|
|
|
|
Thread Tools | Display Modes |
08-31-2005, 12:11 PM | #1 |
Scooby Newbie
Member#: 28823
Join Date: Nov 2002
Chapter/Region:
NWIC
Location: Renton, WA
Vehicle:2004 WRX Wagon Black |
Lean burn on oem ECU - use PIC to offset O2 sensor?
I want to stick with the OEM ECU, but I like the idea of experimenting with lean burn cruising.
Assuming that it is correct that the OEM ECU will not allow targetting closed loop fueling to something other than stoich (not sure if this is correct.. in the very least, it should be possible to compile a new ECU rom with extensive assembly changes.. but it'll probably never happen), I'm wondering if there are any other ways to trick the car into running leaner at cruise. One approach that may work is to have a PIC intercept the main O2 sensor signal, and apply a simple translation to the voltage when the TPS and MAF signals are within a particular range. Actually a PIC is probably overkill if it is just a simple constant change like +.5vdc or something.. in this case maybe just a simple comparator to kick in the circuit when TPS and MAF are low (adjustable set point), and then whatever simple circuit to adjust the voltage this set amount. If you want an actual mapping, then a PIC probably is required and is simplest. One thing I'm worried about though is how the fuel trims would work if you did this. Anyone know if fuel trims are stored over a variety of RPM/Load ranges, so that doing this wouldn't muck with your non-cruise fuel trims? Anyone have a graph of O2 sensor voltage output versus lambda? Thanks! Adrian
* Registered users of the site do not see these ads.
|
08-31-2005, 01:27 PM | #2 |
*** Banned ***
Member#: 2992
Join Date: Nov 2000
Chapter/Region:
NESIC
Location: Connecticut, USA
Vehicle:02 WRX Sedan Silver |
The front A/F sensor is a Denso "limit current" type sensor. The sensor's controller within the ECU applies varying voltage to both sides of the sensor. The sensor's "signal" is the amount and direction of current flowing. See http://forums.nasioc.com/forums/atta...chmentid=26781 for a graph that depicts the current flow / air fuel ratio relationship.
|
08-31-2005, 01:41 PM | #3 |
Scooby Newbie
Member#: 28823
Join Date: Nov 2002
Chapter/Region:
NWIC
Location: Renton, WA
Vehicle:2004 WRX Wagon Black |
Blech, that sure makes it less straightforward.. especially with the supply voltage varying.
From that graph, to lean out beyond stoich would require a drop in current flow, so at least you could get the change you want just be grounding a small amount of the current. Sounds hard. I wonder how hard it would be to hack the assembly.. probably the CL fueling code is not implemented in a way that would make it simple to have different fueling targets in different circumstances. Sounds like a UTEC is far simpler -Adrian |
08-31-2005, 03:57 PM | #4 |
Scooby Specialist
Member#: 59287
Join Date: Apr 2004
Chapter/Region:
TXIC
Location: Houston, Tx
Vehicle:2007 WRB WRX 2004 WRB WRX (RIP) |
It would be a lot easier to do this from within the ECU. Openecu.org is making some progress in disassembling the code.
|
09-03-2005, 02:44 PM | #5 |
Scooby Specialist
Member#: 49087
Join Date: Nov 2003
Chapter/Region:
RMIC
Location: Inside your carnot cycle
Vehicle:1998 RST V8 STi swap '05 R6 '95 BMW V8 hotness |
The reason for that really strange output curve is a function of the ceramic material used in the sensor itself. Different sensor styles have different curves. This is why the really good 4/5 wire sensors all have a calibration table buried in the ECU somewhere. You really aren't TOO concerned with the values in the sensor range where the values start to get really goofy though, so a small shift in the input shouldn't cause things to freak out too much.
|
09-07-2005, 03:10 PM | #6 |
Scooby Newbie
Member#: 15918
Join Date: Mar 2002
Chapter/Region:
NWIC
Location: Seattle, WA
Vehicle:08 STI/02 PGT WRX DGM / Orange |
...definitely a lot cleaner to make this change in the ECU code and reflash. you could just modify the o2 sensor lookup table to alter where it thinks stoich is.
|
09-07-2005, 04:22 PM | #7 | |
*** Banned ***
Member#: 2992
Join Date: Nov 2000
Chapter/Region:
NESIC
Location: Connecticut, USA
Vehicle:02 WRX Sedan Silver |
Quote:
|
|
09-07-2005, 04:41 PM | #8 | |
Scooby Newbie
Member#: 15918
Join Date: Mar 2002
Chapter/Region:
NWIC
Location: Seattle, WA
Vehicle:08 STI/02 PGT WRX DGM / Orange |
Quote:
|
|
09-07-2005, 06:53 PM | #9 |
Former Vendor
Member#: 60836
Join Date: Apr 2004
Chapter/Region:
SCIC
Location: Torrance, CA
Vehicle:2006 Mod Class STI TIMEATTACK!! |
The latest version of Ecutek flashing software allows you to change the target AFR. We, or any Ecutek dealer, could change it.
Regards, Dan |
09-07-2005, 07:02 PM | #10 | |
*** Banned ***
Member#: 2992
Join Date: Nov 2000
Chapter/Region:
NESIC
Location: Connecticut, USA
Vehicle:02 WRX Sedan Silver |
Quote:
|
|
09-07-2005, 10:16 PM | #11 | |
*** Banned ***
Member#: 2992
Join Date: Nov 2000
Chapter/Region:
NESIC
Location: Connecticut, USA
Vehicle:02 WRX Sedan Silver |
Quote:
Or, should we conclude that Harman is absolutely clueless? |
|
09-07-2005, 11:16 PM | #12 | |
Former Vendor
Member#: 60836
Join Date: Apr 2004
Chapter/Region:
SCIC
Location: Torrance, CA
Vehicle:2006 Mod Class STI TIMEATTACK!! |
Quote:
CLOSED LOOP FUELING TARGETS EcuTek just added the parameter. Geez. Dan Last edited by AQ Motorsports; 09-07-2005 at 11:27 PM. |
|
09-07-2005, 11:29 PM | #13 |
Scooby Specialist
Member#: 56321
Join Date: Mar 2004
Chapter/Region:
NESIC
Location: Hopkinton, MA
Vehicle:2006 Honda El Camino Green |
Dan-
When you run the car leaner than stoich, do they provide some mechanism to kick out back to stoich (or richer) immediately upon throttle input? I guess, does it simply change the CL target so that whenever the car is in CL, it runs another value and then the transition to OL is the same as always? I'd think things could get dicey under partial throttle if not tuned correctly. Thanks! Mike |
09-08-2005, 01:37 AM | #14 |
Scooby Newbie
Member#: 90590
Join Date: Jul 2005
Chapter/Region:
NWIC
Location: Seattle
Vehicle:2005 STi OBP |
Oh no, ahains has the PIC addiction!
On the subject of lean burn cruise, couldn't something similar be achieved through attenuating the MAF signal in the low MAP + RPM range? I'm not sure what type of MAF output Subys use, but if it's the same Harman Vortex as my LT1, existing things like the MAF-T can be used, or if the output is 0-5v, it'd be simple to build a PIC based solution. Edit: Doh, just realized we're talking about CL and the ECU will compensate for lowered o2 readings with more inj dc. Disregard above. Back to square one. -Jay Last edited by JayZ28; 09-08-2005 at 01:43 AM. |
09-08-2005, 01:47 AM | #15 | |
Former Vendor
Member#: 60836
Join Date: Apr 2004
Chapter/Region:
SCIC
Location: Torrance, CA
Vehicle:2006 Mod Class STI TIMEATTACK!! |
Quote:
It's a single variable adjustment, so I believe anytime the car was in CL it would target that value. To avoid the chopiness, you could change the parameters for CL operation (kicking the car into OL quickly) as long as you had an excellent calibration on the injectors and fuel basemap. Realistically, I think its more trouble than it's worth - but it could be done. I ran 16:1 at cruise w/ my standalone for a while, but didn't notice a large improvement in mileage - so I canned it. I'd assume if there were cats in the car that running that lean at cruise could cause some issues due to heat, but that's just me talking. This new EcuTek function could prove useful on big cam'd cars though - it's nearly impossible to get a good idle above 13:1 with a lumpy set. Couple that with some crazy OL adjustments and it may just make the ECU that much more flexible. The Subaru MAF is a 0-5v output, not a Karman Vortex. You guys prob. have the right idea - building a seperate controller/interceptor. Interesting idea none the less. Regards, Dan |
|
09-08-2005, 01:59 AM | #16 |
Scooby Newbie
Member#: 90590
Join Date: Jul 2005
Chapter/Region:
NWIC
Location: Seattle
Vehicle:2005 STi OBP |
Dan-
Just out of curiosity, are there any sensor inputs besides o2 which either directly affect target AF ratio or indirectly affect final injector DC while in CL? Things like coolant temp, etc? There may be easier sensor signals to intercept than the O2 signal... just thinking out loud here. Back on my big cammed LT1, I had to resort to open loop idle to have any sort of driveability on the street.. maybe forcing the Suby ECU to operate in open loop during specific MAP+RPM ranges could be an option? -Jay |
09-08-2005, 07:32 AM | #17 |
Scooby Specialist
Member#: 56321
Join Date: Mar 2004
Chapter/Region:
NESIC
Location: Hopkinton, MA
Vehicle:2006 Honda El Camino Green |
The way the car targets AFR values is that it uses the MAF scaling as a starting point and then will adjust fuel trims based on the input from the front O2 sensor. If the starting point is rich, the trims will remove fuel to get the car back to 14.7:1. Basically, it fuels assuming X grams of air per second and then uses the from O2 sensor to fine tune and learn what it is really seeing.
Thanks for the info, Dan! Mike |
09-08-2005, 02:36 PM | #18 |
Scooby Guru
Member#: 8785
Join Date: Jul 2001
Chapter/Region:
SWIC
Vehicle:02 c_turner@ix. netcom.com |
Harman is correct. You can now change the Target AFR for Closed loop. This includes raising it up for a leaner Burn. I have explerimented with this setting around 15.2 to 1 AFR for cruise. I see no real drawbacks at this time and did not see a change in EGT that worried me.
Clark |
09-08-2005, 03:55 PM | #19 | |
Scooby Guru
Member#: 23717
Join Date: Aug 2002
Chapter/Region:
NESIC
Location: Beverly, MA
Vehicle:2005 Big red |
Quote:
TMS |
|
09-08-2005, 04:34 PM | #20 | |
*** Banned ***
Member#: 2992
Join Date: Nov 2000
Chapter/Region:
NESIC
Location: Connecticut, USA
Vehicle:02 WRX Sedan Silver |
Quote:
|
|
09-09-2005, 09:20 AM | #21 |
Scooby Newbie
Member#: 19050
Join Date: May 2002
Chapter/Region:
MWSOC
Location: Hannibal, MO
Vehicle:2005 Black STi Happily rowing 6 gears |
Jon [in CT]... Dunno if it's relavent, but, I monitor A/F with a set of PSI3 guages and a Zietronix WB. The WBO2 sensor is about 3' further downstream than the factory front O2 sensor (if that makes any difference). I usually see 17.5:1 on the wideband, glance down at the front O2 and see 18.2:1. The closer you get to stoich the less the difference is, it doesn't start skewing until around 16.7:1 or so and up (numerically). This would be using OLF in the UTEC under cruise conditions. -Chuck
|
09-12-2005, 02:10 PM | #22 |
Scooby Guru
Member#: 8785
Join Date: Jul 2001
Chapter/Region:
SWIC
Vehicle:02 c_turner@ix. netcom.com |
John. I used an LM1 WB meter to confirm the Closed loop AFR did change. This car held the same 10.6 to 1 AFR in open loop. I do not see this change effecting open loop.
Going to 17.1 to will lower EGT as you are now on LOP. If you are in fact on LOP you are running less effeciently then you where before. Its common to think that if you inject less fuel your mileage will increase but in cases like this you are probably worse off. |
09-12-2005, 06:21 PM | #23 | |
*** Banned ***
Member#: 2992
Join Date: Nov 2000
Chapter/Region:
NESIC
Location: Connecticut, USA
Vehicle:02 WRX Sedan Silver |
Quote:
Operating leaner than stoichometric will improve fuel economy up to a point. Most texts suggest that "point" is in the vicinity of 1.10 to 1.20 lambda. If you lean beyond that "point" you're wasting more and more energy produced by combustion to heat up the excess air that just gets flushed out of the exhaust and your specific fuel consumption begins to rise. It occurs to me that maybe you are remembering that the AFR which produces best torque is also the one which burns fastest. As you dilute the AFR with extra air the burn duration increases. This is important to remember when using an AFR leaner than stoichiometric, given that Subaru engineers designed the ignition advance tables assuming that a stoichiometric AFR would be used while cruising. |
|
09-13-2005, 01:18 PM | #24 | |
Scooby Guru
Member#: 8785
Join Date: Jul 2001
Chapter/Region:
SWIC
Vehicle:02 c_turner@ix. netcom.com |
Quote:
I was referencing Lean side of peak. The purpose of providing the link I did will show Lean side and rich side and Stoich EGT curves in graph form. You are correct in that ignitiontables will need to be adjusted. What needs to happen for me to really Test diff lean AFR's is 2-3 hours of time so I can log and test various AFR and Timing settings while watching IDC, RPM and Airflow to compute actual fuel used. I just dont have that time... But I will start the process today on my own car. Its AG970 ECu is currently set to 14.7 to 1 Closed loop. I will try 15.0 and run some logs. Clark |
|
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
How to extend o2 sensor without wasting 30$ on extension? | Un0RiGiNaL | Normally Aspirated Powertrain | 21 | 05-16-2009 04:45 PM |
Can I use a generic 3wire O2 Sensor? | d1camero | Normally Aspirated Powertrain | 23 | 06-13-2008 06:50 PM |
WTB: Used or new Front O2 sensor | AudiobahnSTI | Private 'Wanted' Classifieds | 5 | 03-14-2006 11:31 PM |
What can be used to plug O2 sensor bungs? | porque | Factory 2.0L Turbo Powertrain (EJ Series Factory 2.0L Turbo) | 19 | 09-17-2005 05:30 PM |
FS: Used '02 - '03 WRX O2 Sensor | gsodonis | Private 'For Sale' Classifieds | 14 | 10-13-2004 07:55 PM |