Honda Insight Forum banner

1 - 13 of 13 Posts

·
Registered
Joined
·
13 Posts
Discussion Starter #1
Anyone have tips on improving the accuracy of the displayed gas mileage? Mine seems to be roughly 8% higher than reality. Had my best ever tank last week, 53MPG, but the pump knocked me down to 49. Still a good number, but not over fifty like I was geared up for.

I don't overfill the tank, and do consistently fill up at Costco. Don't use the same pump refill-to-refill, but I have to assume they're similar.
 

·
Registered
Joined
·
943 Posts
The most obvious is get some larger diameter wheels / tyres.

You will probably find that the speedometer over reads. Get a GPS / satnav and check your speed on a straight flat road at say 70mph. When I'm really doing 70 in my I2, the speedo reads 73-74. The oedometer is off by a similar amount. If you really want to be sure, drive a known distance from map or gps and check it against the trip computer.

One option is factor this into your mpg calculation. If you want to go the whole hog, change the tyres and wheels for something that will give no error and perhaps be a little more careful about speed limits. Don't forget that brand new tyres have a slightly increased circumference.

For the I2, the truth lies somewhere between oedometer / receipt calcs and the MID.

Sent from AutoGuide.com Free App
 

·
Registered
Joined
·
63 Posts
I am getting near the same mileage. Last week I managed 53 on the display, but the reality turned out to be 49ish. I am actually incredibly pleased. I have my doubts regarding odometer error, my car is brand new, with factory tires and I cannot imagine that Honda would put out a product with a consistent +8% margin of error in the odometer reading. I will be doing an experiment to prove my intuitions though.
 

·
Registered
Joined
·
13 Posts
Discussion Starter #4
If you really want to be sure, drive a known distance from map or gps and check it against the trip computer.
I'll try that this weekend.

llpancholl 49/53 is the best I've managed. 44/47 has been more typical, but that fits the 8% margin of error. Curious what the gps experiment will say.
 

·
Registered
Joined
·
113 Posts
MID Optomistic by design

Whether the odometer is accurate or not does not matter in comparing the MID MPG reading to the odometer distance divided by the gallons used, since the MID is using the odometer value divided by the gallons used as measured in the fuel injector delivery. If the gas usage as measured in the fuel delivery system by the cars computer matched the actual fuel used as measured at the pump, the calculated MPG would be the same. The problem seems that on most, if not all Insights, the computer's measure of the fuel usage is low, so the MID MPG reading is 5-7% high. Over ten recent tanks, mine averaged 5.4% high, and over ten tanks last year, 6.7% high. I have never seen a tank where the MID reading was closer than 3.2% above.
Measuring fuel in small amounts through the fuel system is likely to generate a cumulative error, but the fact that it is consistently off on the low side, generating a higher MPG, as reported by many users, strikes me as a systematic over estimation that Honda was probably comfortable with. Most auto reviewers and owners report the MID mileage reading, and this makes for better press and happier customers. A simple adjustment in the calculation that made it error on the high and low side equally was not in their marketing interests.
I would also hazard a guess that the odometer reads high, it is probably based on the new tires circumference, as measured when unmounted. Over the life of the original Dunlops, a 7/32 tread wear results in a 2% increase in the odometer reading. Tire deflection, making the effective diameter smaller, probably contributes another 2%. So the MIDs reading compared to actual mileage as measured by a GPS is probably off even more.
 

·
Registered
Joined
·
943 Posts
Coincidentally, some models of Civic underestimate mpg compared to calculated but that's fairly rare.

I'm inclined to believe that Honda knew precisely what they were doing with the MID on Insights. High mpg means happier customers so they will err on that side within the allowable error within law. We all seem to be getting similar error margins. It wouldn't be hard for Honda to calibrate more accurately if they wanted to, would it? It's easy enough with a scan gauge and that uses the same base data.

I drive a lot of hire cars while using satnav. All over-read by 3-5 mph at 70 mph and all over-estimate mpg compared to calculations at the pump.

It wouldn't surprise me if the speedo, oedometer and mpg all have differing correction factors on some cars. On my old Skoda Octavia SLX tdi you could hack the climate control to get the road speed reading from the ECU. This was bang on accurate, but the speedo on the dash under-read. The car knew how fast it was travelling but it wasn't in the manufacturer's interest to pass that info to the driver.

Edit: Over 2 years and roughly 18,000 miles I've done 51 / 47 mpgUK.

The speedometer / odometer will probably give about 4-5% error. That only gets half way there.

Sent from AutoGuide.com Free App
 

·
Registered
Joined
·
4,006 Posts
Think about it, most people aren't going to actually calculate how many gallons go in. They'll see 47.2 mpg on the display and that's what it'll be in their minds. No matter if when they fill up it was actually 44.0. They'll pay the fill up without looking at the number of gallons and be happy about saving gas.

The human brain is a funny thing like that.
 

·
Registered
Joined
·
13 Posts
Discussion Starter #9
Last night, did a partial test. Used a GPS to verify the speedometer at 40mph. It never varied by more than 1mph, and that would only be for a second or two.

I did not drive a previously known distance. It was around 7 miles, but whether 7.0 or 7.5, I do not really know. But it's hard to believe both GPS and car could get speed right and distance wrong.

That leaves the fuel system wrongly reporting how much fuel it has consumed? That sounds unsolvable, doesn't it?
 

·
Registered
Joined
·
605 Posts
That leaves the fuel system wrongly reporting how much fuel it has consumed? That sounds unsolvable, doesn't it?
yup.

I sent a request to fix, to Honda, a long time ago. ..Ignored of course.

Also, I think a source of error in the actual mpg figure, lies in the process of filling the gas tank...i.e. what exactly is a full tank of gas? ...how much gas has actually been used ...leaving aside any inaccuracies in the gas pump #gals. readout.

To address this (which can be plus or minus 1.5 gal.) all one has to do is, to look closely at the fuel level just under the gas filler cap. Make sure that the fluid level is EXACTLY at the same point as it was at the prior fill up, before recording the #gals. used (indicated in the fuel pump read out). Without checking this carefully, actual gals. used by the engine could be a little higher (hence lower than actual mpg being recorded) or lower (hence higher than actual mpg being recorded). Fouling up the actual/MID mpg difference.

Here in Oregon self service fillups are not an option. So I've tried to correct for this source of error by recording 5 tank fulls of "actual" mpg and MID mpg. I then subtract and find the difference between the two for each fillup. I then compute the average difference for the 5 fillups. That number is 2.92, with the MID higher than "actual".

The outcome is that whenever I'm riding along and see an MID mpg, I just assume that number (whatever it is) is 3 mpg higher than the actual or "real" mpg I'm getting.

FYI
 

·
Registered
Joined
·
1,814 Posts
In my gen 1 insight the mpg is dead on for 75 mpg tanks, anything higher and the car actually shows a little less than the pump. Anything under 75 mpg and the car indicates slightly optimistic. Absolute worst case For a tank over 30k miles logged was about 5% off, typically within 1.5%, but it varys tank by tank.
 

·
Registered
Joined
·
2 Posts
optimistic mpg

just bought a 2010 insight 1 month ago with 15k. i have always done the mpg with all my cars. the only car that had a readout that matched real math was my 2001 acura MDX. after 5 fillups this insight is a full 10% optimistic on the dash calculator. oh well so much for manufacturer integrety
 
1 - 13 of 13 Posts
Top