How to over-ride computer RPM while coasting?

[Ecky]

Originally Posted by mcrews

Same way that coasting at 65 at 1 rpm doesn't double the mileage from the 750 rpm. It's not a simple math equation. I would actually assume it is based on what is required to rim the engine. Engine load, gearing etc.
the issue, imho is consumption of fuel during the brief moments of coasting TO THEN justify a hack that might have unintended consequences.

The math IS pretty simple, you just have to pause and consider what you're mathing.

One mile at 50mpg followed by one mile at 30mpg should average 40mpg, yes?

If so, then it follows that one mile at 5,000mpg followed by one mile at 30mpg would be 2,515mpg, and as we approach engine-off coasting, our average goes toward infinity. We know this is not the case though.

We're not interested in miles per gallon per mile, we're interested in gallons per mile. Burning 0.020 gallons over one mile (50mpg) followed by burning 0.033 gallons over one mile (30mpg) results in a total fuel consumption of 0.053 gallons over two miles, or 37.7mpg.

Doubling your MPG while coasting might look significant, but it's important to figure out exactly how much fuel you're actually saving. It might or might not be significant over a tank, depending on how much coasting you do.

This is why L/100km is a better metric than MPG. Going from 15 to 17mpg actually saves 50% more fuel than going from 60 to 90mpg; MPG obscures this.

~

Just as a thought exercise, if we assume 15mpg @ 60mph climbing one side of a hill for 30 minutes, and coasting down the other side (assume also at 60mph) for 30 minutes, just how much fuel can you save by reducing fuel consumption while coasting? Note that we're assuming 50% coasting in your normal driving - it's probably far less than this.

15mpg @ 60mph = 4 gallons / hour

If you coast at 0.40 gallons per hour on the other side at 3000rpm, you're going to use a total of 2.2 gallons to go 60 miles (again, with 50% coasting), resulting in 27.3mpg.

If you coast at 0.20 gallons per hour @ 700rpm, you're using 2.1 gallons to go 60 miles, resulting in 28.6mpg.

Approximately what percent time are you coasting, would you say?

[Focus-Ak]

So what you have in your example is a 4.8% improvement in MPG, and a 4.5% reduction in fuel consumption. Incremental changes add up, and an increment from a reduced idle seems worthwhile for those who pulse and glide on open roads, especially (from my experience) hilly ones.

On slow city streets I think I probably glide over half the time, and then there is time idling at stop signs and (sometimes engine off) traffic lights. So there is a lot of opportunity there to both save gas and improve mpg (or l/100-km) as well.

------ Later edit, Jan. 28

I have calculated my average idle time from a datalog. It is from a 13 minute trip at mostly 40-50 MPH (some 30-35), no stops, with 23 pulse/glide cycles, starting and ending at the same place.

I used power 26% of the time (average 9 seconds, range 2 to 29 seconds).

Idle RPM coast time was 56% (average 19.3 seconds, range 4 to 60 seconds). Low idle differed on the glides, 600 to 750 RPM, usually about 650 (Focuses have a mind of their own on this).

The remainder was the RPM decline with zero throttle (i.e. dash pot effect) which was 12% of the time (avg. 4.1 seconds, range 3 to 5 seconds). Notably, these drops are much faster than before tinkering when my "low" idle was 1500 to 1600 RPM (idle drop to get there, 7 to 8 seconds). The Focus is notorious for having a high idle (many complaints on other forums).

So, I am saving in two ways; lower idle and faster drop to idle, the latter being important to make the most out of P&G driving.

Lowering my idle as above has made a repeatable difference of about 38 vs 41 MPG on this same route, measured on an UltraGauge, which is nearly an 8% gain.

[hackish]

You need to reflash the ECU for this. Nobody has bothered to develop such a thing because there is no money in it. There is a fair bit of economy left on the table but at the expense of tailpipe emissions. Changing the shifting program is one of the most effective ways to do it without changing the tailpipe output.

-Michael

[LioNiNoiL]

Originally Posted by Ecky

pause and consider what you're mathing.
One mile at 50mpg followed by one mile at 30mpg should average 40mpg, yes?

No.
One mile at 50mpg uses 1/50 gallon, one mile at 30mpg uses 1/30 gallon,
for a total of two miles at 1/50 + 1/30 = 8/150 gallon,
for an average of 2/(8/150) = 37.5mpg

[Enki]

Got the email for this thread today, and it caught my eye so I've read through it. I might have another answer for you guys as to why manual cars in particular do this, and I'm pretty sure it has little to do with the ECU.

First, some details on my own car; it's a 2009 Mazdaspeed 3 with factory drive by wire and electronic throttle body.

Now, for some anecdotal information:
A few years ago, I was redoing the PCV setup on my car to help with oil going all over the place in the intake and whatnot, and I completely bypassed (capped) the line going into the intake manifold (post throttle body). It ran fine except under one condition: Holding the clutch in after hitting high RPMs and letting the car coast.

With no air bypassing the throttle body (via PCV), RPMs would drop sub-idle and usually stall. Sometimes it would hunt for idle RPM a few times, but normally it would stall out. I fixed this issue by installing a hose to the intake manifold with a low spring check valve that goes to the intake, post MAF (so it doesn't pull in unmetered air).

My thought is that there's still some coupling going on which is causing the engine RPMs to run higher than idle (similar to how a manual car in neutral on jackstands will have the drive wheels turn slowly).

I vaguely recall seeing the throttle body reading the same as idle or below idle settings when I was troubleshooting this issue myself, but I've asked a friend to pull some logged data for me to confirm since my car is down. I should have the data within a couple of hours of the time of this post.

Stay tuned!

[Enki]

OK so it looks like my particular vehicle isn't trying to increase RPM on decel (rev matching). It takes about 5 seconds to go from 3k rpm to 6700 at right around 25% pedal, which is roughly the same throttle on his tune (we can change this and mine isn't linear like his is).

It takes 4 seconds for the engine to spin back down to idle, in which it hunts for idle a little bit and modulates the throttle position to do so (no change in pedal). Overall, it's trying to keep the throttle right around what it normally is at idle.

Sorry this theory doesn't pan out on my end, but it might be worth looking into on yours with some kind of monitoring device; I'd check idle valve duty cycle if there is one, as well as throttle position and compare between (hot) idle and when you encounter the issue. It's worth noting that it might be that your clutch isn't floating on the input shaft properly for whatever reason, and might still have a bit of friction against either the pressure plate or flywheel and thus, run higher while coasting.

Just a thought.

[Manualhybrid]

And it may affect the lubrication of your automatic transmission. Check your model of car to see if it is OK for 'dinghy-towing' or 'four-wheel towing' as if behind a motorhome. If it is not OK for dinghy-towing it may mean that with the engine off, lubricating fluid does not cycle through the moving parts of your transmission--even though you are in neutral. I used to engine-off-coast (EOC) in my automatic transmission Matrix till I was advised of checking whether the model was approved for dinghy-towing and, sure enough, it is not approved! I don't EOC anymore, fearing damage to my car. I just coast, engine on, in neutral.

[bob]

My 87 Dodge Daytona does this and it drove me crazy. Discovered different conditions vary coasting idle from 800-1,200 rpm. But, the biggest is when the front windshield defroster is on with the fan switch engaged then it's 1,500 rpm until almost stopped when it drops to 800 rpm. The easy fix is to press the bi level button and move the fan switch one notch higher. Some drivers run a independent power wire to bypass the computer and/or cut the sensor wire to the computer.
The Daytona has a transplanted 2.5 turbo 5 speed with some mild performance mods and fat 225 tires. Unfortunately it's ganged fuel injected that can only be fixed with a ECU change & that's above my pay grade so it only gets 17 ½ in town, 24 mix if lucky. However, highways a different story.
Aero's Work!
I flared the front nose plastic fascia (easy mod that looks good) to cover the tires and installed a black chloroplast belly pan (almost invisible) from the nose to the engine, about $20. The belly pan slips into & overlaps the front bumper cover and attaches with heavy duty Velcro for fast easy removal. This made a massive improvement not only in highway mileage but wind noise reduction. Also installed is a modified spoiler off a little Chevy S10 pickup using R clips for fast removal. I can remove both pan and spoiler in under 1 minute. Although the spoiler doesn't make any mpg difference it does help handling but also scrapes over every speed bump. This old car won't take a Scanguage without major wiring hacks but came with a primitive version stock that matches the odometer readings. Puget Sound highways are all up and down, rarely flat. In 80 degree weather I consistently get between 36-38 mpg at 58-60mph. Retarding the timing a few degrees, flat ground, good gas, a 5 mph tail wind, 55-58 mph, careful right foot and she does 43 mpg. Today I drove in a 18 wheeler's slip stream and did 59 mpg. In a few months I plan to flair the rear fender skirt and install a belly pan to cover the massive parachute bumper hole. Hoping for another 1 mpg highway.

[Enki]

Originally Posted by bob

But, the biggest is when the front windshield defroster is on with the fan switch engaged then it's 1,500 rpm until almost stopped when it drops to 800 rpm.

Most vehicles kick on the AC compressor for defrost; not entirely sure why other than possibly engine load (more heat faster) and temp control of the defrosting air.

[csnyder]

Originally Posted by mcrews

read this.....again
it's easy to get fixated on something like this....had a tarsus that did the same thing.
But there are probably three other things you could do to get better mileage.half of a quarter of a gallon OVER 1 hour is nothing.
maybe you can reduce the rpm in half, but it doesn't use twice as much gas to run 1500rpm (IN NEUTRAL). And honestly, how much time do you really spend coasting....
__________________________________________________ ____
1.) buy a scangauge
2.) aba test your theory
3.) report back

you can discuss all the 'theories' you want until you are blue in the face.

everything in life is a bell curve. It's not worth dealing w the fringe stuff UNTIL you address the stuff in the middle.
SO....... how much time and distance is spent 'coasting'? (the fringe stuff) Which by the way has the LOWEST consumption of fuel on the whole trip probably only 10-15% of driving consumption.
FOCUS on the stuff that matters.

I have a scanguage on the ranger, and coasting at 60kph at idle (750rpm) I will get 65MPG or more, while at 2000 RPM I'm getting closer to 30.
At 100 kph I may get 144MPG at idle and closer to 50 at 2000 RPM. I'll have to record actual mileafe next time I do it - but it is VERY substantial.

I moved the scanguage to the Taurus to see what happens there last summer, and coasting in gear or out makes no difference in mileage because the revs match the speed when in neutral and their is no closed throttle fuel cut - so "pulse and glide" doesn't buy me anything on the Taurus.

With the Ranger (4 liter 5 speed 2.55 ratio and 235/70-16 tires) on my Wedneday trip to the local airport just coasting on the hills changes my fuel consumption from 20 MPG without to 24-27 with coasting - on a 15Km trip, about1/3 in urban traffic - rolling land in the Grand River Valley of Ontario