didn't do the hill test yet (i'll wait until i actually need to go somewhere in the car for that one).
but i did an abridged version sitting in the driveway just now:
started the car, watched the scangauge come alive, then stalled out the engine with the brakes and 5th gear.
tentatively good news: the scangauge "Gallons per Hour" reading immediately dropped to 0, but the SG continued reading data. well, temp anyway: the coolant temp sensor clicked up a couple of degrees in the 30 seconds after i killed the engine.
even better: re-starting with the key didn't reboot the scangauge - it pretty much immediately resumed reporting the GpH data (and everything else). i'll try it on the road next time i'm out, but i suspect all is well with this plan.
i've already scrounged a switch i'm going to use off an old computer power supply. i'll wire it to the 12v power supply to the ignition coil and it can do double duty as a theft prevention device too (it's an on-off switch - which i can leave "off" when parked).
Ok the real question is what type of ignition system are you running points or solid state. The coil is charged up when the points are closed and the 12 volts from the ignition switch is connected to the battery along with other 12 volt stuff in the car. When the points open the magnetic fields collapses and a spark is created in the plug and a 300 volt or so voltage spike appears across the points and condenser which is there to give the points time to open a little before the voltage can jump the gap and to provide a little resonance for the inductance of the coil to make it oscilate and make a better spark. The coil is connected on one end to the spark plug throught he distributor and the other end was connected to ground but you opend that up so the Positive connection to the ignition switch and battery is the only other connection carrying the spark energy back to the ground. Open up the positive side and a couple of things happen - the coil could be already discharged and nothing (this is ideal) or the coil is fully charged and the points are closed and you end up firing the coil at the wrong time and cause a misfire or the ignition is electronic and you possibly spike the distributor points. Why does this not happen normally - well when you turn on the ignition you are powering up lots of stuff besides the coil circuit like the dash lights and sensors and the computer and fuel pump etc so that when you turn off the ignition you still have these devices loading down the charged up coil and helping to ground the coil to the negative earth of the car. Now you are just opening up the connection to the coil with all the other loading removed and I just think you will at some lucky time get this really big spark across that switch. It could be better to open the points side of the circuit but you still have the potential to create a misfire. In any case use high voltage wire with good insulation.
i was trying to think of a way of testing this concept without having to actually wire it up first. ie. test the car & ScanGauge's response to an "engine-off" condition that was not initiated through the ignition key switch.
would this be valid: point the car down a hill at a walking pace and then intentionally stall it out with a high gear.
then coast down the hill and watch what happens to the SG - and also whether i can re-start with the clutch at the bottom of the hill.
remembered to try this test today on my way back from getting the $75 swift ready to go, and it appeared to work.
i coasted down to a walking pace near the top of a gentle hill in my neighbourhood, selected 5th gear, released the clutch momentarily and stalled out the motor.
the scangauge remained on as i rolled down the hill, the "avg" mpg figure (recently reset) climbed steadily as i moved along. i also saw the scangauge odometer tick up an additional tenth of a mile while doing this.
at the end of the coast, i selected 4th, and bump started the engine with no difficulty. the scangauge continued as normal.
the ignition key remained in teh "run" position the entire time.
so this is great news. it means i can install a separate kill switch and do a lot more engine-off coasting without putting the SG to sleep and screwing up its tank/trip FE data (which is the reason i wasn't doing engine-off coasting as much as i could have).
I abuse killing my engine and I'M LOVING IT. I just have to watch out for the cops, as I did it at a red light and I turned off my headlights to save my 12 volt (temps 37.4F or 3C) although someone answer me how long do I have until my headlights drain the 12 volt? And how about them daytime running lights? I wish those can fool the cops but those are too dim at night) I hear this , "TURN ON YOUR HEADLIGHTS!"
I jumped and turned on the car immediately, damn what a frieght there was a RCMP officer waiting in the opposite direction... then when the light turned green I looked around my surroundings and engine killed again ;P
This abuse has dropped my consumption from 6.5L/100km (35mpgUS - 80%city - 20% highway) to 5.4L/100km (43.6 mpgUS) 100% city driving.
Amazing isn't it? Thanks everyone! Now lets hope I don't get ticketed for engine killing... then my savings will be a bust ;(.
If your reading this, then good for you, your saving some gas because your here.
I leave my lights on while coasting, engine off. They do dim a little, but are still effective. With a good battery, you should be able to leave your lights on for at least an hour with the engine off.
I don't think that you can get a ticket for dim lights. Only no lights at all.
Most newer cars (MPI) will shut the fuel off when coasting down hill. It is a function of the idle circuit. If the engine is above the predetermined idle speed (foot off of gas), the system will try to reduce the rpms.
scavenged an on/off switch from an old computer power supply & spliced it into the +ve wire to my distributor pick-up coil today.
haven't mounted it on the dash yet. i'm kind of thinking i'd like to visit the junkyard and get an a/c switch out of another car, and connect it there in the "blank" on my dash. then it can also serve as a simple & relatively stealthy engine immobilizer as well.
seems to work as promised. the only thing i'd like to do is either solder the wires (i just twisted them together), or crimp on some connectors.
anyhoo... looking forward to doing much more engine off/scangauge on coasting.
i've already got a simple experiment in mind: compare mpg from idling/neutral coasting to engine off coasting over the same city route.