ScanGauge: How much calibration required on each tank?
 

I've been using my ScanGauge for a couple months now. After the initial calibration it has proved to be pretty accurate. My main complaint is the deviation from one tank to the next. It can be a pretty large difference. For example, one tank says I got 52 MPG but the gas pump says it was 48. I follow the procedure in the manual and offset the fuel usage to "calibrate" the ScanGauge. The next tank it might indicate 46 MPG, but according to the pump I got 50. According to the manual I should calibrate it every fill, so thats what I have been doing.

So I am just curious if this is common, and do you guys just deal with it? I've been thinking about just leaving the tank-to-tank calibration alone. Obviously making minor adjustments doesn't help the accuracy. It frustrates me because I think I'm getting 50 MPG but its only 46 or something. Maybe I should just leave it alone and not worry about it?

To get it to average out a little better you can calibrate it over 3 tanks or so. After that I'd just leave it. If its off a little its off a little. All the OEM guages are off a little (sometimes a lot). Its not an exact figure, just a real close guess.

Yes, average it out. The more tanks you average it over the better. Record the % correction required each time and average those.

One tip passed along here was to NOT enter the 'whole' error value, but only HALF of it, during each MPG re-calibration. That way your MPG values don't tend to oscillate between OVER & UNDER estimated values with each fillup. This technique "settled" my SGII readings quite quickly.

I thought I'd had mine dialed in for a while, but some tanks it seemed to miss the mark by a significant margin. I recently started tracking the scangauges amount and %correction with my fuel log because I think my sg can be off by +/- 5% depending on type of driving. For tanks that I spend entirely in town I seem to get worse than the sg shows, but for long trips I get better than the sg shows.

I wonder how much more accurate the MPGuino is? Can anyone chime in? I'm really not that concerned about accuracy, but it would be interesting to see if the MPGuino shows the variations that the SG does. Since the MPGuino uses a much more precise way to calculate MPG I suspect it is more accurate.

MPGuino, accurate once dialed in.

Almost the same problem here. I didn't get as much deviation, only a few tenths of a gallon, but one time I suddenly got about 2 gallons off. I've been recalibrating every fill up since and now its usually under by .2-.4 at most. I'd say that might be your cure... At least try it for a number of tanks and if it does get better then try leaving it.

I keep a spreadsheet and I can calculate my total offset since the start. If I use that and leave it, it's more accurate than any single pump fill.

The "adjust it only halfway" suggestion is excellent, though.

...FWIW, someone HERE suggested it first, not me, I just recalled & use it.

I wonder how much of the problem is related to another problem I'm having, which is the SGII randomly shutting down and restarting. It happened to me on a recent 250 mile trip. On the way down, it showed I got 28 mpg (I was going to a lower elevation on average). When I refueled, the manually calculated MPG was about 26, which I kind of expected since I was going back up in elevation. But my fillup gallons used was off by about 2 gallons or more. I have noticed the miles used since last fillup sometimes don't match the miles on the tripometer by quite a lot, which would seem to point to the same problem. And BTW, I don't stop the engine and coast as other more diligent folks do; my old cars tend to have cranky starters.

The scanguage can be an accurate tool, but it has its faults. If it generally seems to be pretty close, I generally ignore the odd 2/10 of a gallon discrepancy. I'm curious though how accurate the nozzle is to pop at the same place everytime. I suspect factors such as weather, temperature of the fuel in the pump and in the car, and how empty the tank was can affect when the nozzle pops. Once upon a time I would refill the tank to the reading on my scanguage (once I had it zeroed in) and it was surprising how close it usually was.

I had stopped using mine on Hägar the Diesel stove as it was too susceptible to changes in driving style.

The way I feel about it, it was more calibrated to how you drive the car, than to how much fuel the car uses.

Since getting the up! , I haven't bothered to install the SG again.

Sounds like my experience. It seemed to respond really well out of the box, but after using it and calibrating it a few times it seemed inconsistent. Now that its been calibrated every fill up for almost 1 year it's a lot better. I wonder if that has anything to do with it or if it only counts back 1-3 calibrations.

My MPGuino is also rock-solid accurate. I have an UltraGauge (cheaper and flashier version of the ScanGauge), and it will vary wildly in its tank fuel economy calculation. It'll tell me that I get anywhere from 19 MPG to 22 MPG, when the MPGuino (and my tank calculations from each fill-up) will reliably report 24-25 MPG.

Too soon to say in my Fit, but in my old Civic, my total error between Scangauge and actual pump gallons was less than 1/2 gallon per year. Once it's calibrated it can be amazingly accurate. Individual tanks would vary by a lot more, but that's the fault of the pump rather than the 'gauge. It all averaged out.

Does anyone know how the SG or UG actually calculate fuel consumption? I asked the scanguage folks & they wouldn't tell me, its proprietary. Since they're not reading injector pulses they must be doing some estimation using throttle position, MAP, and RPM? It seems like you could get it calibrated to give you a fairly close average MPG for a given driving style and type of route, but if you change your style or drive more or less hills, or highway then the calibration could be way off.

Simply:

a) assume A/F-ratio of 14.7:1 (closed-loop ECM);
b) measure Air flow from MAF or MAP sensor;
c) backsolve for GALS/time;
d) and divide: MPG = (MILES/time)/(GALS/time)

That should work as long at the A/F ratio is controlled tightly and the air flow is measured precisely. So why does everyone have so much trouble calibratiing? Why did the SG folks tell me it was a proprietary algorithm?

I'm guessing that most cars run a little rich from 14.7:1. Almost always the calibration is "more gallons" and not less.

I've been using my SGII on a 1995 Thunderbird (Ford had OBDII on 94/95 T-Birds and Cats), a 2001 Mercury Grand Marquis, a '2007 F150 FX4, and starting this past June, a 2008 Mustang GT (all with dedicated cords and velcro). Occasionally I'll slip it on my wife's '03 Forester.

I tried that calibrating when I first got it, way back in 2010 .... but seems to me I was just calibrating it to fuel tank size. Maybe it was calculating based on AFR / O2 sensor readings? Anyway, I haven't calibrated it since 2010 in any case.

I religiously figure my MPG fill to fill using math and I keep the SGII left side X-gages set to MPG instant above MPG average, and I find that when I drive without shutting off (so no resetting) from fill to fill like one may see on road trips (love road trips) so the SGII has it all .... the SGII average reading is nearly always within a tenth or two of what math tells me.

I know it knows "time", it get's a distance from speed devided by time, I didn't know for sure where it got gallons per hour, but appearantly it get's a pretty good sense from whatever my Fords tell it? I do know by using radar (30+ years LEO) in the past and by timing over mile posts multiple times that with OEM tire sizes, my veh's have pretty accurate odo/speedometers.

I use the instant MPG like a vacuum gage of old, which I still use on my '77 F150. It's good to train one to anticipate stops / slowdowns / hills, etc to maximize the average. It's almost like a video game .... I have favorite stations and I find myself trying to beat my normal average from fill up (reset) to home.