Formulating arguement to explain my gas milage

M

Mike Poore

Crabmonster
As many of you know, I've been reporting gas mileage in the 29-30mpg on interstate highways during long trips and at speeds in the 70mph range. (speed limit + 5) Also, I've been blaming the increase from ~26mpg on the installation of the Pro-Guard skid plate.

Others have not seen this increase, and I've been wondering why, so here's a thought I'm working on.

My recollection of the computer control in my ''88 5.0 LX and later the '97 Cobra was that it "learned" by my driving habits to adjust itself from gas sipping (best economy) to "let's boogie". Is this the same situation with the present generation cars, and if this is the case, do the "Tuner chips" do this as well?

Let me stop here, and get your thoughts on this point, then continue.
 
Mike Poore said:
My recollection of the computer control in my ''88 5.0 LX and later the '97 Cobra was that it "learned" by my driving habits to adjust itself from gas sipping (best economy) to "let's boogie". Is this the same situation with the present generation cars, and if this is the case, do the "Tuner chips" do this as well?
Click to expand...
Yes, Mike, the MM EEC has "adaptive learning" which stores certain engine settings for future use. This memory bank is somewhat volatile, it can get wiped clean by disconnecting the battery for a while (maybe overnight?), or, by flash tuning the EEC.

I cannot say the use of a flip chip disrupts adaptive learning, unless it's been reprogrammed to make an adjustment when engaged. Adaptive learning can be switched off in the programming, either in part, or completely. This is why I suggest to those using a HHP to flash the EEC, that they need to do a few WOT passes with the new programming to commit new parameters to memory. However, most seem to do this anyway as a side effect of trying out their new tune. But, if not done accidently, or, intentionally, the car may act up for a day or two, and could lead to an impression of other problems such as hard starting, or rough idle when cold.

With respect to your excellent mileage, have you checked your speedo for correct calibration? You could be over reporting your actual miles traveled by virtue of a tire swap, or, a gear change, and that could lead to incorrect MPG. I know I have a 1 MPH error, and I factor that in when I'm crunching numbers. Like a thyroid, it can be underactive, or, overactive, and you can always check your speed and mileage against a GPS reading, or, during a dynotune. Hope this helps...
 
SergntMac said:
Yes, Mike, the MM EEC has "adaptive learning" which stores certain engine settings for future use. This memory bank is somewhat volatile, it can get wiped clean by disconnecting the battery for a while (maybe overnight?), or, by flash tuning the EEC.

I cannot say the use of a flip chip disrupts adaptive learning, unless it's been reprogrammed to make an adjustment when engaged. Adaptive learning can be switched off in the programming, either in part, or completely. This is why I suggest to those using a HHP to flash the EEC, that they need to do a few WOT passes with the new programming to commit new parameters to memory. However, most seem to do this anyway as a side effect of trying out their new tune. But, if not done accidently, or, intentionally, the car may act up for a day or two, and could lead to an impression of other problems such as hard starting, or rough idle when cold.

With respect to your excellent mileage, have you checked your speedo for correct calibration? You could be over reporting your actual miles traveled by virtue of a tire swap, or, a gear change, and that could lead to incorrect MPG. I know I have a 1 MPH error, and I factor that in when I'm crunching numbers. Like a thyroid, it can be underactive, or, overactive, and you can always check your speed and mileage against a GPS reading, or, during a dynotune. Hope this helps...
Click to expand...
Sure does, Mac, and that was the answer I was hoping for. Now, it's becoming clear what's going on, since Dan made, basically, the exact same trip as me, and didn't report the same results, with his Pro-Guard installed. Now, thanks to your info, I'll go ahead and post my thoughts on what's going on.

Yes, on the GPS. I run a Garmin Street Pilot, and the speedo's right on the money.
 
Last edited:
Continuation of the post.

Ok, with Mac's info, here's what I think is going on: My MM is completely stock. No gears, no "tuner" chip, stock, original tires. It's got 20K on the speedo, and is an '04, December '03 build. I use 93 octane fuel, Amsoil in the crankcase and carrier unit, stock fluid in the transmission, tires pumped to 35psi.

The stock "chip" is an adaptive unit, that is, it compensates for many factors, but most importantly, for this arguement, how you're driving. For instance, by the second pass down the strip, it's in "Let's Boogie" mode; however, if you're on the interstate doing nearly the speed limit, and using the cruise control, it's gonna optimize for fuel efficiency.

So in all eight instances, when I reported gas mileage at nearly 30mpg with the Pro-Guard installed, it was under the following conditions: I had topped off the tank, entered the interstate, put the cruise control at the speed limit +5 and drove to the next gas stop. And, except for the time when we stopped at Barry's, then followed the guys to Detroit, these were the conditions. (we went a little over my 5+ rule, then)

As far as I know, the only difference between my MM and Dan's is the chip, and, perhaps, gears. It may be, that many of the "tuner" chips are not adaptive, and unless you have a hand held unit, may be in "Let's Boogie" all the time, with no adaptive function whatsoever.

So, to sum up, what has the ProGuard done for me? In my case, with everything being completely stock, in an '04, built in Dec '03, with Amsoil in the crank, and rear, and 20K on the odometer, my gas mileage went from 26mpg to 29.6mpg on interstate highways doing, basically the speed limit, with the cruise control on. That's better than a 10% increase ...significant, in my book. :)
 
Mike Poore said:
...significant, in my book.
Click to expand...
Ditto...Mine too, and thanks, Mike, for breaking it out for us.

I think we should expect to learn that there is a lot more for us to learn here, about EEC programming. The SCT Pro Racer crap gives us a whole Hell of a lot of control over adaptive stuff, but it's shy on instruction on how to manipulate it. Jerry W. will be in my area this coming Friday, maybe I'll pick up some more 411 for us?
 
i have an 04 stock just have aftermarket air filter, mobil 1, flowmasters. i get 19-20 mpg out of a tank and i commute 50 highway miles each way. i usually run OD off in the city (OD on at probably 55+)... but with basically 100 highway miles when i drive to work i thought the MPG would be better. dunno
 
Mike Poore said:
It may be, that many of the "tuner" chips are not adaptive, and unless you have a hand held unit, may be in "Let's Boogie" all the time, with no adaptive function whatsoever.
Click to expand...


I don't think the chips have any adaptive function whatsoever. I believe they just contain a new set of calibrations or possibly adjustments to calibrations already loaded on the PCM itself (most likely why you need to provide what is currently loaded in your PCM when ordering a chip). I imagine the PCM checks if anything is present at the port when it 'wakes up' and if so, it goes to work reading the new values to substitute over the preloaded ones.

I don't have any concrete data but I'd find it hard to believe that fuel mixtures would be altered depending on driving habits. The purpose of the whole closed-loop control system is to make sure that given a whole bunch of inputs, the resultant mixture ends up being 14.7:1 (or whatever the situation calls for, but I believe it's almost always 14.7:1). Would a PCM be allowed to fudge the numbers just because it happened to be driven by a more aggressive driver? I'd see the EPA screaming bloody murder...

This story/rumour *might* be a result of a characteristic of factory O2 sensors... they are narrowband, which means their outputs peak in either direction within a very narrow window. Using guessed numbers, it would be like pegging the O2 sensor voltage when the detected ratio was 14.6:1 or lower, and pegging it on the other end when the ratio was 14.8:1, with the full 1 V swing happening between those ratios. Because of the need for richer ratios at WOT (like 12.5:1 etc.), you can forget about using the O2 sensor outputs to run the show, so most if not all factory PCMs disregard the O2 sensor values at WOT altogether.

Since every engine's needs can vary, how does the PCM make sure it's doing things properly for YOUR particular engine? I'm not sure if all PCMs do it the same way but I know many of them would take the data gathered in the operating mode *just* before WOT as a baseline, a sort of educated guess to be used as a base and then offsetted to arrive at the richer WOT results. As has been said before, this is probably why custom on-the-dyno tunes yield such better results than a cookie-cutter tune, because each engine will need a different offset to apply on the original educated guess to get to the best results.

So maybe a more aggressive driver will allow the PCM to learn how to better guess what's needed to do a good job at WOT, but I don't think it works the other way around. At lower power demands, the PCM won't go looking anywhere near the values stored for higher power demands (below WOT) and so it'll just keep doing its best 14.7:1 job.

If I'm wrong, I'd appreciate a good technical explanation of how it really works.


Side note: if driving smoothly really did put the PCM into a snooze mode, I wouldn't have been able to pull the string of 14.6s and eventually the 14.56 after practically tip-toe'ing to the track. Why did I take it easy? Story has it that the LS1s like the one in my WS6 drop down to a less aggressive timing curve after detecting a certain number of knock situations. The idea being that knock has to actually happen before being heard, so rather than risk a situation of knock / pull timing / give it back / knock / pull timing / give it back / etc. it just eventually gives up and backs off for good, until it detects a certain amount of fresh fuel has been added to the tank during a fill up. This helps in situations of getting a batch of bad gas, etc. So not knowing if my Marauder had the same logic, I didn't take any chance having the car pull back timing for good with my half-91 octane / half-94 octane brew that was in the tank that fateful evening. Hence why I drove very gently to the track.
 
Last edited:
TripleTransAm said:
If I'm wrong, I'd appreciate a good technical explanation of how it really works.
Click to expand...

TTA, you're dead on. The idea of adaptive learning is sometimes misquoted. The PCM actually derives info for differing load situations and constantly updates them on a weighted moving average. Load is derived from inputs such as engine speed, throttle position, coolant and air temps., air mass and flow, EGR (On Fords), knock sensor and vehicle speed/gear. Those learned parameters help it to 'predict' what to do when that specific situation comes up next. So it is literally a "learned" response (Similar to a calculator. e.g. 2+2+2=6, 2+3+3=8).

For example, you're driving along at 50mph at a steady cruise and want to pass the idiot backing up traffic. When you change the throttle by pushing on the go-pedal, the fuel enrichment, timing, gear change, torque convertor, EGR and canister purge all default to the "average" of the previous events that match that criteria. Coolant and air temperature only enrichen up to a certain cut-off temperature and then are literally dormant sensors at that point. The knock sensor doesn't have any effect until a knock is detected. Code clearing (Generically) or PCM programming erase all those "learned" responses therefore requiring a brief period of re-learning. After the initial responses are set, they are constantly adjusted each time an individual load zone is encountered (Moving Average).

Virtually all feedback comes from the oxygen sensors - all four of them! All load input is from the above mentioned sensors. Any fault on either end, either feedback or input sensors, sways the moving average. Those stupid little resistors you can buy on e-pay for the air temp sensors are a good example. They lower the ACT reading back to a level that adds enrichment. Of course, the O2 sensors see that and adjust it back to where it supposed to be after a drive cycle or two, eliminating the effect and burdening the system overall. Since I don't have a flame suit, I won't mention the oiled air filter / contaminated airflow meter elements and their resultant effect on performance, efficiency or economy.
:flamer:

Again good post TTA,
J
 
Going back to Mike's 29 highway mpg, I believe the reason is no different than a bonestock MM running a 15.3 quarter mile and another bonestock MM running a 14.9; same track, same day, side by side.

The few times that I had an opportunity to run highway from gas station to gas station my car did 26.3 mpg. N/A.
With the blower, four adults and travel baggage coming home from the Shootout after stopping twice to shop and once to pee, my car got 23.6 mpg.

I think the differences in mpg are more related to the car as an individual and the driver's habits.
 
Oh my gosh, what an awesome read! Thanks for the link!

Dumb question: what are we on our MMs? EEC-IV or V?
 
Oh my gosh, this is SO well written!!!!!!! Seriously, this link has to be sticky'ed somewhere!
 
TripleTransAm said:
Dumb question: what are we on our MMs? EEC-IV or V?
Click to expand...
Good question, /Steve, thanks.

I've been calling this thing an "eek" since I first Jerry W. call it that. The correct term for the Marauder is EEC-VI, and for all Marauders everywhere.

"PCM" is dated. PCMs do not have our adaptive capabilities.

"Intel inside", eh?
 
Last edited:
SergntMac said:
"PCM" is dated. PCMs do not have our adaptive capabilities.
Click to expand...

Actually, all engine-control computers (whether you call them an ECM or PCM or whatever) have adaptive capabilities, as described in that link you provided. It's just that the older ones (like what's in my GTA) have a much lower resolution, no doubt due to the limited processor performance of the early-mid 80s when those ECMs were designed. Lately, these computers control so much more than just engine (for example, transmission, etc.) that they are labelled Powertrain Control Modules. EEC is just Ford's brand name for their ECM/PCM (I think I just recently read an old Motor Trend or R&T from the late 70s where they described the "new" EEC).

In my GTA, there are 16 "cells" arranged 4x4, where one axis is load and the other is RPM, I believe. MAF (no MAP) and other parameters determine load, RPM feeds the next axis, and you arrive at a certain cell. In that cell there are long term ("BLM") and short term ("Integrator") adjustments, with values of 128 indicating "in the middle" and extremities at 0 and 255 (remember, good old 8-bit computing!). Floor it, and the ECM looks at how things were behaving in the cell immediately before going WOT. Each cell has it's set of adjusters, and in reading them I can see that my 5.0l TPI has a tendency to want to run lean at idle or just off idle (because the ECM has adjusters in those cells that are richening the mixture quite a bit). However, the other cells are bang on at 128 or thereabouts. Vacuum leak at idle? Dirty injectors not providing adequate fuel at the 'normal' pulse-widths? No idea. Might actually find out this winter when I tear into it to replace all gaskets in the intake path and maybe *maybe* ship out the injectors to get cleaned/balance-tested.

I don't know the exact details of OBD-II but for sure it ain't just 4x4 cells, from what I've heard. MUCH more resolution. But the adaptive concept remains.

Useless ECM trivia: the 1985 Tuned Port Injected GM F-bodies had a one-year-only ECM model, that was so weak and slow that you had to choose a low-speed data reading speed when connecting a scan s/w to the ECM and driving the car normally, which provided a pitiful refresh rate. Switch to the normal (high) speed data rate, and the ECM would be too overloaded to provide adequate engine protection - you risked inducing deadly knock and other nastiness. 1986 and onwards brought a new and improved ECM with no such limitations.
 
OK, I think I'm getting the picture

This has been one hell of an eye opening discussion, and has gone far beyond what I had hoped, thanks to Mac et.al.

Here's what I think we've learned:

1. All MM's are not the same
2. I believe there's a difference in the EEC of the '03's & '04's (mine's an '04)
3. Adding a "tuner" chip or making changes in the EEC most likely will effect fuel consumption.
4. Changing rear end ratio will effect fuel consumption.
5. Installing the Pro-Guard will decrease fuel consumption at highway speeds.

In my case, when traveling on interstate highways in good weather, using the cruise control, and driving ~70 mph, the addition of the Pro-guard increased my fuel economy by 13.46%. That's using the average of 26mpg before and 29.5mpg after installation.

If you have power adders, ie: supercharger, 4:11 gears, and EEC mods, and are getting 20mpg, on the interstate, then added the Pro-guard, the 13.46% increase would net you 2.7mpg or, ~23mpg. +,-.

Bottom line: Do I think my reports of 29+ mpg on the interstate with the pro-Guard is unusually high? No I don't.

I believe anyone who has a completely stock '04 running synthetic lubricants and uses the cruise control should expect nearly the same results.

If you've added gears, and a chip, and are running around town with your foot in it, most of the time, the Pro-guard isn't gonna do much for you, except keep you from tearing your bumper off when you run over a curb.

If, however, you do a significant portion of your driving at highway speeds, with or without power adders, I think the Pro-Guard will net the 13.46% increase I've experienced.

MATH: Subtract the first number from the second number: 29.5 - 26 = 3.5

Divide the answer by the first number: 3.5 / 26 = .1346 or 13.46%

I've really appreciated the tone and thoughtfulness of the participants in this discussion.
 
Last edited:
:confused: i still drive my basically unmodded MM on 100 mile commutes everyday (highway miles basically no stop and go) and minimal city driving. 19-20 MPG

:confused: i have stock gears no tuner and I run mobil1. so i dont see how i'd get 9 mpg less than another basically stock 04 running mainly highway miles. dunno
 
hdwrench said:
:confused: i still drive my basically unmodded MM on 100 mile commutes everyday (highway miles basically no stop and go) and minimal city driving. 19-20 MPG

:confused: i have stock gears no tuner and I run mobil1. so i dont see how i'd get 9 mpg less than another basically stock 04 running mainly highway miles. dunno
Click to expand...

I'm doing about 23 under those same conditions.

What I'm talking about is topping off the tank, entering an interstate, driving 400 miles, with the cruise control on, filling up, and doing it all over again, in optimum weather conditions. And what I think we've been saying, is that the adaptive function of the EEC is gonna net those results in an unmodified '04.
 
You must log in or register to reply here.
Back
Top