Fooling the Computer for Better MPG

[bobinyelm]

Given the weight, engine size, and the gearing of our Cherokees, the fuel economy generally SUCKS.

I've put up with it to date in my '01 XZ w/4.0, AW4, 3.55 dif, but with gas $3.25 a gallon, I am tempted to actually do something about it.

Some folks seem to get up to 25mpg highway (conservative driving) w/ their XJs, while others are fortunate to get 18-20. I drive like the proverbial old fart and only manage 20mpg around town or on the highway (60mph cruise), but a vehicle w/ Cherokee's stats (3100 pound empty weight, 3.55 gearing w/ OD and locking torque converter) SHOULD do 22-24mpg easily highway.

4mpg doesn't sound like a big deal, but over 20,000mi/yr it's OVER $500 in fuel savings.

I have a '00 TJ (4.0, 5spd, 3.73 gears) that I replaced the O2 sensor in (because every few months it would throw an O2 code), and BEFORE replacement, it ran GREAT and got 25mpg. With a shiny new O2 sensor, driving the SAME way, I can only get 20mpg.

I have a feeling that given the better gearing and aerodynamics of Cherokee over Wrangler, 25mpg is a realistic goal in very conservative driving, and think that if that "bad" O2 could get 25 in the TJ, a MODIFIED O2 circuit or MAF circuit can do the same in the XJ.

Getting 25mpg didn't hurt a thing in my TJ, so I see no reason "tuning" an XJ to do the same could hurt anything in an XJ.

Anyone done or familiar w/ a way to "fool" our computers to lean the mixture slightly? Maybe an Op AMP to increase the O2 output slightly so that when the exhaust actually produces (say) 0.45v (which is the nominal target voltage), the circuit actually SEES (say) 0.50v so the computer will further lean until the computer sees 0.45v (which would mean the ACTUAL O2 output is 0.40v, or leaner than target)?

I have heard of tunable MAFs, or fuel pressure regulators, but given the authority the O2 has over the entire operation, I wouldn't think a modified MAF or fuel pressure would change much (since the O2 circuit would just change the injector pulse width to compensate.


I don't want to eliminate the closed-loop fuel control-I want to MODIFY it so I can perhaps select (with a potentiometer) a certain mixture target and have the computer use THAT value (rather than 0.45v or 14.7 air/fuel ratio). This would give me the option of returning to a richer mixture if I were towing, or a leaner one when it might be appropriate. I naturally have no desire to burn a piston or a valve to save a few bucks on fuel.

Bob

 

[Blue XJ]

If your getting 20mpg around town, your lucky. I don't think you would really get any more around town considering you are driving a box with wheels.

 

[jcshaney]

all the people i talk to with tj's seem to get 14-16 my xy gets 20-22 av.
98xj 4.0 31" allterains ax15 3.07's 160,000 k

 

[srimes]

Would a circuit like what's used on the MAF work w/ the O2? Bump the imput voltage so it comes out higher?

 

[8Mud]

IMO it could be risky. Some places in the power band, the XJ is about as lean as you want any motor to be. Just short of detonation.
About the best thing you can do is keep it well tuned and make sure your sensor suite is functioning properly.
Some small things you can do that actually add up though. I got noticeably better mileage with 6 ply truck tires. Though the ride was harsh. Keeping unnecessary weight out of the back. Turning off your motor during extended stops. Acting like there is an egg between your foot and the gas pedal. Driving way ahead so you have to brake as little as possible.
Since the last gas spike I consolidate my errands (radically). Which has actually been my largest increase in savings. I don't run off to buy something now, I make a list and do it all in one trip. What use to be every other day shopping trips, is now once a week.

 

[bobinyelm]

IMO it could be risky. Some places in the power band, the XJ is about as lean as you want any motor to be. Just short of detonation.
About the best thing you can do is keep it well tuned and make sure your sensor suite is functioning properly.

Hell, LIFE is risky. Risky was when I was flying fighters in USAF in the Vietnam War. Burning a valve is pretty low on the "risky" scale, and with an educated butt, and ear, one can tell when it's THAT lean, anyway.


As I mentioned, my TJ (which is heavier and more "boxy") regularly got 25mpg and had zero problems (detonation, lean hesitation, surge, etc.) so I really don't think a slight leaning bias when not "pushing" would hurt anything. I live in FLAT Texas, and already drive like I have an egg under my throttle foot.


I would WANT to make whatever mod I did adjustable, so if I NEEDED to use more of the power available, I could dial the mixture BACK to where it was for the very reasons you state.

Some small things you can do that actually add up though. I got noticeably better mileage with 6 ply truck tires. Though the ride was harsh. Keeping unnecessary weight out of the back. Turning off your motor during extended stops. Acting like there is an egg between your foot and the gas pedal. Driving way ahead so you have to brake as little as possible.
Since the last gas spike I consolidate my errands (radically). Which has actually been my largest increase in savings. I don't run off to buy something now, I make a list and do it all in one trip. What use to be every other day shopping trips, is now once a week.

As I also mentioned, I already drive like an old fart (as I progress toward geezer-hood), and do all of what you suggest (except stutting the engine down at lights because the potential gas saved is cheaper than a starter motors, though I do shift to neutral to lessen fuel flow), and I keep the vehicle as light as possible w/ the tires inflated to 32psi.


I just think a tad more could be squeezed from a Cherokee with some judicious juggling of computer parameters for 75% of the driving most of us do. The electronic progrqamming is an "average" for all vehicles in mixed service that won't cause mechanical problems, and that will pass emmissions. All I am looking to do is to optimize a bit for specific conditions.


Even if one could get 15% more MPG, yet had to run mid-grade gas at 3% more cost/gallon to avoid detonation, he'd be ahead, for instance.


Bob

I know it's "apples and oranges," but I have a Mercedes Sprinter van (the long/tall one like FedEx and UPS has) and even loaded to 8500 pounds and as boxy as it is I get 27mpg in town OR on the highway. I just sold a 4WD Surburban w/ a 350cu in and a carb that regularly did 19mpg on the highway (16 around town) and weighed 5500 pounds.

24mpg at a constant 60mph (1800rpm in OD) on level ground is not asking too much from a lightweight Cherokee IMHO.

 

[srimes]

Another thing about messing with the o2 sensor is it won't affect WOT (or near it) as the computer goes open loop, so it's not super risky.

 

[badron]

Very doable. And if you will to do Lots of work. It can be done safely. Be willing to pull plugs at the drop of a hat. Expect to open up plug gap. Then expect to change them more often and/or increase the voltage to then. MAY need to run a cooler plugs to. Keep a quit muffler on so you may detect nock before it to late. Watch for NOs to go through the roof. Without a EGR value it may be hard to compensate for high NOs. The types of cat on the 01 would be over kill for a lean burn system but should work.
Adjusting the timing will not be cheap or easy unless your a computer nerd or get some 1000 buck chip.
The later model smog system of the XJ is much more of a balancing act then many. With no MAF, EGR value or knock sensor some compromises were made. MPG was one of then. But hay they saved maybe 30 to 50 bucks per engine.

 

[xjtrailrider]

I noticed my best around town increase when I went to 4.10 gears on 30" tires. I went from 15 to 18 and my highway mileage is close to the same as long as I keep it under 65.

The 3.55 gear is just all wrong for the XJ, it should be at least 3.73 from the factory and 4.10's make for a fun peppy machine that accelerate's effortlessly. My TC lockup now occurs at 30 mph and it will hold down to 27 mph before it kicks back. When it was stock TC lock-up occured at 45 mph and would kick back around 40. Since most in-town driving is around 35 my XJ stays in lock up much more, which in my opinion saves fuel. I don't suggest everyone go out and put new gears in thier Jeep but it helped my 96' out.

Just my $.02

 

[bobinyelm]

Very doable. And if you will to do Lots of work. It can be done safely. Be willing to pull plugs at the drop of a hat. Expect to open up plug gap. Then expect to change them more often and/or increase the voltage to then. MAY need to run a cooler plugs to. Keep a quit muffler on so you may detect nock before it to late. Watch for NOs to go through the roof. Without a EGR value it may be hard to compensate for high NOs. The types of cat on the 01 would be over kill for a lean burn system but should work.

I'm not talking a crazy change, only a slight amount. After all, my TJ got 25mpg w/ the old O2 sensor and it ran fine, no burned plugs, no knock, and passed emmisions just fine. Put a new O2 sensor in it and the mileage dropped immediately to 20 on the very next (and each following) tank. If I hadn't thrown it out, I'd try it on the Cherokee to see if it helped!

Adjusting the timing will not be cheap or easy unless your a computer nerd or get some 1000 buck chip.
The later model smog system of the XJ is much more of a balancing act then many. With no MAF, EGR value or knock sensor some compromises were made. MPG was one of then. But hay they saved maybe 30 to 50 bucks per engine.

Yeah, timing change is a no-go for me.

Unfortunate that they sacrificed fuel economy as they apparently did in the XJ. Hell, I used to routinely get 22+mpg 36 years ago w/ a very "draggy" CJ5 and the Dauntless V-6 they put in 'em back then, and THAT was at 70-75mph when I was young and crazy.

Almost 40 years later, with better technology and we can't do better than 20mpg? Gimme' a break!

I was hoping someone would have an analog operational amplifier circuit that would essentially offer variable gain using the O2 signal as the input, with the output being fed to the ECU. Raising O2 output voltage to the computer would trick it into thinking it should lean the mixture slightly to get to the "magic" 0.45 volt level.

Using a potentimeter, one could mount it in the dash and mark it for best economy, best power, or back to factory setting at will. The computer would be fat, dumb, and happy.

It's been years (37 to be exact) since I designed digital and analog integrated circuitry, but it seems worth a go.

Bob

 

[bobinyelm]

I noticed my best around town increase when I went to 4.10 gears on 30" tires. I went from 15 to 18 and my highway mileage is close to the same as long as I keep it under 65.

The 3.55 gear is just all wrong for the XJ, it should be at least 3.73 from the factory and 4.10's make for a fun peppy machine that accelerate's effortlessly. My TC lockup now occurs at 30 mph and it will hold down to 27 mph before it kicks back. When it was stock TC lock-up occured at 45 mph and would kick back around 40. Since most in-town driving is around 35 my XJ stays in lock up much more, which in my opinion saves fuel. I don't suggest everyone go out and put new gears in thier Jeep but it helped my 96' out.

Just my $.02

The older XJs (up to 91 or so I think) had a switch that allowed earlier shifting and lockup under light throttle.

Yep, TC lockup is right at 42mph under very light throttle w/ mine, which means that around town, it's seldom locked up, since everyone goes about 35mph. Below 42mph I try to get it into 4th, then back off to keep the TC losses low (the lighter the throttle, the less the losses). I can hold the same constant speed (35-38mph) at 1400rpm or 1700rpm with slight throttle pressure differences, and obviously it's using more fuel at 1700.

That's how I get 20mpg around town (suburban, not city driving), but on the highway, once you're in 4th and the TC is locked up, you're pretty much going to get what you're going to get mpg-wise unless you are rough on the throttle. Getting the same mpg at 55/60 as in town just seemes messed up. The aerodynamic drag at 55 shouldn't overcome the efficiency of being locked up in OD.

Bob

 

[Sig220]

I found this thread interesting in a few ways.

Bob, if you figure out how to fool the ecm and get bettere mileage you will have a flock of followers!!

I have had a 84 Cherokee with the 4cy/auto/4wd, that got good mileage, I had a 87 Cherokee with the 6cy/auto/4wd and even though it burned a little oil would get 23 on the Hwy.....easily. Of course, I had to rebuild the damn thing and the mileage fell off to around 17 after that

I have now a 92 Cherokee with the 6cy/auto/4wd that I saved from a junkyard, as long as I keep it undere 2000rpm on the highway it gets right at 20mph......I don't have it on the highway much...as I use it for my hunting lease vehicle...in those conditions, the gas mileage is low....understandably.

My son has a 96 Cherokee with the 6cy/auto/2wd and I would bet his gas mileage on the highway is around 18.....should be better.

I guess what I am trying to say is that all of them are different in how they were broke in/used before we got them and what "wear" conditions they are in. I, too think they should be getting 20-23 in highway driving conditions without a problem, just don't have the time to figure out how to get there!!!

Good Luck, and keep us posted!!

 

[bobinyelm]

I found this thread interesting in a few ways.

Bob, if you figure out how to fool the ecm and get bettere mileage you will have a flock of followers!!

LOL. If I can stay motivated long enough in the face of so many more important tasks bagging for my time, I should be able to come up with something. The ECU sowed the seeds of its hacking by listening to the O2 sensor as its sole advisor. If I can put together a non-inverting Op Amp with a small gain, or a Summing Amplifier with a small add voltage, it should be relatively easy, though how the 4.0 will run when leaned a bit I cannot vouch (except to cite my TJ's 25mpg).

I was frankly hoping that someone already went down this path and had a ready solution.

I have had a 84 Cherokee with the 4cy/auto/4wd, that got good mileage, I had a 87 Cherokee with the 6cy/auto/4wd and even though it burned a little oil would get 23 on the Hwy.....easily. Of course, I had to rebuild the damn thing and the mileage fell off to around 17 after that

I have now a 92 Cherokee with the 6cy/auto/4wd that I saved from a junkyard, as long as I keep it undere 2000rpm on the highway it gets right at 20mph......I don't have it on the highway much...as I use it for my hunting lease vehicle...in those conditions, the gas mileage is low....understandably.

My son has a 96 Cherokee with the 6cy/auto/2wd and I would bet his gas mileage on the highway is around 18.....should be better.

I guess what I am trying to say is that all of them are different in how they were broke in/used before we got them and what "wear" conditions they are in. I, too think they should be getting 20-23 in highway driving conditions without a problem, just don't have the time to figure out how to get there!!!

Good Luck, and keep us posted!!

Internal friction differences alone between engines shouldn't make THAT large a difference (28%) in mileage. Something else is surely at work, but figuring it out is another matter. I wouldn't have guessed that replacing my working O2 sensor (in my TJ) with a new one would have killed 5mpg. Something is AWFULLY sensitive.

Everyone comments that the 4.0 is set to run very lean, but both of mine put plenty of soot in the exaust pipe, and neither uses appreciable oil to explain it. Extra fuel is being burned and not being turned into horsepower.

Bob

 

[xjtrailrider]

I think that fooling the RENIX computer would be easier and less fustrating.

OBD1/2 is going to get complicated, this is one of the reasons I fooled mine with gearing.

 

[bobinyelm]

I agree (RE: Renix), but with OBDII, about the only inputs with unmonitored authority to change fuel trim are the O2 sensor and the air temp sensor. OBDII SEEMS imposing, but if you know what it's looking at, I suspect you can do more than you think.

The O2 sensor IS monitored with regard to response (rapid cycling) and absolute fuel "trim" from preloaded standards, but I suspect staying w/in 10-15% of those parameters won't tip off the ECU police that the signal is being doctored.

Going slightly lower in gearing would actually suit me, because I want to go to 235/75-15 tires and to be able to tow my small (3500#) travel trailer, but the expense would probably never be recouped in actual fuel savings.

Building a variable gain, or summing OP Amp from Radio Shack parts, however, would cost maybe $45 tops, so has a greater pay-back factor, if it works (and if I get off my ass to put it together).

Bob

 

[h2opete987]

New to the forums, but this seems like a great place for me to start.
I am curently working towards my Electrical Engineering major, so have access to the toys and knowledge that will make this porject possible. I have the same problem with efficiency. i consistantly pull 20mpg highway out of my '91 sport, but the engine STILL runs so rich you can smell it. Currently i have a prototype circuit designed, and preparing for constuction and testing. I will post more as i progress through the project.

 

[xjtrailrider]

I agree (RE: Renix), but with OBDII, about the only inputs with unmonitored authority to change fuel trim are the O2 sensor and the air temp sensor.

There is a write up on here somewhere on moving the MAT sensor from the manifold to the air box/air tube. My 90' RENIX has a bad MAT(always reads cold) sensor but they are no longer available. I may have to adapt one from a later model and possibly move it. The disfunctional MAT is giving me poor mileage, but plenty of power!

Going slightly lower in gearing would actually suit me, because I want to go to 235/75-15 tires and to be able to tow my small (3500#) travel trailer, but the expense would probably never be recouped in actual fuel savings.

Towing with my Jeep is what started this whole thing with me, I tow a 3100# boat with my 96' and since I'm a mechanic and set up diffs the cost was minimal for me. I realized very soon after the gear changes that the benefit of having lower gears on near stock sized tires was far better for daily driving than it was for towing. It still tows about the same, just a little better acceleration on the mountains around here. Like I posted before, in town mileage gains were significant. Highway mileage is about the same and that is running 2800 rpm in OD at 65 mph. I think the 4.0 likes to breath more than most people think.

 

[Ecomike]

Who told you MAT sensors for Renix were no longer available? I bought one fron Autozone just about 12 months ago. Stock item.

There is a write up on here somewhere on moving the MAT sensor from the manifold to the air box/air tube. My 90' RENIX has a bad MAT(always reads cold) sensor but they are no longer available. I may have to adapt one from a later model and possibly move it. The disfunctional MAT is giving me poor mileage, but plenty of power!

Towing with my Jeep is what started this whole thing with me, I tow a 3100# boat with my 96' and since I'm a mechanic and set up diffs the cost was minimal for me. I realized very soon after the gear changes that the benefit of having lower gears on near stock sized tires was far better for daily driving than it was for towing. It still tows about the same, just a little better acceleration on the mountains around here. Like I posted before, in town mileage gains were significant. Highway mileage is about the same and that is running 2800 rpm in OD at 65 mph. I think the 4.0 likes to breath more than most people think.

 

[Ecomike]

New to the forums, but this seems like a great place for me to start.
I am curently working towards my Electrical Engineering major, so have access to the toys and knowledge that will make this porject possible. I have the same problem with efficiency. i consistantly pull 20mpg highway out of my '91 sport, but the engine STILL runs so rich you can smell it. Currently i have a prototype circuit designed, and preparing for constuction and testing. I will post more as i progress through the project.

If you are getting 20 mpg, and smell gas out the exhaust you need a new Catalytic converter! I am lucky to get 13 MPG avg, about 9 mpg stop and go and maybe 16 highway and don't smell any gas out my exhaust. If you smell it under the hood, check for leaking injectors, Jeep OEM fuel injectors are bad about external leaking, and not from the o'rings, but from the case itself.

Sounds like an interesting project. Keep in mind that the Renix O2 sensors are different. The ECU supplies 5 volts to the O2 sensor and the O2 sensor acts as a variable resistor. The post 90 O2 sensors produce a 0-1 volt signal that is proportional to the O2 concentration, so matching the correct current/voltage and impedance matching may be real tricky with them.

Have you considered that when the old O2 sensor went out, that it might have had a contributing factor that helped throw the code like a poor ground in the circuit that is keeping the new one from performing at optimal efficiency?

 

[badron]

At what point you TC locks MAY have little to do with good MPG, it MAY have to do with good smog numbers.
My 5 speed (with 3.07) worked well in 5th gear at 30/35 MPH but no real increase in MPG happened until about 45 or later. Smog numbers may have dropped at 35.
A learn burn system is generally one that run at or close to an ideal air/fuel ratio. Some engines run fine all day at 17 or even 18 to one with no problems. I hear of numbers at 20 to 1 but that may be bar talk to.
With a better over all smog/fuel/timing system the XJ would get more MPG AND meet EPG smog spec. But don't hold it on 25.
If you want to out smart the factory you got to understand why they did it the way they did. That ant hard really, First COST then COST then smog then MPG and/or horse power. Think this way and the XJ systems start making since.