Tuning VE for F/I engine - Main load Alpha-n

Hello everyone,

Currently i'm tuning a 3s-gte (toyota) with the VE model on a link extreme ecu (Current map as attached).

Injector data is correctly set resulting in VE matching MAP throughout the entire rev-range (100kpa corresponds with 100%VE).

Fuel/ignition already mapped under lowest positive boost pressure (30kPa).

My question is as follows:

-As MAP rises above 100kpa, VE also increases above 100%.

This is my first attempt on the VE-model and altough everything works fine untill now, I'm not confident this is the correct way so i'm looking for the best and correct approach to handle further mapping under high boost. Hopefully someone can provide me with good advise/experience?

Some further system info:

-Main load sensor is TPS (this gives us the best results in terms of response)

-Fuel system is MAP referenced

Thanks in advance!

Sorry I can't open your map at this time, but can you explain why your load signal is throttle instead of MAP? Do you have individual throttle bodies or something? When you say you get best result in terms of response, how did you decide that?

Hi Raymond,

No, i don't have itb's..

I know that many of you prefer map as main load but in my opinion/experience:

-AFR remains more constant this way.-

-Map as a load-signal assumes VE is only based upon pressure - which is obviously not the case since the throttle plate forms a restriction.

which gives a - however small - delay (Action/prediction from TPS vs. Measurement from MAP).

And since i'm not planning on organising/participating any competitions at Hillclimbing, I prefer to do it this way in most applications.

*Additional benefit is ofcourse the pleasant way of tuning

It does not sound like you have experience with the TPS-setup on F/I Raymond?

I would change the load equation source to load=MAP, I wouldnt normally recommend the BAP/MAP xover mode you are using, unless it is really needed for large camshafts etc.

Also as Raymond mentions, TP may not be the best option for the look up axis of the VE table unless there is some specific reason to use TP such as ITB's or unstable MAP signal. If you want to stick with TP load axis then you will probably need to enable a 4D table referencing MAP as VE will usually change a little with boost level. There is a webinar here somewhere explaining this tuning method.

A couple other things I noticed in your tune:

If you want to use TP on the VE table you should have some finer break points at small throttle openings where air flow changes abruptly. I typically do 2%,4%, 7%, 10%, 15, 20 , 30, 50, etc.

Turn off the IAT trim table, in modeled mode this is already part of the air mass calculation based on approximated charge temp.

It looks like you have a fuel pressure sensor connected so you should change fuel system type to "FP sensor", the ECU will then try to compensate for variations in differential fuel pressure.

When you last saved your tune file there were 2 trigger error counts recorded and also if you look at your recorded ecu statistics you will see a max RPM of 14000 recorded, this suggests you may have a trigger problem. Start by doing a triggerscope to confirm the sensor polarity is correct and the arming thresholds are suitable.

Fuel charge cooling coefficient could do with a suitable number in there - there are instructions in the help file how to test this.

I like to use throttle position as the main load (y-axis) for the VETable, even on turbocharged cars. Only exception is if you don't trust the throttle sensor to always represent the same angle at low throttle positions, for instance is if the TPS will be removed and reinstalled often for some reason. The difference in airflow between 0-30% throttle tends to be pretty significant, so it's important to ensure the throttle position sensor continues measuring those angles consistently after you've tuned the fuel maps.

I'm not familiar with the VE or fueling calculations used by Link, but I wouldn't expect the engine's volumetric efficiency to match the manifold pressure exactly. On our AEM systems that use VE, the shape of the VE table should usually look like the shape of the torque curve. Smaller numbers at low throttle angles, larger numbers at high throttle angles. At full throttle VE will usually have a plateau of high numbers at the RPMs where the engine makes max torque, then drop off at higher RPM where the engine torque is dropping. Also note the VE calculations are just one part of a larger fueling calculation; if the MAP reading is slightly wrong or the injector data isn't correct you may find yourself compensating for that with odd-looking numbers in the VE table.

I would have thought if everything else was working that a fairly flat 100% VE on a production motor would indicate differential fuel pressure isn't what you are assuming it to be or if you are using fuel pressure input, that the injector flow rate is less than nominal.

@Adam:

-I was already started making a 4d MAP overlay, thanks for confirming that!

-I will make sure i'll adjust the tps breakpoints as you have recommended.

-For the IAT compensation: didn't know that was included for the modeled mode, I will change this

-I've had some difficulties with my cranksensor indeed, well noticed!

-The fuel charge cooling coefficient was new to me (left this zero in first instance because all the other info to process regarding VE).

@Scott:

-I will always make sure to calibrate the tps sensor after a re-installation. Never had any problems with this as long as the limits for the throttleplate stay the same:).

-Indeed, after writing the initial post i came pretty quick to the conclusion my VE/MAP relation was just pure coincidence.

@Slides:

I now also suspect the fuel pressure has something to do with my findings. Will give you an answer on this as soon as i'll have some spare time for testing and I have replaced my FP-sensor (it suddely decided to quit on me while testing).

Thanks for your input guys, appreciate this!

I will post an update on this.

MAP as a load signal is pretty straightforward with a fixed cam engine, especially if it's nonturbo but certainly if it's turbo as well. It is more of a hassle dealing with engines that have variable cams. What is the usage of the vehicle? Is this an occasional weekend street car, trailer-only race car, daily driver (seems unlikely with its age)? Do you ever go to higher or lower altitudes with the car? Are you planning to be running closed loop wideband o2?

Keep in mind that a throttle is a butterfly valve, which is a non linear device. Once throttle opening rises over about 30-40% it is basically at WOT. At that point your airflow is regulated by boost pressure in a turbo application.

I have just found the cause of my problem.. My Map-sensor data was divergent:(

During initial setup i have checked the bosch 4b Map-sensor (which matches the one i use).

I have not given it much further thoughts since it matched baro-pressure.. but the sensor I use has other characteristics.

@Raymond:

It's a competition car raymond (700bhp gt-four) and it will run closed loop wideband as i always do.

I was planning doing a 4d compensation for MAP anyway to account for the higher tps/map values.

The car will be used in the national championship (Netherlands - which is almost as flat as a pool table), so no necessity for altitude compensations.

Again thanks for all suggestions!!