1.9 TDI 105cv from 44.7 mpg to 53 mpg higway

[ausias]

How I increased my mileage from 44.7 to 53?
I used:
-A lot of time reading pdf and forum about the 1.9 TDI PumpeDuse Volkswagen Engine
- Some tools 13mm and 18mm explained in many tutorials (visit Fred's TDI Page. TDIClub.com. VW TDI Enthusiast Community or AreaVAG and search for "cam timing", "angle syncrho", "torsion", vag com, vcds...).
- Gloves.
- Aluminum Foil
- Power tape
- 300-400 km of testing and data logging with Vag-Com 311.2 and a Notebook.
- 22 km of non-congested flat Highway, no wind day or night. Start point the same as Finish point allowed ensure there's no climbing each loop.
- A Volkswagen AG car equiped with PümpeDuse Engine.

I started measuring the fuel consumption of my car. After this I blocked the upper grill of my car inspired in Ecomotive and racing versions of my car. You can see some pictures in:
http://ecomodder.com/forum/showthrea...ape-19191.html

My results with upper grill open -> closed

82 km/h: 4.36 -> 4.23 L/100km ; 53.9 -> 55.6 us mpg at 51 mph
94 km/h: 4.73 -> 4.57 L/100km ; 49.8 -> 51.5 us mpg at 58.4 mph
122 km/h: 6.67 -> 6.42 L/100km ; 35.3 -> 36.6 us mpg at 75.8 mph

After my camsahft timing fine tuning this are my figures in fuel consumption also with open and closed grill:
3.90@82 -> 3.80@82 ; 60.3 -> 61.9 us mpg at 51 mph
4.23@94 -> 4.02@94 ; 55.6 -> 58.5 us mpg at 58.4 mph
5.75@122 -> 5.50@122 ; 40.9 -> 42.8 us mpg at 75.8 mph

That means :
at 82 km/h: -13% L/100km -> +15% us mpg
at 94 km/h: -15% L/100km -> +18% us mpg
at 82 km/h: -18% L/100km -> +21% us mpg


I'm using Brisgestone ER300 front and Hankook Ventus V12 Evo K110 at rear, that allow my higher stability (little understeer) and better braking because of more rear grip. Sportly tuned, not ecotires of any make. So it's room form more improvements (ECU, lrr tires of bigger radius, aerodynamic cd I think could be lowered easily to 0.30 and with little tuning of underbody maybe go to 0.27 or 0.28 looking almost the same car...).

If my estimations are correct I broke the 40 us mpg barrier, some full tank filling will show me if i succeeded.

[bitmap]

So what "torsion value" did you end up with to achieve this?

[ausias]

The torsion value is less than -3, values out of the range (-3,3) are shown as 0.0 (some autodiagnostig tools show -6 to 6).
But I did a full tank fuel filling and I'm very surprised of 6.6 L/100km average. I'm wondering if there is interference between cam timing and ECU fuel consumption, torque, etc calculations. I did 868 km with 57.72 L of diesel premium, but near 600 km or more were did in higway driving between 80 to 120 km/h, and VagCom logs showed always fuel conmption around 5L/100km. I did some insane and inefficent urban driving and vag-com said 6.26L/100km, how the average went to 6.6L/100km? MAybe first tank runs made with camshaft timing in +2.9 are guilty.
Aerodynamic addons are working fine (central front spoiler from Ecomotive, grill block) but I just fine tuned my camshaft timing to real 0.0, and I will compare. Also I will test some advance timing full tank comparision. But I'm very disappointed, maybe ECU is adjusting to new camshaft and fuel consumption not being affected or misscalculations are hiding the real behaivour.
Torsion values <0 tend to add torque at lower rpms and noticiable loose of max speed (-5% of the initial top speed). Torsion values of =0, gives me 100% of top speed said by the make (and +2% of my standard) with my aero-addons, Torsion values >0 (+2.9) gave me more torque at high rpms ad the same +2% top speed without aero addons, with aero addons only notidec -5% fuel consumption relative but no change in additional top speed (torque curve is limited from ECU software, so max values could not be exceeded.
With 90% upper grill block I get better fuel consumption and better power because of cold air intake, at 82, 94 and 122 kmh. Lower central Spoiler, and fixed right wheel spoiler wich it was broken.
3.65 L/100km @ 82.65 km/h
3.88 L/100km @ 94.9 km/h
5.67 L/100km @ 122.75 km/h

I also get more ECU better results in my 220 kms trips of 5.4L/100km (110km up 5.7L/100km, 110km down 5.2L/100km) at 122 km/h of cruise control (GPS).

But with cam timing between -1.5 to -1, I got:
4.0L/100km at 82 kmh,
4.53 at 94kmh,
6.5 at 122kmh.

And with cam timing in 0.0 (very little times it says -.5):
4.13 at 82,
4.7 at 94
6.65 a 122kmh

These are my logs (with aeromods which lowered my cd from cd=0.325 to 0.3). 0.5 to 1 L/100km differences on different cam timings seems not being reflected in full tank average, what?

So I think full tank or direct fuel measure is needed because of the possible interference on fuel gauge.
Idle fuel consumption Vag-Com channel 15 says in l/h (+-0.2 l/h precison available only):
cam timing less than -3.0 => 0.4 L/h
cam timing -2.5 to -2.0 => 0.6 L/h
cam timin -1.5 to -1.0 => 0.6/0.8 L/h
cam timing -0.5(somtimes) and 0.0 => 0.6 L/h
cam timing +2.9 => 0.8 L/h

torque values go from 26 Nm to 50 Nm needed to idle.

I think all are related values and someones is not a sensor but a derivated value.
If torque values are correctly measured but fuel consumption its a theoretical calculation maybe more torque needed to idle means real less fuel consumption. Or viceversa, but It must be tested without using MFA (that uses OBDII data from engine ECU) or raw OBDII data.
Power/Torque curve is really affected on this value (not maximum showed in OBDII) but more precise (GPS of 1 or 5 Hz with +-0.1 kmh speed uncertainity, or accelerometer, or dyno bank prefered to +-1km/h, 1.5 to 4.5 Hz from ECU)

I've posted answers to some different questions in http://ecomodder.com/forum/showthrea...ape-19191.html , details on testing procedure, drag coefficient assessment, etc.

[ausias]

I opened a thread where instrumentation issues are under research:
http://ecomodder.com/forum/showthrea...ard-22642.html

[bitmap]

Okay, so I did some camshaft adjustment of my own
I have a Skoda Octavia (czech VW Golf) with BKD (2.0l 16V TDI PD) engine.
It has done about 100,000kms timing belt, injectors etc., everything is original.
Before camshaft adjustment my MFD indicated idle consumption 0.7l/h and there was no way I did a tank under 4.4l/100km.
I rotated both intake and exhaust camshaft slightly anti clokwise and wow, what a difference

MFD now indicates 0.5l/h at idle, engine is much stronger from just above idle up until 2000 rpm and turbo spools up good 300 rpm earlier than before.
Car is now much more pleasant to drive and fuel consumption improvement is just unbelievable.

Just after this adjustment I brimmed the tank and MFD indicated 1490kms range. I thought that I somehow screwed up MFD's calibration, but after I did about 250kms run of quite non-eco driving friendly route with lot of hills, trucks and villages and brimmed the tank again, it took 10.6 litres, so it turnes out It used 4.23l per 100km. MFD indicated 4.1l/100km and there is no way I would make this trip under 4.6l before this adjustment. I hope that car's ECU didn't start "imagining" kilometers and this improvement is real, I'll keep you posted.

Sorry for the rant but I had to share this with someone

[Mario_Marques]

I think something's wrong with your car....you should have better fuel mileage than my car and is actually worst.
That "cam tuning" is only fooling your ecu and you will end up damaging your engine because of that(injection timing and pressure in the injector unit is not ok with that "timing"), i've seen that happening, i'm a vw mechanic...

[mikeboy0001]

Hi ausias
I bought a Lupo 1.2TDI 3L, a PD engine, and I'm trying to fine tune it to have the lowest consumption possible
I'm now messing with the torsion values, and would like to know if you can tell me more of your experience in this subject
First of all, I've found that the torsion values given by an European PD model, or an American PD model are oposite. That is, when you retard the ignition on a European model, the torsion values go positive, whereas in a US model the values go negative. And when you advance the ignition on a European model, the torsion values go negative whereas in a US model the values go positive
What I've also found is that the values displayed by VCDS in consumption change a lot when messing with it
With a torsion value of +4 (retarded) the consumption at idle shows an injection of about 3.6mg/st, 0.3l/h. With a torsion value of -4(advanced) the consumption at idle shows an injection of about 6.0mg/st, 0.6l/h
Now, this makes little sense to me, because theoreticly, with the ignition advanced, the the car would need less fuel for the same power, and the ECU is supposedly adding it. Also with the ignition retarded, the the car would theoreticly need more fuel for the same power, and the ECU is supposedly cutting it
With the ignition retarded, there is a noticeable loss of power at low RPM, and with it at 0 or slightly advanced, there is certainly more torque at low revs
I still haven't done full tanks to accurately measure it, so I would like to know, from your experience, if torsion values in TDI PD's really can change fuel consumption, or if this is only messing with the MFD?

[ausias]

the torsion value must be somewhere between 0.0 and 1.0 advanced (according to VAG engine manual between -3 to 3).
But if you want your engine being in the design point you want 0.0.
But, when you have a value out of range you'll see 0.0, so be careful.

There are loses of max torque and power also when retarded engine (too retarded and engine will not start. Going too advanced and you will blow up your engine, but the bolts will not permit to go too far. I noticed a change in Power from 97 to 107 bhp in the wheel. And in the top speed from 171 to 187 km/h.

Best place for that 1.9 TDI BXE was +0.5 ( 0.5 advanced) althought it wasn'y any differences at all between real 0.0 and this final setup.
+0.5 (-0.5 for your car) was good because in that ECU map the S.O.I. inection timing (in ms.) map (with warm engine) was the best fit and compromise of advancing the entire map for that particular setup. It's hard and long to explain why.

the MFD values are proportional (and you can callibrate them) to the ECU ones (that you read with VCDS), that ECu values are proportional to the voltage applied to the injectors.
There is something wrong in the ECU OBDII values when you change torsion (or some other things) because values that depend linearly with the variable are supposed to be proportional and the VAG ECU give you wrong figures when you change offsets or slopes.

[mikeboy0001]

I'm out of town, so it's harder to catch up with messages
I still haven't done much mileage with the car, but now I've changed the torsion value to an almost exact 0, when hot it goes to -0.6(advanced) on ocasions
I've done this as I've found one of the injectors has a small problem, causing issues on the first start of the day. I've done a Diesel Purge with 2 bottles of Liqui Molly directly on the fuel line, without much results. Now I've made a trip of about 450Km on BP Diesel Ultimate, and it looks better on starts as well on groups 13 and 23 of VCDS, but let's see
When I had the torsion with a small advance (-3), I noticed more torque on low revs, but with the setting 0 it has now, it feels more elastic, or natural
I'm hoping to have the smallest fuel consumption possible, and so from your experience, it isn't worth having the ignition slightly advanced , is that it?

[ausias]

with torsion value you are advanding or retarding the injection in degrees of crank rotation so the injection timing in seconds (or miliseconds) changes depending on what speed (and angular acceleration, boost, temps, and dozens of other values).
Advancing 3º it is not a small adavance, because it means that injectioion pump being pushed to start 6º before and it will represent a large time of advance when low rpms and no noticeable (or worst) changes in performance (in terms of power and/or effiicinecy).
Euro 4 engines and higher have in some regions of the map thing retarded 2º so you only need to fine FINE tune your engine in a range of 1º of crank to compensate, but you are fooling the maths a little on high loads and ranges. IN terms of efficiency sometimes the injection is retarded but the engine has more pressure in the cilinder because of gas recirculation EGR so efficiency is almost the same but with less pollutants being generated.
So big differences in real performance and efficeincy (there are not the same) are between the crank angle too retarded or advanced. And lot of explanation is needed.
To sum up, engine will run and start with Torsion in the range 3º advanced and out of range retarded (I gess a bit higher than 6º retarded).
Volkswagen wants tdi to be between +3 and -3, and some cars have the sensor in the range of -6 to +6. But if you want your injection pump or pump injectors being synchronized with the ECU values and getting all the fuel efficiency and max power you will need to fine tune to +-1º. (For me prefered 0 to 1º adavanced in the 2006 1.9 TDI PD BXE i had)
Elasticity is the way to go because it means that the ECU get from sensors the correct values in all the range so it injects the specified amount of fuel according to air. REmember that your injection is controlled with several limitats and power-torque curves by the ECU. so you are feeling the power that should your engine have. But it was lost because of not fine tuned.

If you want to go further you need to do a custom ECU with pressure sensors, back pressures, etc. maybe Vekke should answer more accurately. http://ecomodder.com/forum/showthrea...-2l-18235.html