Leaning up the fuel mixture.

[alleninpa]

From what I have read, our little S series run rich on the fuel, by design. I was thinking of modifying the air temp sensor to cheat the system into thinking the air was slightly warmer than what it really is. This should lean up the mixture, resulting in better fuel economy. Has anyone tried this?
If you have, do you use a resistor in parallel or series, and what value do you use?

Thanks!
Al

[BarnOwl]

I wouldn't do this. If the mixture is as designed, it's not running rich for this engine. It might be rich for another engine, but not this one. Others may think differently.

[amazinghl]

It's running RICH at WOT, which is normal.

[Travelair]

Does your state test emissons? I can't remember the category, but leaning causes one parameter to be exceeded. Hydrocarbons I think. The EGR system compensates for the rich mixture somewhat.

[danielni]

At part throttle, the system is in closed loop with the O2 sensor. So it would be useless at least. At worst, the ECM would be unable to go closed loop, latching an error code.

[justin_zink]

Originally Posted by alleninpa

resulting in better fuel economy.

Why would you need better fuel economy, all the s series cars get excelent milage, still today only really eclipsed by the hybrids, what gain you would get from the engine would be minimal to the damage that you might cause if it goes to far.

[Travelair]

Originally Posted by justin_zink

Why would you need better fuel economy, all the s series cars get excelent milage, still today only really eclipsed by the hybrids, what gain you would get from the engine would be minimal to the damage that you might cause if it goes to far.

Very true. And if you lean it out too much it won't run at all. Yesterday I got 35 mpg. 38 is not uncommon for me in the winter. I keep the car in tune, the oil fresh and the tires over-filled to 35psi. All of my driving is highway.

[Kevin94SL]

42 mpg in the '02. I don't think that much out there can get this. That is with about 10% city driving and the rest highway.

Really, is there a point to leaning out the mixture. Obviously there is reason that engineers spec'd these cars this way right???

[alleninpa]

The engineers designed the car to practically run forever. The 4K in neutral max, top speed limiters, reverse limiters, ect. They also have it running rich to keep the engine temp down, which, if you are in stop and go traffic in 100 degree weather is good. However, that is not the situation that I drive in, and the fuel mileage heads south in the winter. Therefore, with minor changes, it should be able to sneak another few MPG out of it, with no side effects.

[BarnOwl]

Okay, I've never done this and I don't think it's a good idea, but in closed loop the PCM is paying attention to the oxygen sensor to determine how lean or rich it is running. The voltage is between 100 mV lean and 900 mV rich. So, in theory, if you could get a little more voltage to come in from the O2 sensor, that would make you run leaner. I'm not sure how you'd go about implementing this though and I'm pretty sure that unless you were real careful about it, the PCM would notice and quit paying attention to the O2 sensor.

[Tom92SCm]

A decrease in fuel pressure has shown to improve power at WOT. Therefore I would say that if you have less pressure, you have less fuel flow and therefore better MPG.

Exactly what are your goals by improving your MPG? Are you indicating that you use your car in a manor other than perscribed?

--Tom

[S3aturnR]

Originally Posted by alleninpa

From what I have read, our little S series run rich on the fuel, by design. I was thinking of modifying the air temp sensor to cheat the system into thinking the air was slightly warmer than what it really is. This should lean up the mixture, resulting in better fuel economy. Has anyone tried this?
If you have, do you use a resistor in parallel or series, and what value do you use?

Thanks!
Al

probably not a good idea. the saturn's little brain is programmed to run as close to stoic as possible, based on its output as read from an o2 sensor. wot will run a pre-designated fuel map (open-loop), but that's not where you want to run lean anyway...


s3aturnr

[diamondlarry]

Hey Allen,
I have tried this on my 1997 SL. I have been running this way since April with no noticeable effects. What I did is unplug the the plug from the IAT and stick a combination of resistors that total 110 ohms in the holes in the plug. I recently did a test to find out how much difference it made and I was a bit surprised at the difference. I made two 10 mile runs with and with out the mod. Stock gave me an average of 44.4 and with the mod I got 56.8. Keep in mind these figures are for a 10 mile run with no traffic and only one turn-around. Also, there is a device at www.eagle-research.com called an EFIE that tricks the ECU into thinking the car is running rich. It intercepts the O2 sensor signal and modifies it so it looks like it is running rich. Of course, you don't want to add too much voltage or you will turn on the "Check engine" light.
Larry

[BarnOwl]

Interesting site...That EFIE appears to do exactly what I was suggesting a few posts back and I was just theorizing. Let me know how it turns out, alleninpa.

[Sky King]

I have found this discussion quite interesting. My first concern however with attempting to lean out the fuel mixture goes towards possible major damage. With no actual way to measure the mixture, exhaust gas temps and other parameters tied to mixture, I would proceed with caution. Go too lean and you can start detonation. This in turn can eventually burn pistons. Keep us posted with your progress and results.

[amazinghl]

With all the gas money they save, I'm sure they can afford new engines later.

If you're serious with this, I STRONGLY suggest EGT guage and wideband o2 sensor so you know how your engine is doing.

Forgo the stock ECU and go megasquirt.

[dave.sl]

I did this 35 years ago on a 62 MG Midget and burned out the valves in two months. That car of course was pre computer and had SU carbs but it damaged the engine in no time. Luckily, a quick valve job on a Saturday and resetting the carbs to factory saved the rest of the engine. Someone with the Saturn engineering dept could probably tell you of the potential for engine damage. Maybe the newer cars can take the extra heat but my guess is not. (By the way, for the short time I ran the car like that, I got 44 mpg which was up from the 32 I normally got). That kind of mileage was unheard of back then but it cost me in the end and gas was only 29 cents a gallon.

[stu]

Originally Posted by dave.sl

I did this 35 years ago on a 62 MG Midget and burned out the valves in two months. That car of course was pre computer and had SU carbs but it damaged the engine in no time. Luckily, a quick valve job on a Saturday and resetting the carbs to factory saved the rest of the engine. Someone with the Saturn engineering dept could probably tell you of the potential for engine damage. Maybe the newer cars can take the extra heat but my guess is not. (By the way, for the short time I ran the car like that, I got 44 mpg which was up from the 32 I normally got). That kind of mileage was unheard of back then but it cost me in the end and gas was only 29 cents a gallon.

Dave, you bring back old memories. I had a '62 Midget in high school, until I got run over from behind by a furniture truck on a slippery hill in a snow storm. Its bumber was about ear high on me, but fortunately we slid down the hill together.

FWIT, that midget burned valves up without any modifications. When it was time to do the rings, as well as the head, Dad & I stripped the engine down in frame and lifted it out by hand (1089 cc, no engine hoist). However, it still devoured valves, and Dad kept some grinding compound in the garage.

It was also my introduction to hydraulic clutches, and having the car drive away while the clutch was still depressed. Try holding that one on an incline at a traffic light.

Ah the good old days!

LOL,
stu

[alleninpa]

I didn't try this yet, but I got an aproximate range of the air temp sensor to start with.
It looks like at 32 degrees it is about at 4000 ohms and at 90 degrees it drops to about 2000 ohms. Using thermo laws, assuming air density is linear down to -273f or 0 degrees kelvin, air at 32 degrees f is 13.5% denser than air at 100 degrees f. Therefore, dropping the values aproximately 15% would lean up the mixture by about 2% which is a fairly conservative amount. To decrease the ohms, I am planning to start subtle with the air temp sensor. I am thinking of using a 10K ohm resistor in parallel. This would result in the 4000 ohm value to drop to 3500, and the 2000 ohm to drop to 1666.

[dixonba]

Dont assume the temp sensor is a linear sensor. You should create a calibration curve and extrapolate the results...

That is if you're really serious about it.

One of my friends though he was cool and hooked a potentiometer to his IAT. We would be driving it down the road and he would be adjusting the pot for "more poooower". If anything, it was a good laugh for a saturday night.