35 real world mpg from a MK3 VR6...No? The Toyotecwerke economy Ankor w/up to 38mpg.

From time to time the question of fuel economy pops up for vehicles fitted with "large" engines. One such vehicle is the Golf fitted with a VR6 engine.



Personal experience has shown, my 60ish miles journey into work everyday, would suggest that such a vehicle is capable of achieving an average of 27-29mpg. This reduces down to 22mpg if the grunt of that big 2.8 must always be heard. The typical drive cycle involves B and A roads then some Motorway then A and B roads again. But is there more fuel economy to be achieved?
With fuel prices rising there was no better time to find out.
In the early days of VR6 ownership, long before the ECU was recalibrated, this car was able to achieve up to 31mpg on a similar journey from work to home or work to home, if driven carefully.



A better number on a full trip was not achieved until now.

"Careful driving means" driving within the speed limit, anticipating manoeuvres to avoid large throttle openings, keeping in the right gear to avoid engine lugging and limited revs to a maximum of 3K rpm.
Late in the month of Dec' 10, I was not happy that the mapped WOT fuel mixture had offset richer than desired. This was seen in my last dyno check. After some calibration tweaks to the M3.8.1 ECU, the WOT rich mixture issue was addressed and while the map was being built as yet another iteration, improvements where made to see if the part load fuel economy could be improved. The EEPROM was flashed in the programmer and returned to the car and the vehicle evaluated for response and full load fuel mixture for a few weeks.



It is not known if there are any dyno improvements at WOT is the car has not been restested since, but transitions from low part throttle to higher part throttle to WOT seemed much snappier. This suggested there was more off WOT torque or more optimisation under the WOT curve. With the car feeling right at higher loads it was decided to drive home carefully and determine if the old best of 31mpg could be beaten.
As already mentioned a previous drive from work to home yielded 31mpg. The vehicle was started in the car park and not warmed up. The drive was some fast A road, Motorway, then some B roads.
The most recent drive took the same route. But the vehicle was left to warm up for 6mins or so, which did not help initial FE. MPG checks where taken at 20miles from destination.

33.7mpg average at 39 miles into trip



35mpg average at 50 miles into trip



35.7mpg average at 57miles into trip.



35.2mpg final average at 59miles end of trip.



Evidence of engine running time and distance and average speed.






In conclusion it would appear the calibration changes to part load, in addition to careful driving have improved fuel economy up to 15% at some stages of the drive cycle from previous testing and at the end of the trip the average was a 13% improvement.
I never would have believed it was possible to achieve up over 35mpg average for 60miles in a VR6 until now.

 

just wait till summer, that big phat A/C compressor will claw back the MPG figure

 

Could there be more to be gained with more refining using some dyno time?

I guess an ideal would be live mapping on the dyno in real time but this would require either an emulation package or probably the cheaper/easier option would be to go to megasquirt to control it.

 

Impressive stuff,better than my std AUQ when driving to save fuel.

 

No not for fuel economy with the resourses I have. Dyno time tends to be good for high speed high load optimisation. Already spent time on there previously and know where the WOT LBT mixture needs to be and can determine off dyno.
The rest of adjustments come from studying the logic behind the ecu and making calculated changes.
You can live map these, as well as early MK4 Golf and 3B Passat pre 2000 but it involves obtaining a 16 bit emulator = <1400. There is another option for live emulation which may be a 300 option. Working on it.
Mapping update takes 10 mins and occurs after logging certain channels on VAGCOM, providing the label file is any good. So just a bit longer than emulating.

 

Indeed big nasty thing that... 27-31mpg probably ~280-300miles per tank on these settings.

 

I'd be interested in knowing about the 300 live emulation.

When you say label file is that similar to a damos file or winols project file? Just a list of labels for the different maps in the ecu?

 

No label file is what is used in VDS.

 

Ahh ok with you their now.

 

I assume the MPG figure is going purely by the MPG output readings from the cluster?

I never take any notice of the mpg reading from the mk3 cluster in my mk2 but I used to see 30mpg on a 70mph motorway run and more if going steady. I assumed MPG readings from clusters tend to over-read.

 

Possibly, But you can still see there is an improvement.

 

Changing the subject slighly. Looks like your Top Mounts are duff like Mine.

 

think that pictures old, I don't think there has been a cone filter on this car in a while

 

Well spotted That was probably pre Suspension upgrade.

 

Yes old picture when the car was just bought. Runs factory modified air box.

 

The MPG reading is linked to tank level, wheel speed and ECU fuel comsumption output.
I have actually calculated this by running the vehicle to near dry and brimming tank then attempt to see how many miles I got for my purchase. So far the maximum was 313 miles at about 31 mpg which was the last near enough the displayed value. If there is +ve error in the cluster then it will have this error consistantly.
On my return trip to work I will see if the 35mpg+ indicated can be repeated, then I will dry and refill the tank and try again. I am sure there is an improvement though.

 

If you can do this with a 12v motor is it likely that you could get a better reading with the 24v motor.

 

Pfft...one set of valves at a time....
Actually started looking at the ECU stuff for them too.

 

Toyotec in "that's too many valves" quote shocker!!

Nothing wrong with those strut tops, that is what happens when an anchor is installed.

Technically, a VR6 actually covers less miles than a 4 cylinder engined car over the same journey. This is because the sheer weight of a VR6 engined vehicle actually deforms the earth's surface to the point that the circumference of the globe on which the vehicle is running is reduced, resulting in less distance required between the start and finish points of a journey.

 

So you just tap that chip with an allen key? I might try this tonight!!!