Maxidot MPG - How is it worked out?

[EdmundBlackadder]

Being a bit nerdy when it comes to MPG I have been checking my actual MPG each time I fill up. So far the real figures are showing, based on brim to brim measures, 42, 43.5, 43.5, 44.5, 45. The Maxidot for each full tank has been showing 48-49mpg quite consistently. This is making the reading anywhere between 4-7mpg out. This seems pretty poor to me. How does the computer work out your supposed mpg? Did I have false hopes expecting the computer to be accurate within 1-2mpg?

[Agerbundsen]

It's a bloody good question - and I do not have the answer, but a few comments.

One is prone to assume that brim to brim measuremets are accurate, but how do you know they are? I am not sure of the degree of calibration of the pump counter, but there is definitely a possible error in the cut-off point of differrent pumps and how much the pump is manipulated after the initial shut-off. Variations of a liter or more would not be unusual!!!

Add to that the inaccuracy of the speedometer and the associated question of whether the odometer derives it's values from the same source as the speedometer?

Tyre diameter is another variable, but effects both the maxidot and brim measurements the same.

Finally, when the vehicle was tested on the rolling road for "official" consumption, how many of these variables were corre4cted, or was it jsut measured from an accurately calibrated rolling road and accurate fuel measurement - bypassing all the built in vehicle specific errors?

My take on this is that the official measurements were probably accurate within the specifications - which were NOT driving under actual conditions, but optimized. The real life measurements are only relative - not absolute.

Finally, the official values can be beat by driving optimally to beat them. I have demonstrated better than 54 mpg on a 100 km A-road run - but that's not how I normally drive. That one ought to is another question altogether.

[900000]

At the end of the day you won't prolong your life if you know your MPG is 45 or actually 41 or you thought you drove at 70 when in fact you really are doing 63. So what? I see it as a general rule of thumb of my car's fuel consumption. I've never ever done the brim-to-brim method since I just go by what the car says. Average of 20 means I've been bombing it and an average of 120 means I had the wind in my sails. That is all I need to know. That it might be slightly out this way or that, and even differently between a Shell or a BP pump does not really bother me.

[EdmundBlackadder]

900000 - I know what you are saying but there is no point having a device that is there as a guide if it is wildly inaccurate. Why have any measuring device if it is not close to the actual figure? I did not expect it to be bang on but I had hoped it would be only 1-2mpg out. Surely computers have advanced to a point where it can manage that? Once I knew the margin of error and could trust the computer I would then not have to bother any more doing a brim test.

Is it critical to know your mpg? Not really but fuel is expensive and I like to see if I can drive in a certain way and improve the consumption. It is also interesting to see how the car develops as it loosens up.

[900000]

900000 - I know what you are saying but there is no point having a device that is there as a guide if it is wildly inaccurate. Why have any measuring device if it is not close to the actual figure? I did not expect it to be bang on but I had hoped it would be only 1-2mpg out. Surely computers have advanced to a point where it can manage that? Once I knew the margin of error and could trust the computer I would then not have to bother any more doing a brim test.

Is it critical to know your mpg? Not really but fuel is expensive and I like to see if I can drive in a certain way and improve the consumption. It is also interesting to see how the car develops as it loosens up.

Oh don't get me wrong, I would also like to know how it actually works this out as I'm as much a nerdy geek on these things myself! I just won't cause myself grey hairs about it is all I was trying to say.

[Terfyn]

One thing you can be sure of is the accuracy of the pumps as they are calibrated against a standard volume by Trading Standards. The errors creep in with the filling and topping up.

the most accurate way is to carry out a running total. As the miles add up and the fuel is used then the errors of inaccurate filling and speedos reduce to a negligable level.

So keep a diary. Put in the fuel reading and the mileage over, say, 5000 miles. Divide the total miles by the total fuel for a definitive MPG reading.

Maxidot is but a guide. The higher the Cumulative MPG the better it is on the pocket.

Edited 31 August, 2010 by Terfyn

[countryboy]

Basically you need two things to work out the fuel consumption, a flow meter to find out how much fuel is flowing to and being used by the engine and a distance measurer, i.e. the odometer. From these then the fuel being used per mile or kilometer can be calculated.

Ian

[Terfyn]

Basically you need two things to work out the fuel consumption, a flow meter to find out how much fuel is flowing to and being used by the engine and a distance measurer, i.e. the odometer. From these then the fuel being used per mile or kilometer can be calculated.

Ian

Isn't that what the Maxidot uses to give MPG or Ltr/km etc. It is just that they are not as accurate as we would like.

The summation method reduces the error by providing much larger quantities to work with. The error is reduced to an insignificant level. Standard statistical practice.

Edited 31 August, 2010 by Terfyn

[Agerbundsen]

900000 - I know what you are saying but there is no point having a device that is there as a guide if it is wildly inaccurate. Why have any measuring device if it is not close to the actual figure? I did not expect it to be bang on but I had hoped it would be only 1-2mpg out. Surely computers have advanced to a point where it can manage that? Once I knew the margin of error and could trust the computer I would then not have to bother any more doing a brim test.

Is it critical to know your mpg? Not really but fuel is expensive and I like to see if I can drive in a certain way and improve the consumption. It is also interesting to see how the car develops as it loosens up.

I am sure the computer can add, subtract and multiply and divide, but the old adage of siso (**** in - **** out) data collection still applies. The distance input data are just inherently unreliable so that the data output can only be considered relative, not absolute. It's the same for on-board measurements as the vaunted brim-to-brim measurements.

Bottom line: Don't kid yourself into thinking that one is more accurate than the other. 't aint!

Anyway - by a little judicious attention to my driving habits, I can beat the economy of my old 140 hp PD Octy Combi. Not bad going - but then, i don't often do it, it's toooooo much fun not to.

Edited 31 August, 2010 by Agerbundsen

[shark_90]

It works it out by distance travelled versus injected quantity which is worked out by the ECU. Which is why so many remaps cause the trip computer to be much better than it should be, because some tuners change the injector calibration.

[Terfyn]

I am sure the computer can add, subtract and multiply and divide, but the old adage of siso (**** in - **** out) data collection still applies. The distance input data are just inherently unreliable so that the data output can only be considered relative, not absolute. It's the same for on-board measurements as the vaunted brim-to-brim measurements.

Not so. The large volume of fuel used for measuring starts to make the variance of the "brim to brim" filling insignificant. Example: If you put 60 litres of fuel in every time you fill but are inaccurate by 0.5 of a litre then in:-

120 ltrs inacccuracy is 1 ltr

240 ltr inaccuracy is 2 ltr and so on Add to this that errors in filling can be positive as well as negative then the volume of fuel used can be relied on. The odometer reading will be within set limits but will probably be biased to one side of accurate.

As the volume of fuel put in increases the relative error reduces because the error per fill is fairly constant. Same with the mileage, if your readings are out by 5 miles in 5000 the MPG reading will hardly change.

Edited 31 August, 2010 by Terfyn

[Agerbundsen]

Not so. The large volume of fuel used for measuring starts to make the variance of the "brim to brim" filling insignificant. Example: If you put 60 litres of fuel in every time you fill but are inaccurate by 0.5 of a litre then in:-

120 ltrs inacccuracy is 1 ltr

240 ltr inaccuracy is 2 ltr and so on Add to this that errors in filling can be positive as well as negative then the volume of fuel used can be relied on. The odometer reading will be within set limits but will probably be biased to one side of accurate.

As the volume of fuel put in increases the relative error reduces because the error per fill is fairly constant. Same with the mileage, if your readings are out by 5 miles in 5000 the MPG reading will hardly change.

Terfyn,

No arguing with your math. However, if you fill in 60 liters each time, yu probabaly have to push the SM to the filling station . So on smaller fills, the math becomes more troublesome - if you look at one filing.

If you add up both the gallons filled and the miles covered over longer periods, then the inaccuracies revert to the miles covered - again, assuming they are he same measurement, tires etc. the inaccuracies also are the same, but they are still not absolute, only relative, and do not correlate with the official figures.

It's all academic in any case, and I am not at all dissatisfied with the numbers I read in my case.

Edited 31 August, 2010 by Agerbundsen

[Llanigraham]

3 tanks full of fuel added to mine now.

Consumption has been worked out by dividing amount of fuel into trip mileage gives one figure, which I have then compared to Maxidot consumption average #2. Difference has been around the 10% mark, so quite within statistical accuracy.

Better than I was getting in the past, so I am happy!!

[shrub]

My trip computer has so far been accurate to between 0.5 and 2mpg per tankful. I have used the same pump to fill the tank and so my calculations are as accurate as I can possibly make them. Pretty impressed with the accuracy of mine although given that my screen two resets itself halfway into every other tankful I can't vouch for longer term accuracy. On a tank by tank basis it is pretty impressive though.

[the_raz]

I agree with Shrub. It seems that I am even more geeky than any of you : I have done some testing to find out by exactly how much the Maxidot MPG figure is out, and after a couple of tanks I found out that my Maxidot is off by 1,5 % (the actual cosumption 101,5% of the consumption stated by the Maxidot). As I have access to a VCDS cable I entered 102% for fuel consumption correction in the incar computer, and now when comparing consumption figures over a full tank I get the same figure from my OBD! So even the "stock" setting is fairly accurate as long as you have standard tire dimensions and the car has not been remapped, and any greater errors are probably coming from differences in measuring the amount of fuel consumed.

The testing has been done by using the fuel pumps initial fuel cut-off(always on the same pump), as I assume this is the most correct measurement I can get. I tried filling it up to the brim, but it seem that (at least on 1.2 TSI) the fuel slowy poors down to fill different small cavities in the tank, and I was getting wildly irregular data based on how long I could stand putting a few cups of fuel in the tank and watching it slowly flow down... ... It actually takes several liters after the first click, so this is a great source of error.

the_raz nerd

[Agerbundsen]

Raz, geeky is good, I have learnt something. Interesting that the mileage values can be adjusted.

It still does not alter the difference between the official mileage measured on the rolling road and real life. The rolling road values are only really useful as a guide for comparing different vehicles - not necessarily repeatable in real life.

Edited 1 September, 2010 by Agerbundsen

[Delboyyy]

I have tracked my actual fuel consumption since new (sad I know but it was for work purposes) and also the accuracy of the Maxidot reading.

I have now covered just under 8,000 miles of mixed driving since April and always fill until fuel pump cuts off. Use the same petrol station 90% of the time but not necessarily the same pump but I think my readings are fairly accurate now.

1.2TSi engine: Overall average is 41.47 MPG. (18 fillings of fuel at approximately every 440 miles).

Best average on a between fuel fills has been 44.1MPG and worst 38.9MPG.

The Maxidot seems to indicate a slightly better fuel consumption than actual but I have to say I have been very impressed with the accuracy in my Yeti as it averages around +0.5 MPG error. Best has been 0 and worst +0.84 MPG error from actual and this may be down to pump cut out and other factors meantioned in this thread.

I previously had a Golf Mk5 with a 1.6fsi engine and the fuel computer on this (probably a very similar vag unit) always indicated around 2-3 MPG higher than actual.

Del.

[Simon_B]

All very detailed! I simply use my ave consumption figure as a relative guide. I've been driving my 110 diesel Yeti for 2800 miles now with an ave overall consumption of 51.5. I use the 2nd altenative consumption log for shorter journeys. It regularly gives me a reported 59mpg driving fast A roads and motorways at the legal speed limit using cruise control. Seems more frugal than my previous Honda Civic 2.2 diesel (though not so fast!). I've got used to a different style of driving. Great car! Always feel good even after long journeys!

[kenfowler3966]

I have just done over 3000 miles to date and all fuel is bought on a fuel card via work, so I have an exact record of what I have bought from new. My 110bhp 2wd diesel is showing an overall average of 50.7mpg on the maxidot which has not been reset since delivered with 7 miles on the clock. Actual fuel consumption is 46.5mpg. Bearing in mind the speedo is also about 5% out and over-recording miles according to the satnav, true consumtion is really likely to be around 44mpg. (I have taken account of the initial fill and always fill to the brim)

[rockhopper]

I remember having a debate on the octy 2 forum about this same 'problem'. the maxidot was always giving much better results than a brim to brim did. Ali I know is for an auto it was magnificent for a tweaked 140PD engine. On a long journey, of 300+ miles I could get 64mpg, travelling at 60 up the motorway, but the brim to brim was low 50's. I wasn't complaining as it was better than the official figures. (that trip was boring!).

Mike