Long term fuel trim
Thread Starter
Joined: Sep 2006
Posts: 2,331
Likes: 39
From: Hermosa Beach, CA
Long term fuel trim
Is there any way to zero out the long term fuel trims?
I've read up and some believe the LTFT only affects closed loop, but I think it may change the open loop also. My reasoning for this is because of my LTFT values and A/F ratios.
low/mid/high values were 100/106/126 stock (from evoscan). When my car was dyno'd with the wideband when it was stock, the A/F was fairly constant in it's behavior (steadily around 11.0-11.5) until 5.5k rpms where it just dive bombed in a 200rpm range from 11.0 to somewhere under 10.0. This would seem to correspond to the big jump in the LTFT values going from mid to high.
I would like to have zero fuel trim so that I can better adjust the fuel and timing maps. Also, since I have been tuning my car and gotten rid of the knock, my LFTF high value has gone down to 118 or so last time I checked.
Some more interesting info; I just datalogged a friends car using mitsulogger in Texas. His car is also an 05, but he has a cat-back exhaust. Being in Texas, he gets 93 octane as opposed to the 91 I have to use (and gas there is 70 cents cheaper a gallon to boot!). Anyways, his LTFT values were 3/0/0 in mitsulogger (my values were from evoscan). Regardless, his fuel has been trimmed very very little. I'm guessing the cat-back exhaust caused the car to lean out a hair in the low range which is why some fuel was added there.
However, he's getting knock sums similar to what my car got on the stock ROM. Similiar locations in the RPM range and magnitude. I guess the ECU is content with those knock sums so it has not added more fuel trim.
I've read up and some believe the LTFT only affects closed loop, but I think it may change the open loop also. My reasoning for this is because of my LTFT values and A/F ratios.
low/mid/high values were 100/106/126 stock (from evoscan). When my car was dyno'd with the wideband when it was stock, the A/F was fairly constant in it's behavior (steadily around 11.0-11.5) until 5.5k rpms where it just dive bombed in a 200rpm range from 11.0 to somewhere under 10.0. This would seem to correspond to the big jump in the LTFT values going from mid to high.
I would like to have zero fuel trim so that I can better adjust the fuel and timing maps. Also, since I have been tuning my car and gotten rid of the knock, my LFTF high value has gone down to 118 or so last time I checked.
Some more interesting info; I just datalogged a friends car using mitsulogger in Texas. His car is also an 05, but he has a cat-back exhaust. Being in Texas, he gets 93 octane as opposed to the 91 I have to use (and gas there is 70 cents cheaper a gallon to boot!). Anyways, his LTFT values were 3/0/0 in mitsulogger (my values were from evoscan). Regardless, his fuel has been trimmed very very little. I'm guessing the cat-back exhaust caused the car to lean out a hair in the low range which is why some fuel was added there.
However, he's getting knock sums similar to what my car got on the stock ROM. Similiar locations in the RPM range and magnitude. I guess the ECU is content with those knock sums so it has not added more fuel trim.
Fuel trim has nothing to do with knock. Fuel trim is the correction in fueling needed to maintain a stoich burn during closed loop.
The ECU uses the front O2 sensor feedback to determine if more or less fuel is needed after it bases it's calculations of fueling from MAF, baro, etc. The trims are not used during open loop fueling, so I don't think that is your issue.
Knocking can alter your fueling (and timing) for separate reasons, though. For one, there are routines which pull timing and possibly add fuel due to the presence of knock, and of course once the octane number starts to decrement, there is an interpolation to the low octane map.
Also, different timing adjust can cause AFR changes during open loop. Not to mention the lean spool enable/disables, etc. Basically, there are a lot of factors that can be contributing to what you see. But, I don't think it's the fuel trims as you suspect.
Eric
The ECU uses the front O2 sensor feedback to determine if more or less fuel is needed after it bases it's calculations of fueling from MAF, baro, etc. The trims are not used during open loop fueling, so I don't think that is your issue.
Knocking can alter your fueling (and timing) for separate reasons, though. For one, there are routines which pull timing and possibly add fuel due to the presence of knock, and of course once the octane number starts to decrement, there is an interpolation to the low octane map.
Also, different timing adjust can cause AFR changes during open loop. Not to mention the lean spool enable/disables, etc. Basically, there are a lot of factors that can be contributing to what you see. But, I don't think it's the fuel trims as you suspect.
Eric
Thread Starter
Joined: Sep 2006
Posts: 2,331
Likes: 39
From: Hermosa Beach, CA
If it is fact, then it obviously is a non-issue to tuning the fuel maps.We know the ECU interpolates between the high and low octane and ignition maps once it see enough knock and the 'octane' value goes down. But are we sure the LTFT's don't change the amount of fueling in open loop?
If it is the case that the LTFTs are only for closed loop, then what are are low/mid/high refering too? I didnt' find an absolute definition for all three when searching. There were some guesses but nothing absolute.
I forgot the exact variables that go into it, but low, mid, and high are used at different airflow/load/RPM levels, while still in closed loop.
For example, low is used at conditions at idle, mid is used at cruising conditions, where the airflow, load, RPM will be a bit more, and high I would imagine is a very high airflow, RPM, load cruising situation which will most likely not be reached on a normal basis.
I knew all of the criteria years ago from DSMLink and the DSM ECU, but I have forgotten at this time. If you want to know 100% for sure on the Evo ECU, you can post the question on AktiveMatrix where Bez has just about completely disassembled the ECU.
Eric
For example, low is used at conditions at idle, mid is used at cruising conditions, where the airflow, load, RPM will be a bit more, and high I would imagine is a very high airflow, RPM, load cruising situation which will most likely not be reached on a normal basis.
I knew all of the criteria years ago from DSMLink and the DSM ECU, but I have forgotten at this time. If you want to know 100% for sure on the Evo ECU, you can post the question on AktiveMatrix where Bez has just about completely disassembled the ECU.
Eric
Is there any way to zero out the long term fuel trims?
I've read up and some believe the LTFT only affects closed loop, but I think it may change the open loop also. My reasoning for this is because of my LTFT values and A/F ratios.
low/mid/high values were 100/106/126 stock (from evoscan). When my car was dyno'd with the wideband when it was stock, the A/F was fairly constant in it's behavior (steadily around 11.0-11.5) until 5.5k rpms where it just dive bombed in a 200rpm range from 11.0 to somewhere under 10.0. This would seem to correspond to the big jump in the LTFT values going from mid to high.
I would like to have zero fuel trim so that I can better adjust the fuel and timing maps. Also, since I have been tuning my car and gotten rid of the knock, my LFTF high value has gone down to 118 or so last time I checked.
Some more interesting info; I just datalogged a friends car using mitsulogger in Texas. His car is also an 05, but he has a cat-back exhaust. Being in Texas, he gets 93 octane as opposed to the 91 I have to use (and gas there is 70 cents cheaper a gallon to boot!). Anyways, his LTFT values were 3/0/0 in mitsulogger (my values were from evoscan). Regardless, his fuel has been trimmed very very little. I'm guessing the cat-back exhaust caused the car to lean out a hair in the low range which is why some fuel was added there.
However, he's getting knock sums similar to what my car got on the stock ROM. Similiar locations in the RPM range and magnitude. I guess the ECU is content with those knock sums so it has not added more fuel trim.
I've read up and some believe the LTFT only affects closed loop, but I think it may change the open loop also. My reasoning for this is because of my LTFT values and A/F ratios.
low/mid/high values were 100/106/126 stock (from evoscan). When my car was dyno'd with the wideband when it was stock, the A/F was fairly constant in it's behavior (steadily around 11.0-11.5) until 5.5k rpms where it just dive bombed in a 200rpm range from 11.0 to somewhere under 10.0. This would seem to correspond to the big jump in the LTFT values going from mid to high.
I would like to have zero fuel trim so that I can better adjust the fuel and timing maps. Also, since I have been tuning my car and gotten rid of the knock, my LFTF high value has gone down to 118 or so last time I checked.
Some more interesting info; I just datalogged a friends car using mitsulogger in Texas. His car is also an 05, but he has a cat-back exhaust. Being in Texas, he gets 93 octane as opposed to the 91 I have to use (and gas there is 70 cents cheaper a gallon to boot!). Anyways, his LTFT values were 3/0/0 in mitsulogger (my values were from evoscan). Regardless, his fuel has been trimmed very very little. I'm guessing the cat-back exhaust caused the car to lean out a hair in the low range which is why some fuel was added there.
However, he's getting knock sums similar to what my car got on the stock ROM. Similiar locations in the RPM range and magnitude. I guess the ECU is content with those knock sums so it has not added more fuel trim.
When lean spool is enabled (which it is by defult, from 2500-7000RPM), AFR between the two set RPM points will be lean(er) and when it gets to the stop RPM (7000), AFR go rich, IPW goes up.
Check out this thread
https://www.evolutionm.net/forums/sh...d.php?t=220650
There is also a current thread
https://www.evolutionm.net/forums/sh...d.php?t=259571
edit: It though this occurred at 6.5Krpm for you, but after re-reading and seeing that it happens at 5.5K rpm, lean spool may have nothing to do with your situation.
Thread
Thread Starter
Forum
Replies
Last Post
Raceghost
ECU Flash
22
Jun 1, 2016 12:30 AM
johnnnyjeans
Cobb AccessPORT
3
Jun 4, 2012 09:30 AM
