Thread for those of us who tune our own cars

[Dan03mach]

Originally Posted by jchambers

Doing what you describe is likely completely screwing up your load calculations, amongst other things. Let's presume that for the factory NA calibration your commanded lambda at 5000 RPM and 750 throttle counts (or say a load of 0.899) is 0.879 (12.3 AFR on 14.1 stoich). Now you come along and really want something more like .799 (11.24 AFR on 14.1 stoich). If all you do to drive it to that .799 actual lambda is increase the airflow in the MAF table then you have fictitiously skewed that actual airflow measurement because the ECM is still calculating the pulsewidth based on the original .879 lambda. The actual airflow through the engine hasn't changed. Its the same now as it was originally, you just want to run at a different lambda at those same conditions. Your load will be higher, spark taken from a different point in the BDL table, etc. What you really want to do is change the commanded fuel point. This is done by changing the base fuel table at that point where you desire the change. The fuel needed to deliver that commanded lambda is then calculated based on the original UNCHANGED mass air curve; you're just asking for and receiving a longer pulsewidth from the injector. Load hasn't changed and the airmass entering the engine remains true. Does this make sense?

No I don't increase the airflow in the MAF table to skew the actual airflow for the lambda I want to achive. If you want to keep the stock curve then that is fine but you will still have to modify the MTF to adjust for the 7 x larger then stock fuel injectors I am running. And you will also have to adjust the min. pulse width and rescale the injectors. I think the largest you can scale the injectors to with the software is like 96lbs or something like that and with me running 160lb injectors everything is going to be skewed just a little bit. Now if I were running 60lb injectors I could just rescale the injectors and everything would be fine.

[TKat13]

Originally Posted by Mach828

Bosch sensor. I just replaced it. I should have picked up motorcraft ones. Yes adding fuel stft is 1.25. Maybe the leak is worse than I thought.

Welcome to the tuning party SLOWHITE.

FWIW, when I told my tuner I installed new O2s and that they were Bosch, he told me to either put my old ones back in or buy new Motorcraft O2s.

BTW, I have 2 almost new Bosch O2s just collecting dust, cannot return them since i lengthen the wires. lol.

[jchambers]

Originally Posted by Dan03mach

No I don't increase the airflow in the MAF table to skew the actual airflow for the lambda I want to achive. If you want to keep the stock curve then that is fine but you will still have to modify the MTF to adjust for the 7 x larger then stock fuel injectors I am running. And you will also have to adjust the min. pulse width and rescale the injectors. I think the largest you can scale the injectors to with the software is like 96lbs or something like that and with me running 160lb injectors everything is going to be skewed just a little bit. Now if I were running 60lb injectors I could just rescale the injectors and everything would be fine.

I can see right now Dan that you're not interested in understanding or applying the principles correctly so I'm not going to argue this any further. I was attempting to provide clear, concise instruction on the proper methodology. My 5000+ hours of calibration (not tooning) experience have been in error it would seem. Have at your methods. I would hope that others don't follow suit.

Before I leave, can I ask one question? If the base fuel table isn't referred to under open loop conditions, then just how does the ECM calculate how much fuel (pulsewidth) to add for any given airflow on your MTF curve?

[Dan03mach]

Originally Posted by jchambers

I can see right now Dan that you're not interested in understanding or applying the principles correctly so I'm not going to argue this any further. I was attempting to provide clear, concise instruction on the proper methodology. My 5000+ hours of calibration (not tooning) experience have been in error it would seem. Have at your methods. I would hope that others don't follow suit.

Before I leave, can I ask one question? If the base fuel table isn't referred to under open loop conditions, then just how does the ECM calculate how much fuel (pulsewidth) to add for any given airflow on your MTF curve?

I would have to get infront of the software and my tuning files to answer that and I do not have my laptop with me at the moment. But if I'm not mistaken that is where you actually control and set your pulsewidth for the given amount of airflow/mafcounts your maf is reading. It's actually very simple. In a nutshell you are telling the computer you want X amount of Fuel for this amount of airflow/mafcounts. That is why you have to tune the car from idle all the way up the curve till you reach your highest mafcounts your maf is seeing. You will need your injectors scaled correctly and the correct min. pulsewidth set first before you do anything. So you have 5000+ hours of tuning on cars? Thats a long time. Calibrations is what I do for a living, Instrumentation and controls, tuning and playing with this car is just a hobby of my but I do understand it very well and if we were to sit down you could probably show me a few things and I'm sure I could do the same.

[Dan03mach]

Originally Posted by TKat13

FWIW, when I told my tuner I installed new O2s and that they were Bosch, he told me to either put my old ones back in or buy new Motorcraft O2s.

BTW, I have 2 almost new Bosch O2s just collecting dust, cannot return them since i lengthen the wires. lol.

Amazon has good prices on Motorcraft o2's. I ordered some not too long ago for like $34 each.

[jchambers]

Originally Posted by Dan03mach

So you have 5000+ hours of tuning on cars? Thats a long time.

Yes.....at least. I've got enough miles on my Mustang Dyno to travel around the world several times. I teach calibration to other shops and calibrators, live to dissect the OEM code (when I can get my hands on it), and even develop tuning tools for other calibrators and even the software developers. Enough horn blowing though. Let me know when you find the source code on how airflow (maf counts) is turned into injector pulsewidth. My 26+ years of engineering at NASA and Masters in Engineering Mechanics helps too.

[Mach828]

Thanks for jumping in this thread Jeff. Appreciate your insight. Did you notice any misinformation or mistakes with the info I've posted? I want to make sure I'm tuning correctly and not taking any shortcuts.

I'll order a set of motocraft o2 sensors and stay away from Bosch going forward. Thanks!

[Dan03mach]

Originally Posted by jchambers

Yes.....at least. I've got enough miles on my Mustang Dyno to travel around the world several times. I teach calibration to other shops and calibrators, live to dissect the OEM code (when I can get my hands on it), and even develop tuning tools for other calibrators and even the software developers. Enough horn blowing though. Let me know when you find the source code on how airflow (maf counts) is turned into injector pulsewidth. My 26+ years of engineering at NASA and Masters in Engineering Mechanics helps too.

Not sure on the source code. I know one thing you adjust the MAF data for the desired MAF ADcounts to control the amount of fuel the injectors deliver. Now if this is in closed loop you can data log stft & ltft to calculate the amount of adjustment needed set by the base fuel table at that desired load TB, & RPM. If this is in open loop you will not be able to data log stft or ltft to have any idea how much you need to adjust the MAF data for the desired MAF ADcounts unless you can data log your a/f while driving etc. under different loads. Now I'm sure you already know this but that MAF Data can be wrong if you increase the size of your injectors and can not scale your injectors correctly at that point you will have to adjust that data thru out the complete curve, and with injectors like mine 160lb my sct software will not adjust those to scale in the software so I have to adjust the MAF data in the curve to compensate the difference.

[jchambers]

Originally Posted by Dan03mach

Not sure on the source code. I know one thing you adjust the MAF data for the desired MAF ADcounts to control the amount of fuel the injectors deliver. Now if this is in closed loop you can data log stft & ltft to calculate the amount of adjustment needed set by the base fuel table at that desired load TB, & RPM. If this is in open loop you will not be able to data log stft or ltft to have any idea how much you need to adjust the MAF data for the desired MAF ADcounts unless you can data log your a/f while driving etc. under different loads. Now I'm sure you already know this but that MAF Data can be wrong if you increase the size of your injectors and can not scale your injectors correctly at that point you will have to adjust that data thru out the complete curve, and with injectors like mine 160lb my sct software will not adjust those to scale in the software so I have to adjust the MAF data in the curve to compensate the difference.

You still didn't answer the question. How is the fuel mass determined? If all you're doing is adjusting the MAF curve to hit some fueling target that you've got in your mind, then you're doing what is referred to as 'raping the MAF curve'. When in open loop, the ECM is looking to the base fuel table for the target lambda. Since you've forced it open loop, its looking to this table at ALL times (sans startup) whether you're at WOT or not. When you adjust the MAF curve, you should be seeking to match your measured lambda to this commanded lambda from the base fuel table. Look, if you guys are doing this yourselves you should at least be trying to do it as correctly as possible.

[Dan03mach]

Originally Posted by jchambers

You still didn't answer the question. How is the fuel mass determined? If all you're doing is adjusting the MAF curve to hit some fueling target that you've got in your mind, then you're doing what is referred to as 'raping the MAF curve'. When in open loop, the ECM is looking to the base fuel table for the target lambda. Since you've forced it open loop, its looking to this table at ALL times (sans startup) whether you're at WOT or not. When you adjust the MAF curve, you should be seeking to match your measured lambda to this commanded lambda from the base fuel table. Look, if you guys are doing this yourselves you should at least be trying to do it as correctly as possible.

I see how the ECM is looking at the BFT in Open loop or WOT so how does it get to the desired lambda at WOT or open loop? My MTF is pretty close to what I have commanded in the BFT but not exactly the same. How does the ECM make adjustments to try to get to the commanded lambda in the BFT? I thought if I adjust the #/min. at any AD count then it would either increase or decrease the amount of fuel for that AD count. Now if your BFT is not the same as this desired a/f then you need to adjust the BFT but i'm trying to see how adjusting the BFT makes any adjustment to what is commanded in the MTF.

If someone is knows something that I don't I'm always willing to learn

[Mach828]

In closed loop I think it looks at the field were you can adjust the desired stoich. I think my stock value was 14.67 and I changed to ~14.1 for e10. Its the value you have to change for running different fuel besides pure gasoline, i forget the exact name. Then it uses that as a reference to determine closed loop fueling and then as a reference for the commanded lamba for open loop. So in theory, if this value is 100% accurate for my base 91 e10 tune and then I adjust it to the true stoich for my e85 tune, then the only thing I need to change for my e85 tuning would be the timing tables? I'm sure it won't be entirely perfect and I will need to fine tune, but this is just in theory.

[jchambers]

When forced open loop it is always going to be looking at the base fuel table for the commanded lambda once the stabilized operating conditions have been attained (e.g. cold start modifications have expired). This is the commanded lambda that the computer is targeting. For your EEC-V car, the computer won't know where its at hence the need to properly calibrate it. Its going to take the airflow reported from the MAF (simplistically) and then use the commanded lambda from the base fuel table to calculate the desired fuel mass and further calculate the pulsewidth required to arrive at that desired fuel mass. It will not adjust itself. Again, this is where the calibration comes in. Now when you start talking about the Copperhead platform (2011+) the on-board widebands are always feeding back so that the ECM is constantly correcting the fueling. These always operate in closed loop, there is no open loop even under WOT conditions (again ignoring the cold start regime).

Again the point here is to make sure that your base fuel table reflects the commanded lambda that you want to achieve. Then make your MAF corrections so that your actual measured matches this commanded as closely as possible. In theory, if you've done your scaling correctly then your load calculations will be correct and things will be a lot happier all around.

[Dan03mach]

Originally Posted by jchambers

When forced open loop it is always going to be looking at the base fuel table for the commanded lambda once the stabilized operating conditions have been attained (e.g. cold start modifications have expired). This is the commanded lambda that the computer is targeting. For your EEC-V car, the computer won't know where its at hence the need to properly calibrate it. Its going to take the airflow reported from the MAF (simplistically) and then use the commanded lambda from the base fuel table to calculate the desired fuel mass and further calculate the pulsewidth required to arrive at that desired fuel mass. It will not adjust itself. Again, this is where the calibration comes in. Now when you start talking about the Copperhead platform (2011+) the on-board widebands are always feeding back so that the ECM is constantly correcting the fueling. These always operate in closed loop, there is no open loop even under WOT conditions (again ignoring the cold start regime).

Again the point here is to make sure that your base fuel table reflects the commanded lambda that you want to achieve. Then make your MAF corrections so that your actual measured matches this commanded as closely as possible. In theory, if you've done your scaling correctly then your load calculations will be correct and things will be a lot happier all around.

That makes plenty of sense. I was looking at my tune file and the BFT it looks like I have it pretty **** close to where it needs to be. That is probably why my car is running and driving very good thru the curve of the MAF. Thanks but where do you get the desired fuel mass? Not sure about that one.

[Dan03mach]

Originally Posted by Mach828

In closed loop I think it looks at the field were you can adjust the desired stoich. I think my stock value was 14.67 and I changed to ~14.1 for e10. Its the value you have to change for running different fuel besides pure gasoline, i forget the exact name. Then it uses that as a reference to determine closed loop fueling and then as a reference for the commanded lamba for open loop. So in theory, if this value is 100% accurate for my base 91 e10 tune and then I adjust it to the true stoich for my e85 tune, then the only thing I need to change for my e85 tuning would be the timing tables? I'm sure it won't be entirely perfect and I will need to fine tune, but this is just in theory.

yes in theory that would be correct and it is pretty close but not all E85 is exactly 85% that being said you have to adjust for the difference. The E85 in my area is usally around 88%. I don't run it any longer but that is what it use to be.

[jchambers]

Desired fuel mass is an internal calculation based on a number of parameters. Just from memory (so don't kill the messenger if I don't have it 100% correct), but in its most simplistic form:

AFR = airmass/fuelmass therefore fuelmass (Mfuel) = airmass/AFR

Mfuel = airmass/(AFRstoich*Lambda) => See how the base fuel table is used <=

Now this is very, very simplified. In the speed density world (and somewhat similar in the Ford inferred airflow modeling):

Mfuel = ((Pmap*Nvol*Vcyl)/(Tair*.28))/(AFRstoich*Lambda)

Again, see how the base fuel is used.

Since fueling is what is being commanded (not airflow), the pulsewidth required to deliver that desired fuel mass is then derived from a pretty serious calculation (or series of calculations) using the injector parameters, factored by temperature, pressure, voltage, and a host of other factors.

[Dan03mach]

Thanks for clearing that up. You can't do speed density tunes with the SCT software can you?

[jchambers]

Not easily by any means. The Ford model is an inferred model; the airmass, load and VE are indirectly calculated (hence the inferred) from the set of linear coefficients found in the speed density category. This gets to be real fun and very complicated in later platforms as it is quadratic in nature and is used in the series of checks and balances for the drive by wire system.

[Dan03mach]

Originally Posted by jchambers

Not easily by any means. The Ford model is an inferred model; the airmass, load and VE are indirectly calculated (hence the inferred) from the set of linear coefficients found in the speed density category. This gets to be real fun and very complicated in later platforms as it is quadratic in nature and is used in the series of checks and balances for the drive by wire system.

Just never heard of anyone doing a speed density tune on these cars. Not without a fuel managment system like a BS3.
I just picked up a 2015 GT and don't even want to get trying to tune that car. But that hellion twin turbo system does look Intriguing.

[SLOWHITE]

So after reading all of your comments guys I feel like I am the dumbest one in this group lol. So doen anyone wanna take a look at my tune before I blow my **** up next time I start the car?

[TKat13]

To guys participating, absolutely love the conversation, you guys are educating many of us that are not replying, so, thanks! I am buying the book today.

Even if I do not tune my car right away, understanding everything my tuner tells me is extremely important to me.

[ab_mach1]

Originally Posted by TKat13

To guys participating, absolutely love the conversation, you guys are educating many of us that are not replying, so, thanks! I am buying the book today.

Even if I do not tune my car right away, understanding everything my tuner tells me is extremely important to me.

I Agree!

[jchambers]

Originally Posted by Dan03mach

Just never heard of anyone doing a speed density tune on these cars. Not without a fuel managment system like a BS3.
I just picked up a 2015 GT and don't even want to get trying to tune that car. But that hellion twin turbo system does look Intriguing.

Just wait until you get into the 2015 stuff. There are 32 borderline knock tables, 32 MBT tables, 32 tables for each of the slope, offset and quadratic SD terms, 32 torque and inverse torque tables,.....and on and on and on, all to add to the complexity of the HDFX TiVCT control. Then imagine for trans tuning where you have six upshifts, six downshifts, six locks and six unlocks for each of eight modes of transmission operation.

[01yellerCobra]

Originally Posted by jchambers

Just wait until you get into the 2015 stuff. There are 32 borderline knock tables, 32 MBT tables, 32 tables for each of the slope, offset and quadratic SD terms, 32 torque and inverse torque tables,.....and on and on and on, all to add to the complexity of the HDFX TiVCT control. Then imagine for trans tuning where you have six upshifts, six downshifts, six locks and six unlocks for each of eight modes of transmission operation.

And that's why I keep it as a hobby with just my car.

[watman02]

Jchambers thank you for your much more articulate way of explaining it to dan03mach! I was trying to get there with him earlier but we seemed to be arguing more than anything. No offense dan03mach.
Jchambers well spoken!

[Mach828]

I'm finishing up an excel sheet that will populate the entire load with failed maf table from a datalog of just air flow, rpm, and pedal position. Doing it manually seems like a very time consuming process. I didn't realize how important that table was. I'll upload and share it once I get it all squared away.