotors.

[Auto-Hentzschel]

Yes, Edelbrock doesn't make a kit for the Bullitt. They recommend to use the tuner kit. How I know, the BULLITT TB fit only with a adapter plate. I recommend you also use the Tuner kit with the 103mm TB and the 3.5" 6-rip Pulley. 12:1 compression don't like more than 8psi boost. That fit very close to the Bullitt engine. BTW the reduction pill is still in the bypass line.
If your RPM surge in part throttle, you can reduce the Throttle feedback ans adjust the torque tables. it is possible to get the Bullit running smooth like a factory car. The Edelbrock St II calibration will not work for this project. only the Throttle body data you can get for the 103mm TB.
BTW you log show a lot of Knock retard.

[evolmotorsprt]

A couple pointers, use only known good injector data (sounds like you did), use known good throttle body data and build your torque to load tables using a dyno.

Do not follow the tuning school process of gutting the throttle body model.

When you get that ?surge? from the bypas closing and the supercharger kicking in, that is normal.

it can not be tuned out

Roush and Whipple calibrations have Ford oem bypass logic that handle how the bypass closes to keep that ?surge? from happening and make the transition from no boost to boost less of a light switch feeling.

It can not be tuned out you can only work around it.

some use a pill in the bypass line, some move the vacuum line to a spot right after the throttle body before the rotors.

Thank you for the response. Noted on the ETC Tuning, and boost kicking in. I will be using the factory data for the Bullit TB. I currently do have the pill in the vacuum line. I've tried it in and out, and in definitely makes the transition to boost better. I may reduce the orifice size. 2 questions regarding your response:

1. What torque value are you using to derive you number in the torque tables from? I've seen so many different answers to this question, i.e. %err from etc into the TT histogram, %err derived from indicated vs ETC torque request, pasting indicated torque directly into TT... If I understand it correctly, the coyote cookbook suggests using the torque reading on the dyno and drivetrain loss to calculate actual torque.

2. In your opinion would it be beneficial to flash the complete Roush strategy?

If your RPM surge in part throttle, you can reduce the Throttle feedback ans adjust the torque tables. it is possible to get the Bullit running smooth like a factory car.
BTW you log show a lot of Knock retard.

Could you please define throttle feedback? Do you mean the pedal, desired torque, or the PID for throttle control?

I was under the impression that +KR was timing being pulled and -KR is timing being added. I do have a small amount of +KR when I'm hitting my cyl. pressure limit. Should I not take the PID "Knock Detected: No" at face value?

[MustangGT]

I've tuned quite a few Mustangs but this is my first coyote PD blower car. I've read the tuning school book, the coyote cookbook, and have scoured the forum and internet for good sources of data. I've spent a lot of time breaking down how the car interprets inputs and the flow of commands through the computer. I've also spent a lot of time in excel building worksheets and looking at other's spreadsheets. I'm just stumped because I feel like I can't achieve positive drivability results. BTW, I would like to thank Murfie for contributing as much to this forum as he does.

The car drives great at light throttle and full throttle, but terrible mid throttle. I find this to be a common issue with most threads about this topic. I've been able to accumulate several gen 3 coyote Whipple and Roush tunes on this forum, and have tried copying the majority of the Whipple file (SD, Spark, Torque Tables, etc.) into the Bullit calibration as a starting point. I then got the MAF in line, made recommended tweaks from the tuning school and coyote cookbook and took the car for a drive.

The throttle tends to overshoot and undershoot, oscillates at times, and just isn't acceptable on any level. To fix this I've tried tuning the predicted throttle angle, the torque tables, and speed density independently as well as combined. My results have either made no difference or have made the car drive worse. I've tried locking mapped points to eliminate variables, smoothing spark, slowing down the reaction of the throttle blade, etc, etc, etc...

So what am I missing? I want to understand what I'm doing incorrectly and how to fix it. I feel like I've been banging my head against the wall for a couple weeks now. I have a load bearing dyno all day Monday so I'd like to have a plan.

I'm considering flashing the entire Roush 2650 strategy to the car as a starting point. Is there really a difference between copying the parameters from the Roush file to the Bullit file, and paying for more credits so I can flash the entire Roush strategy?

I've attached some of the Roush/Whipple tunes I'm referencing along with what I currently have in the car. Any help would be appreciated. Thanks.

wanna try my tune ?

let me know if it runs better.

[evolmotorsprt]

wanna try my tune ?

let me know if it runs better.

I'll check it out. Thank you.

[LastPlace]

Thank you for the response. Noted on the ETC Tuning, and boost kicking in. I will be using the factory data for the Bullit TB. I currently do have the pill in the vacuum line. I've tried it in and out, and in definitely makes the transition to boost better. I may reduce the orifice size. 2 questions regarding your response:

1. What torque value are you using to derive you number in the torque tables from? I've seen so many different answers to this question, i.e. %err from etc into the TT histogram, %err derived from indicated vs ETC torque request, pasting indicated torque directly into TT... If I understand it correctly, the coyote cookbook suggests using the torque reading on the dyno and drivetrain loss to calculate actual torque.

2. In your opinion would it be beneficial to flash the complete Roush strategy?



Could you please define throttle feedback? Do you mean the pedal, desired torque, or the PID for throttle control?

I was under the impression that +KR was timing being pulled and -KR is timing being added. I do have a small amount of +KR when I'm hitting my cyl. pressure limit. Should I not take the PID "Knock Detected: No" at face value?

no simple answer on the torque tables, get on the Dyno and start working it out.

Use the tq to load tool and you can get it.

you can’t just flash Roush cal with out using the same iat harness location and several other things

[evolmotorsprt]

no simple answer on the torque tables, get on the Dyno and start working it out.

Use the tq to load tool and you can get it.

you can?t just flash Roush cal with out using the same iat harness location and several other things

I was able to start getting positive results today editing the torque tables and DD per gear.
WRT the Roush Cal, I realized it wouldn't work when I noticed the MAP was enabled in the file.

What I'm trying to figure out now is how to improve the relationship between pedal, TPS, and ETC area.

InkedLog_LI.jpg

I can feel that squiggly ETC line when I'm driving. I hoping to be able to smooth that out.

[LastPlace]

In order to do Roush Cal You would just have to set everything up physically like Roush does

[murfie]

wanna try my tune ?

let me know if it runs better.

That throttle body model is something else. If you wanted to split the TB model by where the bypass closes(5" of vacuum as an example) I would think it would look more like this.

Bypass.PNG

Effective area.PNG

The ETC calculator does not come even close to something that would work for angles on this.
This large effective area is just numbers satisfying the SD model for boost, when the TB model can only model air flowing in one direction and needs the pressure differential to represent that. SD still needs to come out close to actual MAP to keep injector data correct and hence fuel/ air load correct.

If you really think about it, when the bypass closes, air doesn't change how its flowing through the throttle, is should still be in control of the maximum amount of air flow. If it becomes the thing that is controlling the maximum airflow, then there's something wrong with the size of your bypass valve or it isn't functioning correctly. The only thing that changes is instead of the engine pulling the air through the bypass, and through the throttle, the supercharger rotors start pulling air through the throttle body. Air still flows though the throttle body the same, the ETC vacuum might actually be increasing as the vacuum in the supercharger inlet increases. With out a SIP sensor or model, the best you can do is keep the blade at 100% open at WOT. Let the SD calculate boost, and just ignore the SIP. You can model this using the SD, you just have to break it into three spots and get the air load that will correlated to the TB model, bypass open(5" vacuum or more), bypass closed out of boost(5"vacuum-.3" vacuum), and in boost( any thing more than .3" vacuum). In SD MAP inHg thats (0-24.9"), (24.9-29.6"), and (29.6"+) if the Baro you are calibrating in is 29.9" which is standard close to sea level. Air load is determined directly by the MAF, and its the engines VE.