2015+ Control Pack HDFx issues

[Nutter281]

Figured I would just start a different thread rather than continuing to build on the CookBook thread which had a different intended purpose.

Picking back up from that conversation, I'm attempting to tune spark/torque/iTorque by turning off MPs in the tune file. The first thing I noticed (as mentioned) was that my PCM never seemed to use MP1 or MP22 which is what most people were using to tune from. Next thing I noticed, when I turned OP 'off', I am still seeing a great deal of OP weighting (up to 100%) as monitored through VCM scanner.

Finally, I turned off all tables except MP-7, MP-8, and MP-OP. As recorded by VCM scanner, my PCM then began to exclusively use MP-13 and MP-14

Here is the log file and the tune. Clearly there is an MP misalignment either in VCM editor for my control pack tune or in VCM Scanner as it is identifying the weight being used for a given MP. Has anyone else seen this?

austin11_MPs.hpt
log11_a11MPtune.hpl


Thanks,
Austin

[murfie]

The ECU starts at the enable tables to determine the MP to use. If the MP you set do not meet what the distance tables go to in the arrays than it will use the next closes MP.

mode enable> distance table > array > MP

The snap to line is for weights of the arrayed mapped points.

The snap to point is for the weights of the distance.

Because they are connected in this way it takes more than just disabling the mapped point for the ECU to not use it. you would need to remove it from any distance tables and arrays. As well as not allow it in any lines or points.

In your log you only were in fuel economy mode. In the fuel economy distance table this is distances between 1 and 5.5. Looking in the fuel economy distance table and those distances do not get you MP 7 or 8.

The ECU then can use the snap to line and snap to points to try and get to 7 and 8 through Mapped points in other arrays. The snap to line can get it to mapped point 22 via the stability array which is weighted down to mapped point 14. 14 then can spill over into 13 trying to get to 7 and 8.

The ECU could also use snap to point. The closest would be distances 7 8 or 9. This would be allowed in the IMRC tables at distance 9 the lowest loads. The problem is distance 9 is mapped point 21 in the fuel economy array an open IMRC point. So in order to use this load would need to be above .20 once above 1500RPM. If you watch the log as it approaches this load and RPMS raise it moves more weight to 13 as it meets the imrc closed load and that is a closed IMRC mapped point.

If that doesnt make sense

Basically you forced the IMRC open/partially closed by only allowing MP 7 and 8. There are no paths for the ECU to get from the IMRC open MPs to the ones you enabled. the 2015+ coyotes work off IMRC position then cam positions causing the complicated mapped points to become twice as complicated.

[Nutter281]

Murfie, I really appreciate the help man. I've got a MS in Mechanical and Aerospace engineering but this is so unbelievably convoluted it sort of makes a guy want to take a swing at a Ford controls engineer LOL. I mean kudos for wasting 6 months of time fully characterizing the effect of temperature, velocity, and cam position on air density to save $5 on a manifold pressure sensor but FFS at least allow for one to be added as a monitored device in the tune... sorry I digress.

The first part of your post doesn't make any sense 'to me' simply because I am not yet familiarized with those regions within the tune file. I'll get there but it is taking some time. I'm guessing I overlooked them in the LaSota cookbook. Mapped points based upon cam timing make sense but the distance tables are a new concept to me and I need to do some research.

The last part of your post (what I thought was the dumbed down version) also leaves me a bit confused.

Basically you forced the IMRC open/partially closed by only allowing MP 7 and 8. There are no paths for the ECU to get from the IMRC open MPs to the ones you enabled. the 2015+ coyotes work off IMRC position then cam positions causing the complicated mapped points to become twice as complicated.

The majority of the tune file I sent was at idle. The car was in neutral the entire time. I have IMRC to open above 4000 rpm. The best fuel economy wants a load of 1.00 at 1000RPMs for them to open up. What part of the tune is putting the IMRC into a partially open state? I would have expected IMRC to be fully closed at this idle condition but here's where the learning takes place

Even if that sounded sarcastic, it was mostly directed at my own frustrations, I really do appreciate your help!


Thanks,
Austin

[murfie]

Yes Ford did a great job at making it convoluted. I have found more information on the quadratic speed density method Ford uses than how the distance and mapped points are used. Most of it is playing with it and guessing. I tried relating how it works to auto cad and other modeling programs.

The IMRC added two states and to deal with it Ford basically doubled the amount of mapped points. One set for open and one set for closed. Naturally there's going to be a transition point between mapped points. When looking at the load to open and load to close tables you just have to look at the x axis which is the IMRC distance reference. Ford gave the IMRC the ability to open over a range of mapped points and loads by doing it this way. Other wise it would perform like you said below rpm close, above it open. Distance 7,8, and 9 are the most common point where the IMRC transitions open as the loads are easily obtainable at the given RPMs. Also at those distances is the transition from best fuel economy to best drivability.

MP14 was getting all the weight, because that is an IMRC open MP you know your IMRC was open. It would shift to 13 as RPMs went up. So you know it started closing IMRC as the RPMs went up. This is the part of the HDFX that people hate. Some times it can feel like it has a mind of its own. Just realize you are asking it to do the impossible by using only MP 7 and 8 with OP while working with fully populated arrays and distance tables and it just makes it work. I also think different strategies have different fail safes when dealing with tables that don't make sense. I know one thing is it seems like the more the weight is spread out the happier the ECU is with its path in your tables. Plotting graphs against the weights you see the arrays clearer.

[Nutter281]

Quick question, just to clear my mind that something is not amiss with my PCM and/or the software monitoring it. Ford disabled the MP-OP on gen2 coyotes by default.
With it disabled, if I monitor all 28 MP weights, when I go WOT on the car with OP disabled (per stock tune), I will show 100% weighting of OP. This happens and I don't expect it to happen based upon what I had come to understand of the 2015+ PCMs.

Is this supposed to happen? Keep in mind, I have not yet altered any of these tables in any way shape or form from stock.

Thanks,
Austin