Shifting Tactics

[TheSSG]

I've been meaning to try out some experiments on shifting tactics, but I've been unable to get around to doing it.

Downshift on hills
The theory goes that by downshifting, one uses less throttle. Throttle position is more important than RPM goes this theory.
So, by going 55 @1900RPM in 4th, I'd get better mpg than going 55@ 1600 RPM in 5th, as I wouldn't be using so much of the throttle...
Kinda makes sense, eases the load.
I read it in an analogy:
Riding your bike up a hill in High Gear vs Low; requires MUCH more energy. Going up a hill in low gear, you're moving MUCH faster, but it's overall MUCH easier...



Coasting vs Downshifting
The theory goes: "Modern" cars' ECUs cut fuel to the engine when you downshift (Clutch engaged).
When one coasts, the car is idling, consuming fuel.
Therefore, when the clutch is engaged the Engine is merely "going through the motions" as opposed to idling, and using gas.

Only thing is, this works on "Modern" cars. I've read, FOR SURE, OBDII and up.
Since most of do NOT have OBII+, I don't think it will work.




So yeah, I'm curious, but too busy w/ school to remember to try these out...

[SVOboy]

If you look at your car's fuel map, you will notice that it is down by rpm versus load. As the rpm increases, even with the same load, the fuel injected also increases. This is because your injectors have to pulse more often to accomodate the increased amount of revolutions. So in the end, you'll end up using more gas by down shifting, because even if the load is less, you've got to go heavier on the throttle to keep it injecting enough to accomodate the rpms. Such is the reason the new corvette has a sixth gear for high speed cruising and can get 29 mpg highway, because it keeps the revs and load low at higher speeds than with a 5th only.

[Matt Timion]

Only thing is, this works on "Modern" cars. I've read, FOR SURE, OBDII and up.
Since most of do NOT have OBII+, I don't think it will work.


This is the one that I really want to know. In my mind my obd0 car gets better gas mileage when I throw it in nuetral and coast. Perhaps it is simply a obII+ thing. At least I hope it is, b/c breaking this habit will be a tricky one.

[SVOboy]

DrDisco69 told me that no pedal pressure is equal to no fuel injection, which I believe, but also that the engine braking will slow you down quicker and you'll spend more time on the gas/accelerating, and won't have as much cruise. So the idea is to cruise in idle longer than cruising in gear and hopefully using less gas.

[rh77]

I believe the Honda shop manual states that at closed throttle, small pulses of fuel are injected for some reason. I'm thinking of confirming this somehow, but I'm not sure. Maybe I'll tie the fuel injector deactivation to all cylinders and rig a switch into the accelerator pedal that cuts all of them at 0% throttle input. The fuel savings may be minimal, not sure though...

RH77

[SVOboy]

Get a scangauge! You're obdII!!! I forgot, I'd love one, tells you trip mpg, instanteous mpg, the ultimate testing aid, and it's accurate!

[Flatland2D]

I've been thinking about this one a lot lately. I think the idea that the fuel injectors are cut off when the engine is overdriven (coasting with no throttle input) is true. I've been leaving my car in gear when coming to a stop. It'll slow down until about 1000rpm then there's a little kick, like dropping the engine back in gear from being in neutral. That is supposed to be the fuel injectors turning back on to save the car from stalling. My car will still drive forward engaged in any gear without any throttle unput. The higher the gear, the fast it goes at idle (though rpm isn't much different). I'm guessing the engine is just trying to save itself.

So I'd say, coast in gear if you're going to be coming to a stop and use zero gas. This would be when approaching a red light or slowing down for traffic. When you are coasting because you're going slightly downhill and want to resume the same speed you started at, coast in neutral. Otherwise you'd probably waste more gas trying to get back up to speed after each coast in gear. If you're going really downhill and can coast in gear and not slow down, then that would be the best way to go.

[DaX]

I agree, it's a balancing act. The coasting thing doesn't work so well for us with automatics though, as generally when coasting @ zero throttle, the car kicks into the longest gear [in my case, fourth] to keep RPMs as low as possible, and then sequentially downshifts through the gears as speed decreases. I can't wait to install my manual tranny so I can take advantage of injector cut at coast.

[Flatland2D]

Is there any reason why automatics don't do this also? I was thinking they worked the same way. I've noticed in my wife's Civic (automatic) that there's a little blip in the rpm's around 1000 when coasting and coming to a stop.

[SVOboy]

My auto generally runs at around 1500-2000 foot of the gas in gear coasting.

[Flatland2D]

Yeah but as you slow down to a stop from a coast, do you notice anything happenning to the rpm right before the stop?

[Flatland2D]

Forgot to add this in my other post. I've also noticed it's hard to drive at certain speeds like ~35mph because as soon as I step on the gas it accelerates and doesn't just hold the speed. If I let off I'm coasting and slow down. It's a real fine line and can sometimes be really annoying speeding up and slowing down. I originially thought there was some play in the transmission because it's like the power kicks in all of a sudden. Makes the ride kind of jerky.

Now after thinking about the fuel injectors cutting off, I wonder if that's why it's so hard to hold a speed at a low rpm. Instantansous power keeps kicking on and off and makes it real hard to hold a constant speed.

[DaX]

Is there any reason why automatics don't do this also? I was thinking they worked the same way. I've noticed in my wife's Civic (automatic) that there's a little blip in the rpm's around 1000 when coasting and coming to a stop.

From what I gather, that blip is the transmission shifting back into first gear from second. I'm sure the automatic hondas do have fuel shutoff on hard decel [it's hard to imagine they would not implement this just because the car is auto], but the auto transmission by nature tries to keep as low an RPM as possible at low to no throttle input, thus defeating [eventually] the fuel shutoff.

I'm looking at my OBD training manals for Hondas, and for MPFI <b>and</b> DPFI systems, it says "if the ECM receives a closed throttle input (under 0.5V), the engine speed is above 1100 RPM, and the engine is warmed, the fuel injectors will be shut off." There is no distinguishing between auto and manual.

[DaX]

Forgot to add this in my other post. I've also noticed it's hard to drive at certain speeds like ~35mph because as soon as I step on the gas it accelerates and doesn't just hold the speed. If I let off I'm coasting and slow down. It's a real fine line and can sometimes be really annoying speeding up and slowing down. I originially thought there was some play in the transmission because it's like the power kicks in all of a sudden. Makes the ride kind of jerky.

Now after thinking about the fuel injectors cutting off, I wonder if that's why it's so hard to hold a speed at a low rpm. Instantansous power keeps kicking on and off and makes it real hard to hold a constant speed.

If you're talking about this happening in an automatic, it's probably because you are teetering on the shift speed of one of the clutches inside your transmission. If it's a 5-speed [I had this problem on one of my Civics], I'd look into either a vacuum leak or a dead spot in the TPS as the cause for the bucking.

[Flatland2D]

This is in a manual transmission. I'm guessing it could be a dead spot in the TPS. I do have a small vacuum leak, but I don't think that would cause the problem in the way it is effecting my car. Thanks for the info about that.

[Matt Timion]

This is in a manual transmission. I'm guessing it could be a dead spot in the TPS. I do have a small vacuum leak, but I don't think that would cause the problem in the way it is effecting my car. Thanks for the info about that.

If you want I can get a few TPS sensors the next time I'm at the junkyard and send one to you.

[DaX]

If you want I can get a few TPS sensors the next time I'm at the junkyard and send one to you.

You should be snatching those up anyhow...pretty hard to come by these days, and dealerships won't sell them by themselves. You know how to get them off the throttle body, right? My brother dropped and broke one of mine and I about cried.

My FITV got messed up and was causing a vacuum leak at one point and my car bucked horribly until I fixed it.


Oh yeah, if any of you honda heads out there want a copy of these OBD training manuals I have, they are VERY informative for the PGMFI and OBD systems on honda cars up to but not including K-series stuff. The zip file is 8MB...I don't have an FTP or anything but I can email it.

[Matt Timion]

I think I'll go get a bunch of TPS sensors at the junkyard. I'll test them with a multimeter and sell them on ebay. :)

oh yeah...

400th post for me!!

[Compaq888]

Here is how my auto does it. When I'm at speed let's say 30mph. If I want to slow down or just coast I just release the gas and the rpms go to 1000rpm or lower and the car keeps rolling and only loses a little bit of speed.

I usually coast 1-3 blocks If I see a stale green light or yellow or red. If nobody is around I coast as much as 4-5 blocks to the light.

[diamondlarry]

My 1992 Pontiac Gran Prix acts the same way. It seems like it will coast forever by just taking my foot off the pedal.

[worthywads]

I've been doing some experimenting with coasting in neutral vs. in-gear with my scangauge. Here are my observations.

1. In a 2.7L Tacoma I don't think it ever shuts the injectors off complete during in-gear coasting. I tried reving as high as 5500rpm in 2nd gear and then coasted. I saw a drop from 1.9 gallon per hour just after letting off the throttle, then 1.1gph at 4500 and 3500 and then .6gph at 3000 or so. Conclusion, for this engine in-gear coasting doesn't reduce fuel use.

2. During idle my fuel usage is .3 gallons per hour, and during in-gear coasting it is .3gph or greater. In-gear coasting at 2000rpm is .4gph, I plan on going through the range of higher rpms to see how much gph rises. Conclusion, below 1800rpm it doesn't gain anything by dropping into neutral unless in-gear slows you down faster than you want.

3. Going up hills keep it in the highest gear that doesn't lug, higher rpms seem to always mean more fuel use on this truck.

[MetroMPG]

hi worthywads -

1. In a 2.7L Tacoma I don't think it ever shuts the injectors off complete during in-gear coasting. I tried reving as high as 5500rpm in 2nd gear and then coasted. I saw a drop from 1.9 gallon per hour just after letting off the throttle, then 1.1gph at 4500 and 3500 and then .6gph at 3000 or so. Conclusion, for this engine in-gear coasting doesn't reduce fuel use.

i have a scangauge too. my car's service manual states that the injectors are off when engine braking, yet the SG doesn't reflect this.

i've talked to a number of SG owners (including JanGeo on this site) about this and nobody's SG is showing "injectors off". i believe it's a SG bug.

[worthywads]

I have a scangauge too. my car's service manual states that the injectors are off when engine braking, yet the SG doesn't reflect this.

i've talked to a number of SG owners (including JanGeo on this site) about this and nobody's SG is showing "injectors off". i believe it's a SG bug.

Tried to contact SG off their website but I get a error?

I'd like a description of how the gph is measured or estimated.

After the first 3 fillup calibrations I have been off at most .1 gallons on any tank, though every fillup has been actually more by this little bit than the SG. I'd think the Fuel Used would be way off if the SG is wrong about engine braking. Then again if my driving is similar from tank to tank my fuel adjustment could be compensating. I haven't taken a long trip with minimal engine braking, it would seem on such a trip the SG would then be way off.

Can't seem to find any specifics on fuel injector function during engine braking for my Tacoma.

Aargh.

[JanGeo]

Yeah I asked my brother about it and he said that there is a minimum amount of gas being injected with no throttle and by downshifting the rpm goes up and thus they inject the minimum amount at a higher rate thus using more gas. I got 66mpg 3rd gear 35mph 120-150mpg top gear in this case I think using the brakes a little will be better as they need some use once in a while anyway. I also see the slowdown in top gear until about 1000 rpm (20mph) then the engine kicks and it's time to throw in the clutch. I did one ScanGauge fillup calibration - second one is today as the gas light has been on for a few miles - in town short trips with really sucky mileage like 17mpg 0.5mile trips up and down hills. Maybe we should wire LED's to the injectors to monitor the injector pluses and get a light show to boot!

[Sludgy]

I suspect that auto engineers don't map the fuel to zero when coasting in gear because of pollution control:

If the fuel is shut off completely while coasting, the catalyst will cool off quickly because unheated air will flow through the exhaust. The car would never meet emission standards if the catalyst was cold every time you go back on the throttle.

[krousdb]

I suspect that auto engineers don't map the fuel to zero when coasting in gear because of pollution control:

If the fuel is shut off completely while coasting, the catalyst will cool off quickly because unheated air will flow through the exhaust. The car would never meet emission standards if the catalyst was cold every time you go back on the throttle.



My Honda shop manual says that the fuel shuts of when the throttle is closed and rpm's are above 1100. The Honda OBD training manual states the same thing. And the SuperMID confirms this showing the injector pulse width as zero when the engine is at normal operating temperature. So it would appear that Honda has figured out how to do this without affecting the catalyst.

[Sludgy]

I'll bet Ford and GM haven't figured it out....

[philmcneal]

Here is how my auto does it. When I'm at speed let's say 30mph. If I want to slow down or just coast I just release the gas and the rpms go to 1000rpm or lower and the car keeps rolling and only loses a little bit of speed.

I usually coast 1-3 blocks If I see a stale green light or yellow or red. If nobody is around I coast as much as 4-5 blocks to the light.

maybe that's why camry's automatics get such good mileage. They coast for such such a long time if almost putting it neutral does not affect its coast at all (unless hill is super steep).

[JanGeo]

Hey A few tests done going up hills in different gears - steep hill I go up and down every day seems 3rd 18mpg 4th 21mpg 5th 25mpg speed 29mph about 10 degree slope so higher gear lower rpm better mileage.

Coasting in gear (standard) the faster you go the higher the mpg so if in neutral you coast faster and get better MPG unless you really want to slow down put it in neutral and coast faster. Fuel shut off calculations in the SG are not possible division by ZERO!!! DUH!!! but I see the same kick at about 18mph to 20mph depending upon idle speed hot cold etc. when slowing down in 5th and it seems to provide constant engine braking until about 20mph and the fact that it shows a decreasing mpg reading indicates that it is injecting a constant amount of fuel over the range of speed. Killing the injectors will not change the SG reading because it is monitoring the computer not the injector signal. It will perhaps vary the braking action as you kill the small amount of fuel injected if it is injecting. Again with no change in mpg with engine turning faster or idle then it comes back to the SG minimum injector setting.

Another recent thing I have had happen to my in the xB is sticky / no fast idle with engine cold and wet low temps so that when I back off the gas to shift the car quickly tries to slow down before I can get the clutch in where it usually cuts out a little after the clutch is in unless I let off the gas a little ahead of time before clutching.

[MetroMPG]

hey, is that the official scangauge builder's response on the "injectors off" issue? did you write to him?

the scangauge seems able to divide by zero (i mean figuratively - it uses some other calculation mode) when i kill my engine with the new switch.

good info on the hill climb gear/mpg - thanks for posting that.

[rh77]

I rented a Toyota Matrix (Pontiac Vibe) on a few occasions and found that from a cold start, under heavy throttle, the ECU won't let the RPMs go higher than around 3000 RPM until the coolant temp is high enough -- perhaps an engine saving or emissions feature. I think the ECU is pretty complicated and un-traditional and may apply to the Scion line (similar engine probably). Maybe the computer doesn't react to closed throttle as expected.

I need to get a ScanGuage...injection pulsation is good info to have.

RH77

[JanGeo]

Not official just a guess - any way you could check the exhost for fuel use - oxy sensor feedback? to see if gas is being injected?? Heat rise of a cold engine may idicate something also. Fuel pump current?

Just thought of another test - climbing hills - going fast and going slow should yield the same mpg unless the engine has a sweet spot. Maybe a great way to test power levels without going too fast to eliminate air drag factors. Going up hill with a cold engine really gets the engine temp up quickly - keep rpm low to increase heat absorbsion inside the cylinders.

Coasting in/out of gear - if you have to time a light then slowing down sooner will preserve more speed when getting to the light green. Also stopping on a down hill with a light - better to stop higher up on the hill if you can and use the hill to speed up when the light turns green. Stop signs don't work with this technique.

[MetroMPG]

I need to get a ScanGuage...injection pulsation is good info to have.

it actually doesn't show that directly (though it must USE that info for the instant/avg MPG calcs). it doesn't show fuel trim either.

[rh77]

it actually doesn't show that directly (though it must USE that info for the instant/avg MPG calcs). it doesn't show fuel trim either.


I wonder if it's possible to get a "Y" OBD-II splitter, so I can use my data logger to get fuel trim, etc. and plug the ScanGuage into the other port.

RH77

[Matt Timion]

it actually doesn't show that directly (though it must USE that info for the instant/avg MPG calcs). it doesn't show fuel trim either.


I wonder if it's possible to get a "Y" OBD-II splitter, so I can use my data logger to get fuel trim, etc. and plug the ScanGuage into the other port.

RH77

I bet that you could make one really easily. I also bet that you could contact the guys at www.rywire.com to get one made for you.

[SVOboy]

I rented a Toyota Matrix (Pontiac Vibe) on a few occasions and found that from a cold start, under heavy throttle, the ECU won't let the RPMs go higher than around 3000 RPM until the coolant temp is high enough -- perhaps an engine saving or emissions feature. I think the ECU is pretty complicated and un-traditional and may apply to the Scion line (similar engine probably). Maybe the computer doesn't react to closed throttle as expected.

Similarly, my friend's 01 civic sets a rev limit based on certain factors. For example, I can't rev it up in park, or with the brake on, so there's no possibility to peeling out unless I neutral bomb the thing, but I'd feel bad doing that to someone else cars, :p