Intakes, Fuel Trims, and AFR
Here is a test I conducted yesterday:
Today Gear Grinder and I conducted multiple tests on his Evo X. One of the was to confirm, yet again, the impact of LTFTs on the AFR using the famous CARB legal intake.
We flashed the car back to the stock 020 rom id when the trims were at +6.4%. They went to 7% by the time we logged. We ran two back to back WOT 4th gear runs.
Then we ran the car at a steady 60 mph on the freeway and headed back to the same on-ramp that we did the first two logs. By the time we got to the on ramp, the LTFT was maxxed out @ 12.5%. We did 2 back-to-back WOT logs
Here are the results, confirming the previous results
The AFR became richer when the trim went up from +7% to +12.5%. At 6500 rpm the AFR richened up from 11.05:1 to 10.6:1, 0.45 of a point richer. The math supports the findings: the trim rose 5.5%. 11.05x0.055=0.6-11.05=10.45. 10.45:1 is very close to the 10.6:1 that we logged when the trims maxxed out at 12.5%.
So now it is confirmed that the famous CARB legal intake can max out the LTFT at 12.5%. It can even trigger a P0171, system too lean code as it has done on some cars running in CA.
Last edited by nj1266; Aug 2, 2009 at 09:16 AM.
https://www.evolutionm.net/forums/ec...ight=funky+afr
Just logged my LTFT cruise, im pegged at 12.5% idle is 4.6%. Looking through ecu flash looks like we do not have access to the MAF scaling map yet is that correct? Is adjusting the latency really the proper way to fix this as the root of the problem is likely the intake? Even if the LTFT is not used at wot shouldnt we still try to get that LTFT cruise back to within the 5% range simply as a fail safe in case we get into a situation where the car may need extra fueling. If mine is maxed at 12.5% the ecu can not richen any further correct?
Just logged my LTFT cruise, im pegged at 12.5% idle is 4.6%. Looking through ecu flash looks like we do not have access to the MAF scaling map yet is that correct? Is adjusting the latency really the proper way to fix this as the root of the problem is likely the intake? Even if the LTFT is not used at wot shouldnt we still try to get that LTFT cruise back to within the 5% range simply as a fail safe in case we get into a situation where the car may need extra fueling. If mine is maxed at 12.5% the ecu can not richen any further correct?
Don't let anyone tell you that LTFT do not carry over into WOT. They are WRONG, dead wrong.
When pegged at 12.5%, then the ECU can no longer add fuel to correct what it sees as a lean condition during cruising. If this continues a P0171 "system too lean" is triggered.
well i have never seen LTFT-High get off 0% (on CT9A)
And that would be the MAF-Hz bin for most WOT operation...
I havn't really logged the trims on the X yet due to stock injectors - i'll go have a look tommorow.
ps - Subaru != Mitsu
ps2 - MAF Scaling is out there, I have it defined on my ROM and I have seen it defined on USDM ROM's
for 53040010:
5268XXXX
And that would be the MAF-Hz bin for most WOT operation...
I havn't really logged the trims on the X yet due to stock injectors - i'll go have a look tommorow.
ps - Subaru != Mitsu
ps2 - MAF Scaling is out there, I have it defined on my ROM and I have seen it defined on USDM ROM's
for 53040010:
Code:
<table name="MAF Scaling Part 1" address="5753e"> <table name="Volts" address="61626"/> </table> <table name="MAF Scaling Part 2" address="57596"> <table name="Volts" address="6167e"/> </table> <table name="MAF Scaling Part 3" address="575ee"> <table name="Volts" address="616d6"/> </table>
Code:
<table name="MAF Scaling Part 1" address="5753a">
<table name="Volts" address="615ea"/>
</table>
<table name="MAF Scaling Part 2" address="57592">
<table name="Volts" address="61642"/>
</table>
<table name="MAF Scaling Part 3" address="575ea">
<table name="Volts" address="6169a"/>
</table>
It does not...It just carries over from closed loop cruise into WOT operation. So even though the LTFT cruise is not adding/subtracting fuel at WOT, it adds/aubtracts the fuel it has calculated in closed loop to WOT. So if LTFT cruise is X, then the AFR becomes, target AFR+X. X stays the same in WOT operation as it was during closed loop cruise. It is simply added to the target AFR.
Last edited by nj1266; Aug 2, 2009 at 09:18 AM.
well i have never seen LTFT-High get off 0% (on CT9A)
And that would be the MAF-Hz bin for most WOT operation...
I havn't really logged the trims on the X yet due to stock injectors - i'll go have a look tommorow.
ps - Subaru != Mitsu
ps2 - MAF Scaling is out there, I have it defined on my ROM and I have seen it defined on USDM ROM's
for 53040010:
5268XXXX
And that would be the MAF-Hz bin for most WOT operation...
I havn't really logged the trims on the X yet due to stock injectors - i'll go have a look tommorow.
ps - Subaru != Mitsu
ps2 - MAF Scaling is out there, I have it defined on my ROM and I have seen it defined on USDM ROM's
for 53040010:
Code:
<table name="MAF Scaling Part 1" address="5753e">
<table name="Volts" address="61626"/>
</table>
<table name="MAF Scaling Part 2" address="57596">
<table name="Volts" address="6167e"/>
</table>
<table name="MAF Scaling Part 3" address="575ee">
<table name="Volts" address="616d6"/>
</table>
Code:
<table name="MAF Scaling Part 1" address="5753a">
<table name="Volts" address="615ea"/>
</table>
<table name="MAF Scaling Part 2" address="57592">
<table name="Volts" address="61642"/>
</table>
<table name="MAF Scaling Part 3" address="575ea">
<table name="Volts" address="6169a"/>
</table>
I have tested the Evo IX, Subrau STI, Subaru WRX, Evo X and all of them CARRY the LTFT-cruise over to WOT operation. The LTFT cruise is NOT active at WOT, it is just added (carried over) to whatever the target AFR is in the fuel table.
I am wrong to ask a question? 
Your testing seems reasonable, but still leaves quite a few things open on the table. I'm not saying it is wrong, I'm just saying it might not be end all be all conclusive.
Do we know if the X has any lean spool equivalent? That could seriously play a part in your testing and be very unpredictable in relation to how you were driving before testing each pull.
I appreciate what you have done so far, but it is only a step in the right direction...
I did take a look at a few older logs and noticed that 200+ load is achievable with <1580hz on the IX in the lower RPMs. This still leaves the question (for me, at least) if FT applies during open loop operation or if that *only* turns off the trim feedback. I still need to do some logging of my own to see if this stuff applies to my car as I think it does.
A good positive solution to this problem would also be to find the fuel trim airflow hz switchover point and lower it significantly. If I remember correctly these values were just before the fuel table in the IX.
Your testing seems reasonable, but still leaves quite a few things open on the table. I'm not saying it is wrong, I'm just saying it might not be end all be all conclusive.
Do we know if the X has any lean spool equivalent? That could seriously play a part in your testing and be very unpredictable in relation to how you were driving before testing each pull.
I appreciate what you have done so far, but it is only a step in the right direction...
I did take a look at a few older logs and noticed that 200+ load is achievable with <1580hz on the IX in the lower RPMs. This still leaves the question (for me, at least) if FT applies during open loop operation or if that *only* turns off the trim feedback. I still need to do some logging of my own to see if this stuff applies to my car as I think it does.
A good positive solution to this problem would also be to find the fuel trim airflow hz switchover point and lower it significantly. If I remember correctly these values were just before the fuel table in the IX.
Last edited by fostytou; Aug 2, 2009 at 10:08 AM.
[quote=fostytou;7348774]
There is no lean spool on the Evo X. It is turned off from the factory.
LTFT cruise is NOT active during WOT. The LTFT does not add/subtract fuel when in WOT. What happens is the LTFT cruise that has already been stored is added to WOT operation.
I do not know why it is done that way. I do not know why the ECU does not simply read what is in the fuel map w/o adding LTFT cruise to the fuel map.
Do we know if the X has any lean spool equivalent? That could seriously play a part in your testing and be very unpredictable in relation to how you were driving before testing each pull.
This still leaves the question (for me, at least) if FT applies during open loop operation or if that *only* turns off the trim feedback.
I do not know why it is done that way. I do not know why the ECU does not simply read what is in the fuel map w/o adding LTFT cruise to the fuel map.