G4 Xtreme 2007 USDM STI runs leaner in low gears, best solution?

[Eric Godden]

I installed my G4 extreme not to long ago and have been slowly tuning the fuel map in my free time. One issue I have run into is that when the car is in say 4th gear the fueling is spot on (14.7 cruise etc), however when in 1st and 2nd gear the afrs will be much leaner than they should. Say 16afr etc. This occurs most times during clutch release for accelerating from a stop and also when moving slowly in say those gears. I haven't gotten into doing any power runs all loads are below 0MGP. If I tune the main fuel map for proper AFRs in the lower gears , my cruise afr is now 14afr. I have messed with the accel enrichment table but this seemed to have little or no affect as it runs lean at constant throttle openings not tip-in. Has anyone else experienced this problem and what was your solution? Thanks

[martin battye]

i think unless you have missed something you'll need to set up an overlay table for TPS vs rpm.

i first found this trim useful when tuning a supra years ago on AEM that had a large TB. no matter where you tried to retune the cells would always fall into two running areas and cause AFR differences...

[Eric Godden]

I will give that a go thanks

[Eric Godden]

Making some better progress using a 4d table tps/rpm to add a bit. However I'm a little lost. Should a I tune the fuel table in say 3rd gear for as many cells as I can with this 4d table off and then adjust the table to get 1st and 2nd gears correct or some other avenue. Essentially the problem I am running into is if I adjust the 4d table to get 1st and 2nd where I want them then if I happen to be crusing in 5th or 6th my afrs are 13.5 or so instead of 14.7. I'm trying to avoid using closed loop wbo2 as I consider that a patch. Any input?

[martin battye]

i meant use the TPS overlay for low throttle angle correction... ..

 

the fact your running leaner in lower gears is most likely due to the engine load being slightly different.. (most likely seen when you use a fixed duty for boost control).. thus the ecu reading from slightly different cells on a WOT pass. Log MAP and see the difference...

[Peter Giljevic]

I have had similar situation with my Mazda RX7 if that makes you feel any better, and to add I tune every type of ECU commonly available and the Link G4 is the only one ecu that I have experienced this on to date, it is very complex and time consuming to set up 100% correctly for all operational ranges. All I can say is you will need to use all of these extra dimensions to get it to work properly, and it is pointless to use the closed loop AFR trim feature until you have the map working as it should without any AFR input loop. I need to stress so long as all influencing variables are in check and not influencing your situation (as they did not in mine, its a pure ECU function quirk) but its worth stating as many issues can sometimes be other systems playing a part.

[Guest Dave]

Hi Eric

Can you please email me a copy of your tune so far as most times it is simple things not setup correctly

Like Peter removing the load axis off the default air temp tables on his car    some other ecus do corrections behind the scene for air temp etc   Personally Like the control of the link , the fact that all the tables are adjustable means it can be setup as you need and with the sensors you want to run  very rarely do you need to use 4d tables  but they are there if needed   use the default correction tables as a start point  make sure you have an air temp sensor fitted   this makes a huge difference to mixtures at different air flows as well as its position on the intake track, also make sure it is a fast response type sensor and not a closed in thimble one.  

Regards

Dave Heerdegen

[email protected]

 

 

[Eric Godden]

Have you had a chance to review my map dave? Read below but I still would like any opinions you have on it. I did have a bit of luck. I reloaded a map and had to recal the e-throttle when it did that it seems to have calibrated it differently than previously. Now I require more throttle plat request to achieve the same idle speed (open loop e-throttle control) but this has almost removed all discrepencies between the low and high gears as far as afrs. I have a few spots still but I think these are just untuned cells. I still have some problems with my idle, essentially when I go to neutal or clutch-in and the car returns to idle it will idle at 12.9 ~13.9 afrs, usually starting at 12.9 and then slowly increasing to 13.9. If I turn the car off and restart it the afrs at idle will be where I have set them ~15.0afr. If I lean out the cells to remove the rich afrs then I idle very lean at other times. Another issue I am having is that if for example I drive for 20min, park, then restart the car about sitting for maybe 10min or so the car will idle very very lean 16.0 afrs +-.5, this is with the same settings as I have when the idle is rich. If I go drive the car for just a minute or two this lean condition will go away. When this is occuring the car is hot and there is usually no post start or warmup enrichment., I will start another post with these problems for neatness. Thasnks for all the help. -Eric

[Guest Dave]

HI Eric

I have had a quick look at your map and first things are the type of air temp sensor used is to slow  also air temp correction table needs adjusting.

 The reason you are getting different afrs in different gears is due to the cam control y axis is set to tps

what happens in this case is the car is driving down the road at a fixed throttle and a fixed map reading cam control is at a set degree and you tune the fuel cell to suit   then you change gear and have same throttle setting but map reading is slighly different  so different mixtures  then you back of the throttle and cam control moves and you are in a different map cell for fuel and mixtures are different again (read chasing your tail)

In short  tune cam control based of map and then you will find it more stable as the cells will then be the same no matter what the load    it just takes more time to do.

Regards

Dave

[email protected]

 

[Guest Dave]

The base map is a base map only    designed to get the car running which it does well

but there are different e- throttle cals and also tumbler control cals depending on model 

just change the axis over to map and span appropriatley  if you dont adjust the cam control axis then you will never get the mixtures the same in different load situations

the v8 sti sensor can be 'decapped' but that is on you  if it fails i would use either a bosch or delphi open element one

you wont ruin your cams   the only way this would happen is if you had after market cams that had huge lift and the pistons did not have sufficent pockets to suit and then it would damage a valve and a piston

 

Regards

Dave

[email protected]

 

[Guest Dave]

how much of the thermister is protruding ???

yes Bosch ntc    can confirm this when the engine is cold should be close to engine temp reading

re e- throttle cal not a biggy as long as you calibrate the foot position sensor and the throttle body as well

Regards

Dave

[email protected]

 

[Guest Dave]

I also forgot to mention another variable ........fuel pump speed control     depending on what pump and what size injectors you have fitted you may need to change the fuel pump cross over points as it may be switching at the wrong rpm and load for your set up which can change fuel pressure in extreme cases

the cross over points are as per subaru factory    but this may need adjusting if your fuel pressure is fluctuating

 

Regards

Dave

[Eric Godden]

I took off roughly 2/3 of the plastic so there is still a bit on there. The IAT sensor never reads exactly what the ect sensor reads if memory serves me but I will confirm that. As far as fuel pump speed I am usning a walbro 255 which I am pretty sure should use the same as factory. I used it with the original ecu without changing anything.

[Chris Conway]

I have used the 4D table for a very similar reason and have set it up as follows: Used Vitual AUX 1 as the activation by setting ECT>85C and MAP < 80kPa. The reason for this is that I did not want the 4D table to be active while the engine was warming up as this would cause the AFR to be too rich. Also did not want the 4D table to be modifying the main map when on boost. The table was then set up using the Gear number as the Y-AXIS and RPM as the X-AXIS. For gears 0,4,5,6 I set all the cells to '0' so as to have no influence. I then set gears 1, 2 ,3 with various % increases to get the AFR I wanted (around 14.2). The whole reason I set this up was to be able to run lean (16.2 AFR)at cruise loads (< 100kPa)for gears 4,5,6. When I used lower gears then the car would be a bit rough so I need to run at richer AFR (14.2) to get stability. I works fine and I get great fuel economy (around 8 L/100km).

[Eric Godden]

I think this is a great idea. I will have to try this.