Trip computer accuracy

[BTB 500]

Top Gear mag. has a long-term W204 C220 CDI, which they think is fantastic.

Interested to see though that they reckon the trip computer is a rather optimistic. Over 2,371 miles their fuel added/distance covered calculation gave 36.3 MPG, whereas the trip computer said 38.8 MPG (nearly 7% out).

 

[stats007]

Isn't it speedo based - whereas TGs are from GPS data?

 

[BTB 500]

I guess they use odometer mileage for their calculations - which should be the same distance data the trip computer works with. So a difference in MPG would be down to fuel usage ... maybe missing injector pulses from time to time (if that's how it's measured by the trip computer)? That would explain why the MPG figure was higher.

 

[stats007]

In the tests Evo do it's done using GPS data - as far as I know the fuel consumption just uses the odometer and the fuel gauge measurement.

 

[BTB 500]

These are long term test cars though, that they keep and use daily for 6-12 months. I'm sure they just record odometer mileage.

Fuel consumption goes up and down as you drive along (many other cars give an 'instantaneous' MPG figure), so it must use some kind of flow measurement. I was guessing that counting injector pulses (from the engine management system) would be a likely method, rather than a flow meter in the fuel line.

 

[oldcro]

Top Gear mag. has a long-term W204 C220 CDI, which they think is fantastic.

Interested to see though that they reckon the trip computer is a rather optimistic. Over 2,371 miles their fuel added/distance covered calculation gave 36.3 MPG, whereas the trip computer said 38.8 MPG (nearly 7% out).

What about the errors in the fuel used?
Petrol pumps are NOT accurate, they are accurate within acceptable tolerances, but could deliver more (unlikely) or less than a Litre. The error maybe minute, but over hundreds of Litres does add up.
Secondly, if the temperature of the fuel delivered changes from fillup to fillup then the amount indicated also changes.
At the end of the day these calculations are all an approximation.

 

[Dieselman]

I was guessing that counting injector pulses (from the engine management system) would be a likely method, rather than a flow meter in the fuel line.

It does. A given pulse width delivers a given volume of fuel thus it can be accurately calculated.

This is why a tuning upgrade can cause inaccuracy due to the calculation no longer being correct.

 

[Mactech]

What about the errors in the fuel used?
Petrol pumps are NOT accurate, they are accurate within acceptable tolerances, but could deliver more (unlikely) or less than a Litre. The error maybe minute, but over hundreds of Litres does add up.
Secondly, if the temperature of the fuel delivered changes from fillup to fillup then the amount indicated also changes.
At the end of the day these calculations are all an approximation.

Quite right! The biggest area of aproximation is in the calibration of the injector flow rate. Production tolerances allow these to vary as does the fuel pressure. The computer is just adding up the time each injector has been open and assuming that the flow rate is correct. Most of the add on 'chips' to increase power on the diesels increase the fuel pressure so sending the computer out by approx. the % fuel/power increase.
That said, the fuel computers on the latest V6 diesels are at least as accurate as my attemtps at measuring the fuel into the car...

 

[BlackC55]

Remember gents.......Its only a guide.

 

[stats007]

If the ECU controls the pulse width (from a map) and this value is used for calculating fuel consumption - tuning would be irrelevant?

 

[Mactech]

If the ECU controls the pulse width (from a map) and this value is used for calculating fuel consumption - tuning would be irrelevant?

Most add-on boxes to increase performance of diesels increase the fuel pressure without reference to the ECU map. So each injector pulse injects more fuel (due to the pressure increase) than the fuel computer assumes so skewing the figures to make the car appear more economical (by the computer) than it is in reality.

 

[stats007]

Ahh - didn't think about that - tend to think in petrol terms!

 

[BTB 500]

What about the errors in the fuel used?
Petrol pumps are NOT accurate, they are accurate within acceptable tolerances, but could deliver more (unlikely) or less than a Litre. The error maybe minute, but over hundreds of Litres does add up.
Secondly, if the temperature of the fuel delivered changes from fillup to fillup then the amount indicated also changes.
At the end of the day these calculations are all an approximation.

The legally allowed tolerance on a forecourt pump is + 1% and – ½ % (i.e. maximum error on a 50 litre fill-up would be + 0.5 litres or - 0.25 litres). They are checked by Trading Standards fairly regularly.

Fuel temperature affects density and can be a factor (hence recent rows in F1 about some teams abusing minimum fuel temp. rules). It's much less of an issue in most filling stations though where the storage tank are buried underground and the temp. stays fairly constant.

 

[crockers]

The legally allowed tolerance on a forecourt pump is + 1% and – ½ % (i.e. maximum error on a 50 litre fill-up would be + 0.5 litres or - 0.25 litres). They are checked by Trading Standards fairly regularly.

Fuel temperature affects density and can be a factor (hence recent rows in F1 about some teams abusing minimum fuel temp. rules). It's much less of an issue in most filling stations though where the storage tank are buried underground and the temp. stays fairly constant.

I never realised they were allowed to be -0.5% - I thought that if the erred it had to be on the plus side only................. well we all live and learn..