From my understanding, when you flash the PCM... your truck loses all of its prior learning with that tune on how to adapt to certain situations and conditions. It almost seems counter productive to me then to have multiple tunes, such as towing and performance. For instance, say I've been running my performance tune for the past 3,000 miles and its gotten everything nice and figured out. A weekend comes where I'm going to be towing some stuff and really want to utilize my towing tune.... So I slap the XCal on there and flash over the towing program. I just wiped all prior learning that the PCM did for the performance tune!!!??? Also while having that towing tune freshly flashed over it is (the truck) going to spend a few hundred miles or so learning that tune, correct? So I'd essentially not be gaining as much performance from either tune as I should be.

Just curious... anyone? Feel welcome to correct me if I'm wrong. Or does the tuner copy over the learned information when you swap tunes using the programmer?

 

First, it learns all over again in a few drive cycles. Secondly, it gets rid of all the bad habits it has learned. Some of us routinely disconnect the battery for about 15 miniutes while doing maintenance, which achieves the same results. It wipes the Adaptive Driving Memory clean and starts over.

 

Also, if the truck could continue to learn at the same pace as it did stock then the truck would lose it's performance characteristics over time. For example, when my truck has been running stock for a long time, it runs a 9+ 0-60 and about a 17-18 1/4 mile. After resetting the computer it will run mid to late 8s on 0-60 and around 16.5-16.8 on the 1/4. Another instance I had recently was a really bad batch of fuel. With the custom performance tune the truck would barely go down the road and even had issues idling. I set it to stock and 20 miles later it was driving like it was supposed to.

 

Adaptive controls (learning) only make corrections for closed loop operation when the O2 sensors detect errors from the programmed result, i.e. fuel trim errors. If the tune is properly written, the errors should be minimal and there should not be that much learning to affect the way the vehicle operates. Now if there are corrections, those corrections will ultimately be applied to open open loop operation. To compound the problem, adaptive learning only occurs within certain programmed parameters/limits. Suffice it to say a bad tune or a departure from stock configuration for the air intake tract (i.e. CAI) could have drivability issues or more disastrous effects.

 

In other words, if your truck has become used to driving with certain modified parameters, and then you flash your truck.. it suddenly does not know how to deal with whatever caused it to modify the program in the first place. Assuming that by having JUST the negative battery terminal down, I've cleared my PCM of any prior learning, and it runs just as good if not a little bit better... would it be safe to say that a properly constructed tune shouldn't have negative effects... again assuming that the new tune is properly setup and not drastically changing fuel/air ratio functions and gaining most of its power from something like... say, shift points and what not?

I'm also guessing that these things that trigger the PCM to "Adaptively learn" are also the same things that the PCM will flag as an DTC in the event that the event is far from normal, not just slightly off.... correct?

 

 

No, not really. It has nothing to do with how you drive. Because of variations in most mass produced components, the purpose is to make minor corrections based on based on O2 feedback to maintain the correct a/f ratio in closed loop. Here are a couple of links that may do a better job explaining.
http://fordfuelinjection.com/files/H...trol_works.pdf
http://members.shaw.ca/blk99sho/ford...ng%20notes.pdf

You want to clear the learned parameters when you change tunes, although you'd end up in the same place eventually. For example, if the system mades corrections to lean the fuel delivery and your other tune will ultimately richen it, you could have a problem
by being too lean when the tune is first loaded if the old corrections were employed.

If you're noting differences in performance after clearing memory, you might have a dirty MAF sensor or you might be running a CAI that has changed how the MAF interprets air flow.

 

There is a little bit of learning based on how you drive. This will usually have a minor effect on shifting though.

The part I have bold is the exact reason why tuning hampers the learning ability of the PCM. If it didn't then like Windsor said, you would end up in the same place. This is also why the P1000 code only appears after you program back to stock and not between the other power levels.

 

Windsor, I have read your post several times and I am still confused. I know nothing about how it works or what is learned. All I know is tha over a period of time of driving around town and a lot of slow acceration, and stop and go, my truck gets more unresponsive. It slips up on me as it is gradual. If the battery is disconnected for any reason, then it is all peppy and like a jack rabbit again.

 

There is some learning based on transmission functions, which isn't related to adaptive controls. Those are beyond my ability to understand and describe.

 

Do you have a programmer or any sort of CAI or other types of modifications?

 

I have Gryphon 87 Performance and 93 Performance tunes with the Gotts mod. But, it was like that when totally stock.

 

The transmission learning is based on pedal input compared to other parameters. Similar to the engine adapting fuel trims to input from multiple sensors the PCM adjust the shifting to fit the driver. This also goes hand in hand with the learning for the engine. The computer is always monitoring how all of the systems are working. If the pedal position has a tendency to go to 70% while at low speeds then a logical adjustment would provide later and firmer shifts. This results in a more "performance" oriented strategy. I'm sure this sounds like by hot rodding a stock vehicle all the time you'll get a free performance tuned computer but then we go back to the limits the vehicle is allowed to learn. This limit is very moderate so it wont ever match the effects of a programmer. This also works the other way (as bluejay observed), regular driving makes the truck drive more lethargic. Most regular driving includes cruising and slow acceleration. This will cause the computer to make adjustments that should improve mileage like milder fuel trims and earlier/softer shifts.

 

Bluejay, if you're running a K&N filter, you might want to clean your MAF sensor. Sometimes oil from the filter can settle on the element and give a false reading that adaptive control will make changes to correct. Aside from that, unless the transfer function for the Gryphon is off, I'm not sure if what you feel is the transmission logic making changes. Although that doesn't seem to be what you're describing.

 

Bluejay, do what Windsor said with cleaning the MAF sensors and such. Is it getting slow and unresponsive with Bill's tunes?

One thing you can do is get a datastream like this after you either reprogram or disconnect the battery. Later when the truck is feeling slow, do another and compare the two. Also helps if you have a spreadsheet program like excel where you can look at the actual numbers instead of a graph's representation.


I find this the most usefull. It consists of one extremely slow and even acceleration from 0-60 in 60 seconds (makes it easy to time the acceleration with the clock and speedometer) followed by a 0-90ish WOT and then a regular traffic acceleration.