Some hypermiling techniques illegal in my state!

[rh77]

Also, if you're into tinkering under the hood, you can bypass the fuel injection electrical system by installing a switch inside the car to manually cut injection. I did this with 2 cyldinders during an experiment a while ago (basically, you re-route the wire that powers the FI to the switch and back to the injector series). Since my car does this already (didn't know it at the time) it was moot to flip the switch on decel. Now, I just play around with N and D to coast and engine-brake where necessary. Only if I had a clutch

RH77

[GasSavers_Erik]

I tried a crude fuel injector cut off earlier this year. I would disconnect one injector on my 87 Acra intergra (1.6 DOHC MPFI) when I would make a long trip (3 hours one way) on the interstate. I would also disconnect the oxygen sensor so the computer wouldn't richen the mixture when it saw all the oxygen coming through the dead cylinder. The engine vibrated a lot under 3000 rpms, but at 65mph (3500 rpm), it felt smooth (although sluggish). I didn't see any noticeable mpg gains- probably because disconnecting the oxygen sensor sent the computer into open loop mode. I tried it on 2 of those long trips (800 miles total) and then abandoned the idea because of the lack of discernable mpg improvement and possibility of increased engine wear.

I think that a selective cutoff during coasting as described above would be a good improvement of the concept and that it coudl be rifined further for better mileage.

Another possibility would be an oxygen sensor in one end of an exhaust manifold (or on one runner of a header) so that fuel could be cut off to other cylinders and yet the oxygen sensor would stilll get information from only the cylinder which was still firing and the ECU could adjust fuel quantity to get the best mixture.

[cfg83]

Silveredwings and rh77 and Erik -

Thanks for the info. I didn't know these details. I will experiment to see if my car will go into "0 throttle + engine engaged => gas cutoff" mode. I couldn't tell this before without the ScanGauge. If it works, I can add another strategy to my MPG driving toolbelt.

CarloSW2

[zpiloto]

Originally Posted by rh77

Also, if you're into tinkering under the hood, you can bypass the fuel injection electrical system by installing a switch inside the car to manually cut injection. I did this with 2 cyldinders during an experiment a while ago (basically, you re-route the wire that powers the FI to the switch and back to the injector series). Since my car does this already (didn't know it at the time) it was moot to flip the switch on decel. Now, I just play around with N and D to coast and engine-brake where necessary. Only if I had a clutch

RH77

What conditions allow the fuel cutoff to work and at what point do the injectors come back to life?

[cfg83]

theclencher -

Originally Posted by theclencher

I tried a 3-cylinder mode once and did it ever suck! Bad vibes...

Which cylinder did you cut off? When I was talking to a friend of my mechanic, he said that it should "just work", but with caveats. For a regular 4 cylinder, he said that cylinder 1 is needed for ECU timing and I think cylinder 3 for EGR data.

CarloSW2

[Silveredwings]

Originally Posted by zpiloto

What conditions allow the fuel cutoff to work and at what point do the injectors come back to life?

Empirically, I've determined cutoff begins when:
1. engine coolant is at operating temperature.
2. catalytic converter is at operating temperature.
3. foot is off the gas pedal.
4. car is coasting at such a speed that with whatever gear is engaged, the engine rpm is at or above ~1500 - 1600 rpm.
5. conditions 1 - 4 continue for ~1 - 2 seconds.

Cutoff continues until either:
1. the car speed slows so that the engine rpm drops below ~800 - 900 rpm.
2. you put the clutch in and the engine rpm drops below ~800 - 900 rpm (or around idle speed for your car).

So this means that if you are going to be slowing down in the very near future, you want to choose a gear in which to coast so that you can maximize the time you can spend in this cutoff mode before you have to do one of 4 things:
1. put the clutch in and stop (engine is now idling).
2. put the clutch in and coast (engine is now idling).
3. put your foot back on the gas.
4. switch gears.

[GasSavers_Ryland]

From what I can tell from honda shop manuals, fuel cut off happens when:
Engine is fully warmed.
Manifold vaccum is high (foot off gas, or fast running engine at low throttle).
vehicle is going above 10-15mph (varries depending on modle)
Engine is turning over 1,200rpm
Clutch is engaged.

Honda, and a number of other compenies started doing fuel cut off with carburated engines in the '70's, maybe earlier.

[rh77]

Originally Posted by Erik

I tried a crude fuel injector cut off earlier this year...when I would make a long trip (3 hours one way) on the interstate....The engine vibrated a lot under 3000 rpms, but at 65mph (3500 rpm), it felt smooth (although sluggish).

I did the same thing with my Integra about a year or so ago (article) and had the exact same results with 2 cylinders. It had to run above 3K or else it would rock and vibe like mad. I saw no appreciable gain either. I made a long highway trip with the setup and had to flip back and forth for mild load (OK) and accel/heavy load (bad under 3000 RPM).

I'm not sure when the FI's kick back in, but the comments thus far seem to make sense (what I don't get is that the SG doesn't account for it -- it should go to 9999 mpg like in engine-off mode).

RH77

[cfg83]

rh77 -

Originally Posted by rh77

I did the same thing with my Integra about a year or so ago (article) and had the exact same results with 2 cylinders. It had to run above 3K or else it would rock and vibe like mad. I saw no appreciable gain either. I made a long highway trip with the setup and had to flip back and forth for mild load (OK) and accel/heavy load (bad under 3000 RPM).

I'm not sure when the FI's kick back in, but the comments thus far seem to make sense (what I don't get is that the SG doesn't account for it -- it should go to 9999 mpg like in engine-off mode).

RH77

Egg-zactly. I did this test today. I know that when the engine is off and I am coasting, I can see insane instant MPG, something like 370 MPG @ 35 MPH. Notice that I don't see infinite "9999" MPG. I think this is because the ScanGauge has some kind of failsafe against divide by zero. When I idle in Neutral at 35 MPH, I see something like 150 MPG. When I was in gear and essentially "engine braking", the MPG was still good but not as good as either of the numbers above, probably because the RPM was higher than idle (Note, these are not absolute numbers from a controlled test, but ballpark numbers I observed on my normal commute. But they do represent the kind of numbers I have seen in the past).

What this tells me is that either the engine is still being fed gas during the above idle RPM engine braking (bad Saturn, Heel!), orrrrrr, the ScanGauge is extrapolating the MPG from other data input (sees above idle RPM in engine braking and does not assume fuel cutoff despite 0 TPS condition). The ScanGauge does not see the actual fuel being squirted as data input.

Segue Way - If the ScanGauge saw the fuel being used, why doesn't it offer that as an instantaneous output, aka # mg (milligrams) or something?

Caveat - Since your Engine/Computer are different, you may see different results.

CarloSW2

[striegel]

Originally Posted by rh77

[URL="http://www.gassavers.org/showthread.php?t=243"]

I'm not sure when the FI's kick back in, but the comments thus far seem to make sense (what I don't get is that the SG doesn't account for it -- it should go to 9999 mpg like in engine-off mode).

RH77

But doesn't the ScanGauge documentation show that it only calculates the fuel consumption rate? I seem to remember reading that. So it probably just estimates from the engine displacement and the air flow rate.