So here is my question for your consideration:

How much MPG improvement will you get with a new air filter? The commercials say "up to 10%" The Gov't says the same thing. I heard another ad say "up to 2 MPG".

I have thought about it and I think that in a fuel injected car, the savings will be minimal, unless you have been driving off road a lot or somewhere really dusty and the filter is clogged beyond normal levels.

I have not changed my filter in about 3 years and I have a new one ready to put in. I have a very accurate trip computer (ScanGuage II) that I have had for a while and I have it dialed in, so I can run a scientific test comparing old filter to new filter.

I am looking for ideas on what "scientific" tests to run and what results you expect me to find. I have some ideas that I have discussed with some engineers. Does anyone have any thoughts on this?

__________________
1998 C230 330,000 miles (currently dead of second failed EIS, yours will fail too, turning you into the dealer's personal human cash machine)
1988 F150 144,000 miles (leaks all the colors of the rainbow)
Previous stars: 1981 Brava 210,000 miles, 1978 128 150,000 miles, 1977 B200 Van 175,000 miles, 1972 Vega (great, if rusty, car), 1972 Celica, 1986.5 Supra

A clean air filter isn't going to save you any fuel.

I, too, have been thinking on this subject.
In the days of carburated engines, a dirty air filter could affect mileage to a considerable extent.
However, on the milti-point fuel injection systems, with Oxygen feed back from the exhaust stream, I do not see how the air filter can have much effect. The computer monitors to the exhaust and varies the mixture almost instantaneously to obtain the stociometrically correct ratio. If the intake is restricted ( dirty air filter), the computer will only allow enough fuel to burn the available air.
A dirty air filter may restrict the available power, but I see only a minimal impact on economy.

__________________
1982 300SD " Wotan" ..On the road as of Jan 8, 2007 with Historic Tags

Think pumping losses. Any restriction in the airflow is going to increase pumping losses. Grab your bycycle pump. Lay the hose loose on the ground and start pumping. Pretty easy, huh? Now connect it to a tire. Start pumping. A bit harder? Same amount of air is moving, but it takes more work to move that air. (yeah, I know, presure differential vs. simple reduced orifice but most folks don't have the stuff laying around to demonstrate it exactly). Same thing if you are moving a bunch of air through a restricted air filter. Work performed = fuel burned.

It wouldn't be anywhere near as bad as what already happens with the throttle plate.

A clogged air filter on a EFI engine actually means better MPG since there is less airflow which means less fuel injected by the computer. A carb engine would have lower MPG since it would result in a rich mixture, same with the mechanical Diesels.

That analysis works on cars with carburettors or mechanical fuel injection, but not on cars with computer-controlled engines. The computer offsets any reduction in airflow caused by the air filter with a corresponding reduction in fuel.

If the restriction caused by the air filter results in insufficient power for the driving conditions, the driver simply applies more throttle, thereby reducing the flow restriction. The low restriction air filter allows more airflow at full throttle, but otherwise I don't think it makes any difference on a computer-controlled, fuel-injected engine.

I heard that ad as well and was leery of the claims.

Since it's a MythBusters thread, I've wondered at what point you're better off shutting a vehicle off vs. letting it idle. Does a vehicle use any more fuel at start-up than it would idling for a period of time (like the fluorescent lightbulb theory of using more energy at start-up than leaving it on-which was busted unless you we're only returning to the room about 18 seconds later or something) What's the point where fuel savings comes into play? 3 min.? 5 min.? I'm curious as to whether it could make a difference with our drivers/salepeople when making multiple 1-3 case deliveries over the course of the day.

__________________

1980 300TD-China Blue/Blue MBTex-2nd Owner, 107K (Alt Blau) OBK #15
'06 Chevy Tahoe Z71 (for the wife & 4 kids, current mule) '03 Honda Odyssey (son #1's ride, reluctantly) '99 GMC Suburban (255K+ miles, semi-retired mule) 21' SeaRay Seville (summer escape pod)

Up to 2 mpg. Of course that includes no increase.

Tom W

__________________
[SIGPIC] Diesel loving autocrossing grandpa Architect. 08 Dodge 3/4 ton with Cummins & six speed; I have had about 35 benzes. I have a 39 Studebaker Coupe Express pickup in which I have had installed a 617 turbo and a five speed manual.[SIGPIC]

..I also have a 427 Cobra replica with an aluminum chassis.

The computer can do whatever it wants to adjust the mixture to meet emissions and power demand. It is still a mechanical loss. The engine has to work harder to produce the same rear wheel horsepower. Granted at lower RPM the losses are less and they increase with increased throttle opening. Still comes down to pulling air through a soda straw vs. pulling air through a paper towel tube. If you need to fill your lungs, which will be easier on you? And the thing with pumping losses (and all mechanical losses) is that they are there all the time- like rolling resistance from under-inflated tires. Constantly costing you a little bit more each mile you drive. If you don't think it is a problem, why do you EVER change your air filter.

Thank you for the interesting replies! We are all thinking the same way generally, but I now have come to a new conclusion. I will try to put all my responses together here:

*One disclaimer: Since we are getting into the issue of the throttle plate, this discussion (currently) will be limited to gasoline cars, since I don't think diesels have a throttle plate. In fact, I have heard that this fact is why Diesels are more efficient than gasoline engines, so maybe diesels would benefit from a cleaner air filter...

ForcedInduction: I think you are correct, there will be NO difference. You make an excellent point about the throttle plate, which I will get back to in a minute.

MS Fowler: Yes I agree on the carb. However the gov't and advertisers are pretty far behind if they are talking about carbs. I think the last car sold with a carb was the '89 Subaru Justy, nearly 20 years ago.

BobK: This is exactly what I WAS thinking, that there would be (small) losses due to the "sucking through the soda straw" idea, pumping losses.

dculkin: I think you are where I am now but saying it a little different than I will say it.

Swampyankee: Idleing losses are very significant. There is no "startup loss" in a warmed up fuel injection engine. My car when fully warmed up uses .33 GPH at idle. Even after a 30 min trip on the highway I can watch my trip MPG melt away as I idle at a stop light. But you have to weigh the savings against the wear and tear on the starter.

Here is my conclusion: I thought there would be "pumping losses" as Bobk describes. However ForcedInduction got me thinking about the throttle plate. The throttle plate acts as a resistance in series with the filter. The engine cannot tell one from the other, it just sees a total resistance. The throttle opens until the resistance is such that the amount of air you "request" (with the gas pedal) goes into the engine. If the filter is more clogged, then the throttle will open more, but the total resistance WILL BE THE SAME. The engine will have no way of "knowing" the difference. There will be NO IMPROVEMENT in MPG. The only effect of a dirty filter would be a limit on maximum possible horsepower. (and possibly letting dirt in I suppose), but NO MPG improvement as advertised.

Now since we probably don't all agree, and I myself am not POSITIVE, I propose the following tests:

1) At idle, with engine fully warmed up after a long trip I will record:

Engine temp
RPM
GPM fuel flow rate
HP

I will do this 10 different days with the old filter, then again with the new filter.

2) I will run a measured course on the highway at a fixed speed, measuring the total fuel consumed over a specific trip segment. I will do this with the old filter and then with the new, both on the same day when the air temp is stable and there is no rain.

Comments?

__________________
1998 C230 330,000 miles (currently dead of second failed EIS, yours will fail too, turning you into the dealer's personal human cash machine)
1988 F150 144,000 miles (leaks all the colors of the rainbow)
Previous stars: 1981 Brava 210,000 miles, 1978 128 150,000 miles, 1977 B200 Van 175,000 miles, 1972 Vega (great, if rusty, car), 1972 Celica, 1986.5 Supra

Not always true. Put an airflow meter up on a scope and restrict airflow. The percieved airflow increases. i dont know if all airflow meters do this but bosch meters do and so do the ones GM used in the mid 90's. The engine computer would see a greater airflow signal with a restricted air filter. Once on a volvo turbo I worked on the computer would limit turbo boost and the car would not produce much power. Did a little reading and turns out the computer doesnt look at manifold pressure just airflow. ShoNuff the airfilter was plugged. Put in a new one and lit up the tires.

Not true. As an air filter loads up with dirt, the airflow, and therefore the power output, decreases. To make up for the decrease in power, the driver opens the throttle some more, which reduces the airflow restriction. The engine doesn't know the difference between a dirty air filter with a wider open throttle and a clean air filter with a less wide open throttle. The only time it makes a difference, at least in theory, is at full throttle (subject to the peculiarities cited by omegaman)I'm not sure, but I think that a clean air filter might do a better job of protecting the engine. Also a clean air filter provides more power at full throttle.

Wouldn't that support my theory? With a restrictive air filter, the vacuum between the air filter and the throttle becomes stronger, but the pressure in the manifold should stay the same for a given air flow, shouldn't it? I dunno.Do you know whether it also performed differently at part throttle?

That's what I was trying to say.

The engine itself wont know the difference but the computer will. The airfilter is before the mass airflow meter and they read higher when the air filter is restricted.