Trying to Better Understand how OEM ECU Learns
12-03-2010, 03:56 PM
#1
Thread Starter
Trying to Better Understand how OEM ECU Learns
OK, I know the stock ECU learns and adjusts to the proper air/fuel ratio. So I assume the ECU has some sort of standard settings or range for the air/fuel ratio at different RPMs and throttle inputs.
I also know that bolt on breather mods, such as intake, headers, high flow cat, test pipe and exhaust create a slightly leaner air/fuel ratio, since more air is moving in and exiting easier. As a result of the slightly leaner air/fuel ratio, you produce a little more power.
For example, I have a my00 with a richer ECU. Say my AUT cooling plate, K&N drop in filter and Greddy Spectrum Elite dual make the wide open throttle, high RPM air/fuel ratio leaner, maybe from 12:1 to 12.5:1 (not sure if these numbers are correct; just guessing). I know the mods produced a little amount of power, because I can slightly feel it and I've seen dynos.
Makes sense except for one part. If the ECU learns and adjusts the air/fuel ratio, wouldn't it just adjust it back to the stock range by sending in more fuel to compensate for the extra airflow the mods produced? I guess I'm just trying to figure out how S2Ks run leaner after breather mods. It seems like the ECU would learn and bring the air/fuel back to stock, negating the new airflow (even if you did reset it after install). Can the ECU just not learn enough or what?
Thanks in advance!
I also know that bolt on breather mods, such as intake, headers, high flow cat, test pipe and exhaust create a slightly leaner air/fuel ratio, since more air is moving in and exiting easier. As a result of the slightly leaner air/fuel ratio, you produce a little more power.
For example, I have a my00 with a richer ECU. Say my AUT cooling plate, K&N drop in filter and Greddy Spectrum Elite dual make the wide open throttle, high RPM air/fuel ratio leaner, maybe from 12:1 to 12.5:1 (not sure if these numbers are correct; just guessing). I know the mods produced a little amount of power, because I can slightly feel it and I've seen dynos.
Makes sense except for one part. If the ECU learns and adjusts the air/fuel ratio, wouldn't it just adjust it back to the stock range by sending in more fuel to compensate for the extra airflow the mods produced? I guess I'm just trying to figure out how S2Ks run leaner after breather mods. It seems like the ECU would learn and bring the air/fuel back to stock, negating the new airflow (even if you did reset it after install). Can the ECU just not learn enough or what?
Thanks in advance!
12-03-2010, 05:25 PM
#2
Thread Starter
Ha, when someone like you chimes in, it shows how relatively little I know about this topic. Thanks for the detailed response. I had to read it a couple of times to make sure I got it all!
So to clarify, under closed loop operation (not WOT), the ECU will adjust for breather mods and maintain the stock air/fuel ratio. But under open loop (WOT), the ECU cannot adjust due to fixed maps. This means more air is going in with breather mods, but not extra fuel at WOT, hence the slightly leaner air/fuel ratio and power increase.
And I'm definitely not into trying to trick the computer... or even tuning at this point.
So to clarify, under closed loop operation (not WOT), the ECU will adjust for breather mods and maintain the stock air/fuel ratio. But under open loop (WOT), the ECU cannot adjust due to fixed maps. This means more air is going in with breather mods, but not extra fuel at WOT, hence the slightly leaner air/fuel ratio and power increase.
And I'm definitely not into trying to trick the computer... or even tuning at this point.
12-03-2010, 07:28 PM
#3
Registered User
"Tricking" the computer is a half *** attempt at tuning. Get a real EMS if you plan on tuning.
I'd recommend reading a EFI/tuning book like "How to tune and modify engine management" by Jeff Hartman. The book goes over the basics of OEM computers and aftermarket EMS's and is also easy to read. They can be found on ebay or amazon.
I'd recommend reading a EFI/tuning book like "How to tune and modify engine management" by Jeff Hartman. The book goes over the basics of OEM computers and aftermarket EMS's and is also easy to read. They can be found on ebay or amazon.
12-04-2010, 06:45 AM
#4
Thread Starter
Originally Posted by H22toF20,Dec 3 2010, 08:28 PM
"Tricking" the computer is a half *** attempt at tuning. Get a real EMS if you plan on tuning.
I'd recommend reading a EFI/tuning book like "How to tune and modify engine management" by Jeff Hartman. The book goes over the basics of OEM computers and aftermarket EMS's and is also easy to read. They can be found on ebay or amazon.
I'd recommend reading a EFI/tuning book like "How to tune and modify engine management" by Jeff Hartman. The book goes over the basics of OEM computers and aftermarket EMS's and is also easy to read. They can be found on ebay or amazon.
12-04-2010, 08:25 AM
#5
Originally Posted by 2ndTimeAround,Dec 4 2010, 03:38 AM
Learn isn't really the right term, often used to describe idle adjustment, idle (or value in question) correction (fuel, spark & TP trimming) would be a more accurate description. In closed (feedback) loop the ECU (computer) polls input from sensors and plugs that data in to it's calculations. Depending on the data present it can result in a number of adjustments. This mode functions after initial start up / warm up is complete under idle and part throttle conditions.
It is not that the ECU can not make enough corrections, it is fully capable in close loop operation. The lean conditions that pose a problem are in open-loop (non-feedback). Usually during cold start/warm up and WOT, in this mode the ECU relies on fixed maps. There are several reason for this, often due to hardware or data limitations (not enough samples for correction, rapid change in conditions, engine speed, etc, etc). Fuel and spark maps are fixed and the ECU cannot adjust them. That being why there is not correction under WOT.
However there are ways to ad fuel or even deceive the computer. I have not tested any on a Honda. On other cars I've seen larger injectors (10-20%) installed, closed loop can trim them, open loop will have the fuel needed. Over 10% usually doesn't idle well on speed density cars and WOT is rich as pig, this can be counter productive. This method along with a cooler set of plugs is sometimes used on quick spool OE engines, that run out of air in high rpms going to a larger turbos that feed those rpms. Another methods using an adjustable FPR to raise pressure therein adding enough fuel for WOT, yet small enough that part throttle trims are minimal. Large trims like those induced by larger injectors may throw CELs on some ECUs. Outside of an adjustable FPR with fine tuning (rate and pressure), I wouldn't want to trick the computer to much. Keep in mind bad data could ultimately lead to bad results. Ultimately why many are opposed to MAP or other signal adjusters.
It is not that the ECU can not make enough corrections, it is fully capable in close loop operation. The lean conditions that pose a problem are in open-loop (non-feedback). Usually during cold start/warm up and WOT, in this mode the ECU relies on fixed maps. There are several reason for this, often due to hardware or data limitations (not enough samples for correction, rapid change in conditions, engine speed, etc, etc). Fuel and spark maps are fixed and the ECU cannot adjust them. That being why there is not correction under WOT.
However there are ways to ad fuel or even deceive the computer. I have not tested any on a Honda. On other cars I've seen larger injectors (10-20%) installed, closed loop can trim them, open loop will have the fuel needed. Over 10% usually doesn't idle well on speed density cars and WOT is rich as pig, this can be counter productive. This method along with a cooler set of plugs is sometimes used on quick spool OE engines, that run out of air in high rpms going to a larger turbos that feed those rpms. Another methods using an adjustable FPR to raise pressure therein adding enough fuel for WOT, yet small enough that part throttle trims are minimal. Large trims like those induced by larger injectors may throw CELs on some ECUs. Outside of an adjustable FPR with fine tuning (rate and pressure), I wouldn't want to trick the computer to much. Keep in mind bad data could ultimately lead to bad results. Ultimately why many are opposed to MAP or other signal adjusters.
The Ecu reads the AFR jumps from the stock Lambda Sensor. It's not an Wideband Lambda, so the ECU recognizes in closed loop operation only "to rich" or "to lean" and makes statistical calculations to write a value into the correction table called "Short Term Fuel Trim" STFT. The correction can be +/- 25% to the Fuel tables. this works only in closed loop. But the reason for working only in closed loop is not the limited capabilities of the sensors, it's just because the engine needs rich AFR under high loads.
There's another statistic calculation that's read these values from the STFT and writes a correction parameter into the "Long Term Fuel Trim" LTFT Table. This can be up to +/- 15% (I guess). this correction is working in closed and open loop!
When you reset the ECU by powering it off for a few seconds, it erases the stored values in these tables and the ECU has to learn some time for the new corrections.
So any modifications that have influence on the AFR of the engine will be corrected after some time by the LTFT Value - even under WOT!
12-04-2010, 08:42 AM
#6
[QUOTE=bpaspi,Dec 4 2010, 06:25 PM] Sorry, but there are some false assumptions..
The Ecu reads the AFR jumps from the stock Lambda Sensor. It's not an Wideband Lambda, so the ECU recognizes in closed loop operation only "to rich" or "to lean" and makes statistical calculations to write a value into the correction table called "Short Term Fuel Trim" STFT. The correction can be +/- 25% to the Fuel tables. this works only in closed loop. But the reason for working only in closed loop is not the limited capabilities of the sensors, it's just because the engine needs rich AFR under high loads.
There's another statistic calculation that's read these values from the STFT and writes a correction parameter into the "Long Term Fuel Trim" LTFT Table. This can be up to +/- 15%
The Ecu reads the AFR jumps from the stock Lambda Sensor. It's not an Wideband Lambda, so the ECU recognizes in closed loop operation only "to rich" or "to lean" and makes statistical calculations to write a value into the correction table called "Short Term Fuel Trim" STFT. The correction can be +/- 25% to the Fuel tables. this works only in closed loop. But the reason for working only in closed loop is not the limited capabilities of the sensors, it's just because the engine needs rich AFR under high loads.
There's another statistic calculation that's read these values from the STFT and writes a correction parameter into the "Long Term Fuel Trim" LTFT Table. This can be up to +/- 15%
Trending Topics
12-04-2010, 09:03 AM
#8
Originally Posted by bpaspi,Dec 4 2010, 06:49 PM
LTFT is also working in WOT.
That's an Info I have from Hondata.
I have written, that STFT is only applied during closed loop.
That's an Info I have from Hondata.
I have written, that STFT is only applied during closed loop.
LTFT is only applied during closed loop, so not used in open loop.
If you monitor your LTFT trim values with an obd2 reader you will also see this when in WOT however some will show its value as its stored, but this doesnt means its actually getting applied.
12-04-2010, 10:07 AM
#9
I stay to my statement - LTFT is also applied on WOT.
OBD doesn't show any changes to the LTFT values, because they are altered from the STFT values. And these (only STFT) are changed only during closed Loop.
But the LTFT corrections are also valid during open loop.
I have this statement not only from Hondata, also Bosch confirmed this.
It makes sense, because the LTFT prevents the engine from damages due to wrong AFR when something is out of the specification. e.g. a dirty airfilter....
OBD doesn't show any changes to the LTFT values, because they are altered from the STFT values. And these (only STFT) are changed only during closed Loop.
But the LTFT corrections are also valid during open loop.
I have this statement not only from Hondata, also Bosch confirmed this.
It makes sense, because the LTFT prevents the engine from damages due to wrong AFR when something is out of the specification. e.g. a dirty airfilter....
12-04-2010, 10:34 AM
#10
Registered User
Join Date: Apr 2008
Posts: 176
Likes: 0
Received 0 Likes
on
0 Posts
I always taught that open loop is commanded fuel, target based on ECT, IAT, MAP/MAF, etc... O2 sensors are out of the loop in open loop...
https://www.s2ki.com/forums/index.ph...ic=803783&st=0
Cheers
https://www.s2ki.com/forums/index.ph...ic=803783&st=0
Cheers