Tuning for E10 - Injector Scalar or something else?

Guest_1u53r · 11-26-2006, 04:54 PM

Is it better to change the injector scalar to a 6.3% lower flow number or to change something else (MAF cal, fuelmap, etc) to help account for E10 fuel?

I am using Enginuity on a mostly-stock 06STI.

It makes sense to me that I still want my lambda targets to be almost the same, so I don't want to change the whole fuel map to try to account for this but instead to just reduce the scalar to increase fuel flow overall, including closed-loop.

Also, I'm having trouble finding where to alter the closed-loop fuel targets in Enginuity. What am I missing?

WolfPlayer · 11-26-2006, 05:17 PM

Nick,

Fuel is in the low-det fuel map via AFRs.

Here's the deal with E10 as far as I can figure out on my own. The stock ECU is calibrated for petrol. That means that the stock O2 sensor is interpreted in a specific way by the ECU. Usually, this is done via 'switchover points' in mV. To configure the stoichiometric lambda for E10 gas, you would have to reconfigure the O2 crossover points (which by default means that you have the knowledge of knowing what mV corresponds to the stoich of E10). The O2 stuff isn't exposed in Enginuity as far as I know (I could be wrong though).

Long story short: Changing the injector constant won't get you what you need. The ECU would basically just apply a long term fuel trim to get the car back to where the ECU wants it. I would consider richening up the fuel map according to what your car needs. This will tell the ECU to shoot for a different fueling (captain obvious - yea, I know). I am suggesting this because E10 is working pretty good at 14.7 on my car in closed loop locations. I did have to richen up WOT areas accordingly. A good place to start is to use Cobb's E85 information (which they have ALSO supplied a column for E10!) contained in their spreadsheet worksheet linked on their website.

t

Guest_1u53r · 11-26-2006, 06:18 PM

OK I have to point out some things here that I am sure you know, but may not have yet considered.

First of all, E10 burning at its stoich point of 13.8 will produce an O2 lambda of 1.0. Petrol burning at its stoich point of 14.7 will also produce an O2 lambda of 1.0. Since we know that the map is really just lambda targets, we can also assume that no matter what fuel is used, closed loop fueling will be adjusted to achieve the target lambda (1.0 for the case of idle). This means if you adjust _nothing_, using E10 you will always be running at least 6% higher AF Learning trims than if you were running proper petrol because the ECU senses that the pre-tabled injector pulsewidth is not adequate to meet the lambda.

I have seen this behavior personally... Car properly calibrated so that AFR trims are low everywhere, then fill up with E10 and watch the trims rise.

Now, you mention that E10 on your car is working fine at 14.7, but I question this claim because A), 14.7 is vastly leaner than E10's stoich point and B), your logging software doesn't realize that you're using E10 and the sensor doesn't report AFR, it's reports lambda voltage. So an 'indicated' 14.7 using E10 is merely indicative that the software is using a petrol multiplier. In Innovate's Logworks, you can specify a different multiplier so that AFR is reported accurately, but neither Deltadash, StreetTuner, nor Enginuity actually have this option so any WBO2 logging done with them will be 'incorrect' in terms of AFR. You CAN display lambda instead of AFR though, but that would only tell you that it is still hitting the proper lambda. If you were to LOG the lambda, then use a multipler column in your logged spreadsheet, you would see that what you previously logged as "14.7", when properly calculated, is actually 13.8

Now if you just adjust the fuel maps to a 'richer' AFR value, you will run richer for sure, but realize that the displayed tables are interpreted by the software from a natively lambda value in the actual ROM. As such, adjusting the AFR in the displayed table is just adjusting the Lambda target and that is NOT what I want to do.

Further, if you were to ignore this part and just adjust the fuel tables to account for the lower stoich point, it would probably run fine until the corrections from the closed loop learning process started to get applied to those higher load points and start throwing your values way off. You need a truly global adjustment for the changing fuel, imho. Adjusting the MAF table (when we are not actually seeing a change in airflow) will affect the calculated load and thus will affect timing and fueling improperly, IMHO.

Adjusting just the open-loop fuel maps, needs to be accompanied by a corresponding adjustment in closed-loop targets (which is why I also asked about those, since it's not apparent where they are in enginuity).

It strikes me that the most direct way to alter the injector pulsewidth is via the flow scalar. We want to spray more fuel (by volume) for a given airflow than before so you skew the ECU's idea of what pulsewidth flows what amount of fuel, thus increased injected volume.

What am I missing in this reasoning?

RainMaker · 11-26-2006, 08:08 PM

Damn. Suddenly I figure out wy my fuel trims suddenly went to 7-8 out of the blue.

*grumble*

LOLSTi · 11-26-2006, 08:12 PM

I hope Shell diafs.

Great thread guys.

Guest_1u53r · 11-26-2006, 08:14 PM

Quote:

Originally Posted by RainMaker

Damn. Suddenly I figure out wy my fuel trims suddenly went to 7-8 out of the blue.

*grumble*

Did you recently switch to E10 as well? I didn't know OR went E10. Seems to be getting more popular. That was the first thing I noticed that keyed me in to needing some kind of adjustment... my trims went to **** all of a sudden when they had been perfect for months.

crystal_Imprezav · 11-27-2006, 05:36 AM

I have been bothered by this for a long time as well, but luckly I can get the good stuff down here for most of the year. I personally like to calibrate the MAF tables up to the point of closed loop cross over. Increasing the MAF until right around the 3 volt area so that during all closed loop operation the trims stay as close to zero as possible. Of course this takes a lot of time but with out adjusting the upper MAF the important fuel and timing loads are not changed. In the fuel tables I set all the values to 14.7 or 1 lambda so that the ecu is targeting that during all closed loop loads and rpms. I also run a lot more timing in these areas to effectively make the best gas mileage as I can get. Playing with the AVCS in these areas can also help lean out or richen up your targets. With all that finally done, lowering the closed loop modified load and some of the delays throttle position calibrations will help the cross over point to hit exactly where you want it keeping the closed loop trims where you want them. Just my 2 cents on it and what I have done to calibrate the fuel. Effectively its almost the same as tuning alky/meth since it does spray at some points during closed loop operations where your trims go wild. I tried playing with the fuel table, injectors tables but none have yielded the same results as calibrations to the MAF table.

R4ND0M_AX3 · 11-27-2006, 07:46 AM

I asked this before, why don't we just use lamda instead of AFR and not worry about the fuel type?

WolfPlayer · 11-27-2006, 07:57 AM

Nick,

You aren't missing anything. I was (go figure)

After reading your post it hit me. I wasn't thinking in terms of lambda with regard to the ECU.

Now, with that said, the injector constant idea absolutely makes the most sense to me.

t

WolfPlayer · 11-27-2006, 08:00 AM

Quote:

Originally Posted by R4ND0M_AX3

I asked this before, why don't we just use lamda instead of AFR and not worry about the fuel type?

My guess ... guess ... is that more people are comfortable talking about AFR vs. lambda. However, lambda does make more sense. FWIW, I did some searches on innovate's forum and they have had similar discussions.

t

R4ND0M_AX3 · 11-27-2006, 11:20 AM

Quote:

Originally Posted by WolfPlayer

My guess ... guess ... is that more people are comfortable talking about AFR vs. lambda. However, lambda does make more sense. FWIW, I did some searches on innovate's forum and they have had similar discussions.

t

I'm switching. I'm gonna change my XD-1 to read lamda so I get used to it.

Guest_1u53r · 11-27-2006, 12:02 PM

Lamda For Life!

Christian. · 11-27-2006, 09:03 PM

Nick, WolfPlayer, and all. I have updated the worksheet to include some more goodies like BSFC, Injector Calculator, IDC calculations, some graphs, etc. The main thing I wanted to get across is how the stock 1:1 FPR works, let me know if that worksheet and graph makes sense? Fuel pressure is dynamic and I wanted to make sure that people tuning their vehicles understand that. We are currently working on the DISI engines and they are even more dynamic with fuel pressure...running anything from 60 psi to 600+ psi (some of the latest fuel systems will be operating at 1000+ psi!). The next version of the Tuning Worksheet is going to have some information on MAP sensor calibrations and some additional information about pressure conversions.

FWIW, I would just focus on fueling in Lambda...it really makes tons more sense, is easier to tune with, and will make life easier.

Take care,
Christian.

Quote:

Originally Posted by WolfPlayer

Fuel is in the low-det fuel map via AFRs.
Here's the deal with E10 as far as I can figure out on my own. The stock ECU is calibrated for petrol. That means that the stock O2 sensor is interpreted in a specific way by the ECU. Usually, this is done via 'switchover points' in mV. To configure the stoichiometric lambda for E10 gas, you would have to reconfigure the O2 crossover points (which by default means that you have the knowledge of knowing what mV corresponds to the stoich of E10). The O2 stuff isn't exposed in Enginuity as far as I know (I could be wrong though).
Long story short: Changing the injector constant won't get you what you need. The ECU would basically just apply a long term fuel trim to get the car back to where the ECU wants it. I would consider richening up the fuel map according to what your car needs. This will tell the ECU to shoot for a different fueling (captain obvious - yea, I know). I am suggesting this because E10 is working pretty good at 14.7 on my car in closed loop locations. I did have to richen up WOT areas accordingly. A good place to start is to use Cobb's E85 information (which they have ALSO supplied a column for E10!) contained in their spreadsheet worksheet linked on their website.
t

R4ND0M_AX3 · 11-28-2006, 07:21 AM

Quote:

Originally Posted by Christian.

...I have updated the worksheet to include some more goodies like BSFC, Injector Calculator, IDC calculations, some graphs, etc. ...let me know if that worksheet and graph makes sense? ...

There's an external link in the worksheet pointing to another XLS file I don't have.

Christian. · 11-29-2006, 12:18 AM

Quote:

Originally Posted by R4ND0M_AX3

There's an external link in the worksheet pointing to another XLS file I don't have.

I found the link and deleted it, a newer v1.13 should be up by tomorrow. Thanks for pointing that out.

Take care,
Christian.

11-26-2006, 04:54 PM	#1
Guest_1u53r Banned Fav Mod: knowledge Location: Bristow, VA Posts: 392 Join Date: Apr 2003 Trader Rating: (1)	Tuning for E10 - Injector Scalar or something else? Is it better to change the injector scalar to a 6.3% lower flow number or to change something else (MAF cal, fuelmap, etc) to help account for E10 fuel? I am using Enginuity on a mostly-stock 06STI. It makes sense to me that I still want my lambda targets to be almost the same, so I don't want to change the whole fuel map to try to account for this but instead to just reduce the scalar to increase fuel flow overall, including closed-loop. Also, I'm having trouble finding where to alter the closed-loop fuel targets in Enginuity. What am I missing? This ad is not endorsed by this member. Please register or login to hide this ad.

11-26-2006, 06:18 PM	#3
Guest_1u53r Banned Fav Mod: knowledge Location: Bristow, VA Posts: 392 Join Date: Apr 2003 Trader Rating: (1)	OK I have to point out some things here that I am sure you know, but may not have yet considered. First of all, E10 burning at its stoich point of 13.8 will produce an O2 lambda of 1.0. Petrol burning at its stoich point of 14.7 will also produce an O2 lambda of 1.0. Since we know that the map is really just lambda targets, we can also assume that no matter what fuel is used, closed loop fueling will be adjusted to achieve the target lambda (1.0 for the case of idle). This means if you adjust _nothing_, using E10 you will always be running at least 6% higher AF Learning trims than if you were running proper petrol because the ECU senses that the pre-tabled injector pulsewidth is not adequate to meet the lambda. I have seen this behavior personally... Car properly calibrated so that AFR trims are low everywhere, then fill up with E10 and watch the trims rise. Now, you mention that E10 on your car is working fine at 14.7, but I question this claim because A), 14.7 is vastly leaner than E10's stoich point and B), your logging software doesn't realize that you're using E10 and the sensor doesn't report AFR, it's reports lambda voltage. So an 'indicated' 14.7 using E10 is merely indicative that the software is using a petrol multiplier. In Innovate's Logworks, you can specify a different multiplier so that AFR is reported accurately, but neither Deltadash, StreetTuner, nor Enginuity actually have this option so any WBO2 logging done with them will be 'incorrect' in terms of AFR. You CAN display lambda instead of AFR though, but that would only tell you that it is still hitting the proper lambda. If you were to LOG the lambda, then use a multipler column in your logged spreadsheet, you would see that what you previously logged as "14.7", when properly calculated, is actually 13.8 Now if you just adjust the fuel maps to a 'richer' AFR value, you will run richer for sure, but realize that the displayed tables are interpreted by the software from a natively lambda value in the actual ROM. As such, adjusting the AFR in the displayed table is just adjusting the Lambda target and that is NOT what I want to do. Further, if you were to ignore this part and just adjust the fuel tables to account for the lower stoich point, it would probably run fine until the corrections from the closed loop learning process started to get applied to those higher load points and start throwing your values way off. You need a truly global adjustment for the changing fuel, imho. Adjusting the MAF table (when we are not actually seeing a change in airflow) will affect the calculated load and thus will affect timing and fueling improperly, IMHO. Adjusting just the open-loop fuel maps, needs to be accompanied by a corresponding adjustment in closed-loop targets (which is why I also asked about those, since it's not apparent where they are in enginuity). It strikes me that the most direct way to alter the injector pulsewidth is via the flow scalar. We want to spray more fuel (by volume) for a given airflow than before so you skew the ECU's idea of what pulsewidth flows what amount of fuel, thus increased injected volume. What am I missing in this reasoning? Last edited by nhluhr : 11-26-2006 at 06:21 PM.

11-26-2006, 05:17 PM	#2
WolfPlayer Moderator & ECU Tech Car: 05 CGM STI Fav Mod: mo boostaz Location: Fredneck, MD Posts: 3,703 Join Date: Jun 2004 Trader Rating: (6)	Nick, Fuel is in the low-det fuel map via AFRs. Here's the deal with E10 as far as I can figure out on my own. The stock ECU is calibrated for petrol. That means that the stock O2 sensor is interpreted in a specific way by the ECU. Usually, this is done via 'switchover points' in mV. To configure the stoichiometric lambda for E10 gas, you would have to reconfigure the O2 crossover points (which by default means that you have the knowledge of knowing what mV corresponds to the stoich of E10). The O2 stuff isn't exposed in Enginuity as far as I know (I could be wrong though). Long story short: Changing the injector constant won't get you what you need. The ECU would basically just apply a long term fuel trim to get the car back to where the ECU wants it. I would consider richening up the fuel map according to what your car needs. This will tell the ECU to shoot for a different fueling (captain obvious - yea, I know). I am suggesting this because E10 is working pretty good at 14.7 on my car in closed loop locations. I did have to richen up WOT areas accordingly. A good place to start is to use Cobb's E85 information (which they have ALSO supplied a column for E10!) contained in their spreadsheet worksheet linked on their website. t

11-26-2006, 08:08 PM	#4
RainMaker The Next Petter Solberg Car: 08 Z06 Corvette Fav Mod: Jack Jack Location: Portland, OR Posts: 883 Join Date: Oct 2005 Trader Rating: (3)	Damn. Suddenly I figure out wy my fuel trims suddenly went to 7-8 out of the blue. grumble

11-26-2006, 08:12 PM	#5
LOLSTi S204 Racer Car: RIP 02 WRX Fav Mod: 2004 Aspen White WRX STi Location: Cincinnati Posts: 3,621 Join Date: Dec 2005 Trader Rating: (1)	I hope Shell diafs. Great thread guys.

11-27-2006, 05:36 AM	#7
crystal_Imprezav Spec C Club Car: Slow 05 STi Fav Mod: Loping Idle Location: Chewy's Garage Posts: 1,881 Join Date: Mar 2005 Trader Rating: (12)	I have been bothered by this for a long time as well, but luckly I can get the good stuff down here for most of the year. I personally like to calibrate the MAF tables up to the point of closed loop cross over. Increasing the MAF until right around the 3 volt area so that during all closed loop operation the trims stay as close to zero as possible. Of course this takes a lot of time but with out adjusting the upper MAF the important fuel and timing loads are not changed. In the fuel tables I set all the values to 14.7 or 1 lambda so that the ecu is targeting that during all closed loop loads and rpms. I also run a lot more timing in these areas to effectively make the best gas mileage as I can get. Playing with the AVCS in these areas can also help lean out or richen up your targets. With all that finally done, lowering the closed loop modified load and some of the delays throttle position calibrations will help the cross over point to hit exactly where you want it keeping the closed loop trims where you want them. Just my 2 cents on it and what I have done to calibrate the fuel. Effectively its almost the same as tuning alky/meth since it does spray at some points during closed loop operations where your trims go wild. I tried playing with the fuel table, injectors tables but none have yielded the same results as calibrations to the MAF table.

11-27-2006, 07:46 AM	#8
R4ND0M_AX3 Grassroots STI Racer Car: 05 Subaru WRX STi Fav Mod: Cobb StreetTUNER Location: Elk Grove, CA USA Posts: 592 Join Date: Jul 2004 Trader Rating: (0)	I asked this before, why don't we just use lamda instead of AFR and not worry about the fuel type?

11-27-2006, 07:57 AM	#9
WolfPlayer Moderator & ECU Tech Car: 05 CGM STI Fav Mod: mo boostaz Location: Fredneck, MD Posts: 3,703 Join Date: Jun 2004 Trader Rating: (6)	Nick, You aren't missing anything. I was (go figure) After reading your post it hit me. I wasn't thinking in terms of lambda with regard to the ECU. Now, with that said, the injector constant idea absolutely makes the most sense to me. t

11-27-2006, 12:02 PM	#12
Guest_1u53r Banned Fav Mod: knowledge Location: Bristow, VA Posts: 392 Join Date: Apr 2003 Trader Rating: (1)	Lamda For Life!