Mapped Point Isolation Calibration

[Higgs Boson]

How can we tune things like Speed density or Spark or Torque/Inverse tables if they are by mapped point and the points are always blending.

Is there a way we can keep the car in individual points to tune each MP table separately or am I being ridiculous?

[AKDMB]

mp1.PNG
Don't know if this is what you are asking for, but you can disable individual mapped points. As far as the speed density, spark and torque tables, that's a question for someone whose not me, and actually knows what they are doing.

[Higgs Boson]

I guess that's what I am asking is what is the process? Disable all but one mapped point and tune each one individually? Would we have to fill the distance tables with the mapped point we are using also?

[txcharlie]

One way to do it, is make a spreadsheet that records the differences between the MP tables and one you use as the base. Most likely the OP table would be your base.
Disable all tables except your base and tune the base table using a dyno(or the street). Re-enable all MP tables and use your spreadsheet(multiply the percentage difference to your base) to populate new MP tables.

For example:

Using 2014 GT A6 Stock file,
OP Borderline .10/550 (load/RPM) = 40*
MP 1 Borderline .10/550 = 45*
45/40=1.125
1.125 is the percentage difference. (12.5% greater)
If you tuned your base table(OP) and found that 47* was the sweet spot, at .10/550, you would multiply 47* by the difference(1.125) and get 52.9* at .10/550 on MP 1 Borderline table.
This is why you need spreadsheet a because once it's done, you can do an entire table with a few clicks.

I would post mine, but HPT thinks my Open Office files are invalid and won't let me post them.

Murfie posted his XL file in this thread, post #10. I haven't used his, but he says it works fine.
There are issues when one of the values is 0 or less, but Murfie came up with a solution in his XL file.
Another solution for this is to make all 0 or less values = 0.
http://www.hptuners.com/forum/showth...-auto-populate

[murfie]

I see 28 tables as flexibility and the distance tables/arrays are the path Ford set for the ecu to try and stick to. If the ecu notices knock occurring or not occurring at a certain spot in this path it has the flexibility to tweak the path a little. It is just another way ford keeps the ecu at "good/safe" numbers.
Subaru had something similar. there was a program called learning view that specifically showed the different types of knock learning. They also had a number that when it reached a certain point the ecu was done learning and happy with the path.The old trick was to hold both the gas and brake at 50mph to speed up that learning process. This is how you would defeat the learning to make your engine run harder but less safe. It usually lead to a blown motor. Not sure if we have any thing similar to that in hptuners scanner or if the learning /tweaking is a constant that is happening.

That being said. I would set all spark tables the same. Pick a table any table. Tune it to 0 knock or as close as possible with maybe a few degrees negative here and there. you don't want any positive knock numbers. Drive the car let it learn and see what it does with your values. Repeat until the car is happy. Then based off that table modify the rest with a spread sheet. Don't use the spread sheet until your values and cars learning are happy. This is exactly what txcharlie described.

If you want to you can take that a step further and tune it around the imrc low rpm and high rpm. Choose like table 7 for any tuning below 3500 and table 14 for tuning above 3500. Then from there do the percentage difference between all other mapped points with the same imrc position.

The next step would be looking at the open and close tables for IMRC and the snap to line points. You will see two lines one 0-4(IMRC open) and one 6-8(IMRC closed). I believe these lines are for the transition from open IMRC to closed IMRC at low rpm/ low load. The open/closed IMRC tables follow the array points not the actual mapped points thats why they can open and close at mapped points that the timing mapped points say they should be closed all the time at. So you could even tune one point from each of those groups or mapped point 5 as a transition point that would be most prone to knock. Then of course use the spread sheet to give the ECU its HDFX back.

For speed density the ECU starts at the offset values for every MAP calculation. They are just calculated MAP for given RPM so it shouldn't be to difficult to modify correctly. With Speed density(desired air mass) disabled compare the MAF calculated MAP,MP, and RPM to the stock values and modify accordingly. After getting the values from the MAF in the speed density what you could do is unplug the sensor, but Unplugging the MAF will also unplug the IAT sensor. So you have to lower the MAF fail RPM under engine diagnostics- air flow(Above this RPM threshold the MAF will not be considered failed) and tweak them from there based on the drivability or dyno numbers to give you a more accurate speed density. this is the easy way If you don't want to go through all the math behind the speed density calculations.

MAF failed RPM: Above this RPM threshold the MAF will not be considered failed. Is this a typo not really means now?

As for the torque values that are in the ECU I don't think they need to be the actual torque numbers the engine is producing. To me they are just a number that relates the pedal position with the throttle position and the throttle position with a load and RPM relationship. I see many people who start to change the torque values to compensate for a change in the physical engine parts. This is the wrong way to go about tuning the ECU. A better way to do this would be to change the model(throttle body, speed density, and intake manifold) that describes the physical engine parts rather than numbers that are based on the old engine parts model. Also change the values the torque numbers are related to through the load/ RPM relationship. The actual power/torque the engine is producing shouldn't matter to the ECU. The only reason it should know is so when I tell the ECU I want x torque by the throttle angle. It has a number to navigate through tables for values that will best get the engine to do what I want it to do. Im satisfied with fords throttle mapping and only want to change the values the mapping leads to. I ran into this with Subaru tuning a lot and until I left the stock throttle mapping alone I had a very difficult time tuning the engine well. If one throttle map tells the ECU I want 250ft lbs torque and another tells the ECU I want 300ft lbs torque but the path through the tables both lead to the same values of air, fuel, and spark the engine will produce the same actual torque. With that said I see no reason or gain from telling the ECU a different torque value when the engine is only going to see spark, fuel, and air values.

Seems to be a reoccurring pattern in tuning fords. disable ECU calculated values- log sensor values- modify calculated values. Then you do it the other way to fix the speed desity. unplug sensor- log calculated values- modify calculated values for improvement.
At least it makes actually tuning easy you just change what you want until you think you are at a sweet spot then modify the ECU calculated values and logic.

So the easiest way to tune these different parameters are in breaking it down into groups of mapped points. The percentage distribution between the mapped points that you logged needs to be applied to the value you want to change and distributed properly through out the mapped points. The key is small throttle position changes so that you do not get the load distributed through out too many mapped points at once. Cruise control is great for this.

[murfie]

playing with the snap to point/ snap to line/ knock advance I think I noticed something. The MBT tables are the ignition the car would like to run to make the most power. The borderline tables are where it starts. the borderline tables assume the worst case scenario. From there it listens to the knock sensor and if everything is good it starts its path toward the MBT value and gets as close as it can before the knock sensor stops it. Stock most of the tables values are further apart than the knock advance could make up for. The few exceptions to this are the tables the fuel economy array refers to. They are actually come extremely close to the MBT in spots. The system seems to work as much as it can to monitor knock and keep retard or advance at 0 or as close to it as possible and if using a different path than what the fuel economy array says to achieves this then it will. I set up my snap to point/ snap to line to target MP 26 instead of 14 and it tried really hard to go to 26 but MP 21 was closer to MBT and knock advance was less, so in reference to what the knock sensors were saying that's where it went. The closer I got borderline MP 26 to MBT 26 the more it went to 26 and the less knock advance was needed until knock started happening and then it went back toward 21.

I think to truly target a MP you need to get the ignition timing borderline path happy then tune the knock sensors off and it should just follow what ever the snap to line/ arrays are set as. Then you can re-normalize torque tables and turn the knock safety net back on. I don't know if this is true, but I'm just trying to make logical sense of what i'm seeing. This also means the engine needs the knock sensors to make optimal power safely. Bumping the borderline tables up just pushes the knock advance deeper into unsafe territory but that's where the power comes from. There probably isn't anything you can do to improve how the knock sensors are calibrated by ford. After bumping up the borderline tables to be safe I would dial back how aggressive the knock advance is so that the car can still use it to a beneficial degree. When you are 1 degree away from knock you need it advancing .25* a second instead of trying to to go for 2* or 4*. I think you can get it down to .05* a second. That's pretty accurate and I don't know if fuel quality will allow that. Maybe a more knock resistant fuel like high quality E85.

[txcharlie]

You can adjust the sensitivity of the knock sensors in Spark/Knock Sensors/Knock Threshold. I believe the higher the value, the less sensitive the KR sensor becomes.
You may be onto something with the speed of KR advance. I was having positive knock in the same areas the KR had just advanced spark, so maybe my advance rate for KR was too aggressive. It happened on 93 and E85.

[murfie]

Yeah you can make it less sensitive. I just think like the advance rate needs to be reduced to still have it be used with more aggressive borderline do you want to really reduced it. Excluding something like a noisy PD blower. You then could make them more sensitive but then it's possible to pick up more than necessary knock. I think they are pretty well calibrated to this engine from Ford and don't need much tweaking if any at all. It would be nice if someone else could confirm what I see with the MPs and advance rate.

[txcharlie]

FYI, the Roush tunes come with the Knock sensors turned off. I believe they must have dealt with enough false knock from the blower that they decided that was their only option. That's why the Roush tunes are considered conservative. You would have to be conservative without the safety net of the Knock sensors.

[bcmarek]

I thought I read somewhere, in this forum or another, that the early Roush tunes had them turned off, but the later ones had them turned back on.

Also, not to hyjack this thread, but did you figure out where your knock was coming from? I am still getting knock readings sometimes right after spark source shows tip-in, at 35-45% load in low RPM's.

Stock timing and no changes to tip-in settings. I've even removed timing in those cells with no luck. Maybe it's just showing the amount of spark it's pulling for tip-in.

[Higgs Boson]

Torque Management spark reductions often show up in the knock retard pid, have to be careful there. Does the same thing with GMs. Not sure if it's the nature of the beast or the HPT Scanner.