Dan
08-24-2007, 07:32 PM
ScanGaugeII Calibration Worksheet (http://www.cleanmpg.com/forums/downloads.php?do=file&id=14)
I've had some requests to go into a bit more detail on the calibration worksheet. Here's the basic premise on it, posted in the Tips and Tricks thread...
Calibration
The math may get a bit fuzzy, but I'll try to clod trough it. There are two settings you can set, one is Speed plus or minus whole percents (±1%, ±2%, ±3%...). This effects MPH, DIST, MPG, DISTANCE TO EMPTY (think that's it). So if you see that SGII is reading too low on MPG, DIST, or MPH, then bump up SPEED, till MPH on the gauge and on the dash match (or until DIST on the gauge and on the odometer match).
The other setting you can adjust is in FILLUP and that controls how it monitors fuel use. This is a plus or minus percent meant to get the gallons you put in the tank to match the gallons SG calculated you burned. This effects GPH, MPG, DIST TO EMPTY. So if SG has calculated that less gas was burned than what you just pumped in, the set calibration to some positive percent till they come in line. The manual (http://www.scangauge.com/support/pdfs/SGIIManual.pdf) (search fillup) goes over this to some detail.
Now the interesting part comes when you start comparing your MPG. Since for me the MPG is the most important figure. MPG is Miles/gallon. Another ratio to state it is MPH/GPH (GPH=gallon per hour).
So increasing MPH increases MPG.
Decreasing MPH decreases MPG.
Increasing GPH decreases MPG.
Decreasing GPH increases MPG.
Sounds simple, but it's easy to get mixed up. So if SG reports a lower MPG than we calculate (either at the pump or on the FCD) then we need to increase MPG. We can do this by either increasing (+%) the speed calibration or decreasing (-%) the fillup calibration. And conversely if SG reports a higher MPG than we calculate (either at the pump or on the FCD) then we need to decrease MPG. We can do this by either decreasing (-%) the speed calibration or increasing (+%) the fillup calibration.
Confused yet? I've added a spreadsheet to the uploads section to do the math for you.
ScanGaugeII Calibration Worksheet (http://www.cleanmpg.com/forums/downloads.php?do=file&id=14)
Now there is a little trick I found on the FILLUP calibration. I found it best to do FILLUP>DONE without calibrating, then when gallons consumed is 0.0 do FILLUP again (before you start the car), that way you can set the calibration in arbitrary tenths of a percent 1.1%, 1.2% ... etc. Try it on your next tank to see for yourself.
Confusing... yeah I know. I'm not too gifted at turning math into English. Hopefully by restating the points there will be sufficient convergence.
So the spreadsheet basically takes a known good estimate of your mileage and compares it to Scangauge's view of your mileage. Works good if your known good is a car maker's fuel consumption display (like on the Prius), but not so good when the "known good" is a gas receipt. Problem is, it's hard to fill your tank to the same level every time. So when you under-fill on one fill-up and over-fill on the next, the calibrations vary wildly. So what is needed is an averaging mechanism (that I left out of the first draft). I'll correct it this weekend, so we might be able to fix it.
Let's take what's out there now.
The SGII comes with a calibration method that adjusts the actual gallons used and the measured gallons used. This is covered in the manual, so I won't dwell on it. My problem with this is that it doesn't offer enough granularity. Both speed calibrations and gallon calibrations effect the measured MPG. So what the spreadsheet does is predict the effect of both of these calibrations on a given tank.
Time for an example:
Lets say at your next fillup your ODO says you made 386.4 miles on your tank.
Now lets say SG reported your tank distance as 384.1.
SG underreported distance by 0.5988 %
Let say that at this fillup you put in 17.513 gallons
And SG reported 17.3 gallons consumed.
SG over reported gallons by 1.0565 %
So now at fillup, you change 17.3 to 17.5. Done right?
Lets see how that lines up.
SG(now) = 384.1 / 17.5 = 21.82386
Actual = 386.4 / 17.513 = 22.06361
So even after the calibration, SG is under reporting MPG by 1.10%
Now try with the spreadsheet.
Put zero next to MiAdj, RQMi, GalAdj
Put 386.4 next to MiFCD
Put 384.1 next to MiSG
Put 22.06361 next to mpgFCD
Put 21.82386 next to mpgSG
Now you will see that RQGal has changed to -1.09. The spreadsheet is suggesting you adjust gallons down by 1.09%. Close as we can get on SGII is 1.1%. So test it by putting -1.10% next to GalAdj.
Now you'll see that the mpgTST field went to 22.07. This is what SG would have reported WITH the calibration. So the spreadsheet got you a lot closer (22.07) than the normal calibration methods (21.82).
Now since distance is borderline we could test what would happen if we did a 1% adjustment on distance. So since SG underreported distance by 0.59% we could round up to 1%.
Put 1% next RQMi
Put 1% next to MiAdj
The spreadsheet now suggest dialing gallons down by 0.10%. So try putting -0.10% next to GalAdj, and you will see the prediction changed from 22.07 to 22.06.
So those are two calibrations you could try in this hypothetical
{Speed = 0%; Gallon = -1.1% } = 22.07 mpg
{Speed = 1%; Gallon = -0.1% } = 22.06 mpg
both of which are trying to zero in on your tank numbers of 22.06361
I'll be making some revisions for averaging soon... So stay tuned.
11011011
I've had some requests to go into a bit more detail on the calibration worksheet. Here's the basic premise on it, posted in the Tips and Tricks thread...
Calibration
The math may get a bit fuzzy, but I'll try to clod trough it. There are two settings you can set, one is Speed plus or minus whole percents (±1%, ±2%, ±3%...). This effects MPH, DIST, MPG, DISTANCE TO EMPTY (think that's it). So if you see that SGII is reading too low on MPG, DIST, or MPH, then bump up SPEED, till MPH on the gauge and on the dash match (or until DIST on the gauge and on the odometer match).
The other setting you can adjust is in FILLUP and that controls how it monitors fuel use. This is a plus or minus percent meant to get the gallons you put in the tank to match the gallons SG calculated you burned. This effects GPH, MPG, DIST TO EMPTY. So if SG has calculated that less gas was burned than what you just pumped in, the set calibration to some positive percent till they come in line. The manual (http://www.scangauge.com/support/pdfs/SGIIManual.pdf) (search fillup) goes over this to some detail.
Now the interesting part comes when you start comparing your MPG. Since for me the MPG is the most important figure. MPG is Miles/gallon. Another ratio to state it is MPH/GPH (GPH=gallon per hour).
So increasing MPH increases MPG.
Decreasing MPH decreases MPG.
Increasing GPH decreases MPG.
Decreasing GPH increases MPG.
Sounds simple, but it's easy to get mixed up. So if SG reports a lower MPG than we calculate (either at the pump or on the FCD) then we need to increase MPG. We can do this by either increasing (+%) the speed calibration or decreasing (-%) the fillup calibration. And conversely if SG reports a higher MPG than we calculate (either at the pump or on the FCD) then we need to decrease MPG. We can do this by either decreasing (-%) the speed calibration or increasing (+%) the fillup calibration.
Confused yet? I've added a spreadsheet to the uploads section to do the math for you.
ScanGaugeII Calibration Worksheet (http://www.cleanmpg.com/forums/downloads.php?do=file&id=14)
Now there is a little trick I found on the FILLUP calibration. I found it best to do FILLUP>DONE without calibrating, then when gallons consumed is 0.0 do FILLUP again (before you start the car), that way you can set the calibration in arbitrary tenths of a percent 1.1%, 1.2% ... etc. Try it on your next tank to see for yourself.
Confusing... yeah I know. I'm not too gifted at turning math into English. Hopefully by restating the points there will be sufficient convergence.
So the spreadsheet basically takes a known good estimate of your mileage and compares it to Scangauge's view of your mileage. Works good if your known good is a car maker's fuel consumption display (like on the Prius), but not so good when the "known good" is a gas receipt. Problem is, it's hard to fill your tank to the same level every time. So when you under-fill on one fill-up and over-fill on the next, the calibrations vary wildly. So what is needed is an averaging mechanism (that I left out of the first draft). I'll correct it this weekend, so we might be able to fix it.
Let's take what's out there now.
The SGII comes with a calibration method that adjusts the actual gallons used and the measured gallons used. This is covered in the manual, so I won't dwell on it. My problem with this is that it doesn't offer enough granularity. Both speed calibrations and gallon calibrations effect the measured MPG. So what the spreadsheet does is predict the effect of both of these calibrations on a given tank.
Time for an example:
Lets say at your next fillup your ODO says you made 386.4 miles on your tank.
Now lets say SG reported your tank distance as 384.1.
SG underreported distance by 0.5988 %
Let say that at this fillup you put in 17.513 gallons
And SG reported 17.3 gallons consumed.
SG over reported gallons by 1.0565 %
So now at fillup, you change 17.3 to 17.5. Done right?
Lets see how that lines up.
SG(now) = 384.1 / 17.5 = 21.82386
Actual = 386.4 / 17.513 = 22.06361
So even after the calibration, SG is under reporting MPG by 1.10%
Now try with the spreadsheet.
Put zero next to MiAdj, RQMi, GalAdj
Put 386.4 next to MiFCD
Put 384.1 next to MiSG
Put 22.06361 next to mpgFCD
Put 21.82386 next to mpgSG
Now you will see that RQGal has changed to -1.09. The spreadsheet is suggesting you adjust gallons down by 1.09%. Close as we can get on SGII is 1.1%. So test it by putting -1.10% next to GalAdj.
Now you'll see that the mpgTST field went to 22.07. This is what SG would have reported WITH the calibration. So the spreadsheet got you a lot closer (22.07) than the normal calibration methods (21.82).
Now since distance is borderline we could test what would happen if we did a 1% adjustment on distance. So since SG underreported distance by 0.59% we could round up to 1%.
Put 1% next RQMi
Put 1% next to MiAdj
The spreadsheet now suggest dialing gallons down by 0.10%. So try putting -0.10% next to GalAdj, and you will see the prediction changed from 22.07 to 22.06.
So those are two calibrations you could try in this hypothetical
{Speed = 0%; Gallon = -1.1% } = 22.07 mpg
{Speed = 1%; Gallon = -0.1% } = 22.06 mpg
both of which are trying to zero in on your tank numbers of 22.06361
I'll be making some revisions for averaging soon... So stay tuned.
11011011