Changed Tune from MAF to SD mode and suddenly AFR went horrid, Help Please

[SultanHassanMasTuning]

use your stock LS2 map sensor
E40 does not accept negative values in map scale from memory

also as mentioned your LS3 injector data is off

[Dave_In_VA]

Ok, not sure, but got all changes into tune file.

However, its idling too lean 16x to 17s.

I added appx. 11% of the MAF calibration Airflow vs Freq. Low table, across the entire low table. The attached log is after adding the 11% extra fuel.

Ideas anybody on how to get the idle AFR towards 14.7?

I'm seeing that VE and Mass airflows are really different.

Dave

[Dave_In_VA]

Ok, not sure, but got all changes into tune file.

However, its idling too lean 16x to 17s.

I added appx. 11% of the MAF calibration Airflow vs Freq. Low table, across the entire low table. The attached log is after adding the 11% extra fuel.

Ideas anybody on how to get the idle AFR towards 14.7?

I'm seeing that VE and Mass airflows are really different.

Dave

Ok, I had a couple of setting that were incorrect, in particular my Injector Offset Select was set to MAP with no associated MAP values :-(

It's running better now, but basically I'm starting the tuning process over.

However, this time from a solid foundation. Thanks everybody!

[Dave_In_VA]

https://forum.hptuners.com/showthrea...orrect+airflow

Yes. The idea is if your MAF is available as a reference point for airflow, then it can be used to correct the VVE. No fuel trims are even necessary unless the MAF calibration is not accurate.

Question, I have an E40 ECM that does not do VVE.

Does the above guide work for E40?

Dave

[eXo3901]

Yes. Once injector and sensor data is correct then as long you understand what you are adjusting, correctly gather, and then filter your data. It should work.

At the end of the day, airflows should closely mirror each other on a chart. Put MAF airflow, Dynamic airflow, and the calculated airflow from the GMVE values (do not log VE airflow, log only GMVE which is a direct lookup to your VE table values) in a chart. If your fueling(airflow) is correct, all three values will closely overlap with mainly the MAF value going off course during tip-in/out. You should see dynamic airflow follow closer to VE during those events. ?Assuming the dynamic airflow settings have not been monkeyed with to much.

[Dave_In_VA]

All, the attached logs don't seem to make sense to me.

In MAF mode, the ECU seems to pull select VE airflow over MAF airflow. I'm 98% ((( ???) certain I've turned off all SD items and in MAF mode only.

For example, at time 11:50:09, VE was 1.62, Mass was 1.89, and dynamic selected 1.617.

Another example, at 11:50:09, VE was 1.90, Mass was 2.28 and dynamic selected 1.896. This resulted in an AFR of 15.07 when commanded was 14.7. If more Mass had been selected the AFR would have been lower.

Anybody got any ideas?




Separately, I've gotten several recommendations to go back to the OE VE and MASS tables with new injector and MAF parameters. I tried this last night, AFR was very rich. To do this, I used OEM VE table, and the LS7 Mass tables. I cannot use the OEM Mass tables as they are designed for the round LS2 housing Mass sensor and I now have an LS3/7 blade style MAF sensor and the LS7 tables are the best approximation when changing the LS2 E40 to blade MAF. This resulted in a very, very rich at idle, about 11:1 AFR.

[eXo3901]

You still do not have the correct MAP sensor data. You are using an LS3 sensor, you need to input LS3 sensor data.
Your LS3 MAF sensor has an IAT sensor and unless you are using a breakout harness you need to use LS3 MAF sensor IAT data. The data you entered seemed to match the breakpoints for an LS3 sensor, but the resolution was reduced significantly, for whatever reason.
You had the MAF fail HI set to 10hz I believe it was.
You did not copy all of the tables blindsquirrel provided you.

I made these changes in the attached file along with some others that I think should be made. Give it a try or at least copy over the IAT, MAP, and remaining injector data.

Also, some will say to not log both MAF airflow and VE airflow at the same time as it could skew data. So, remove VE airflow and log the VE (mg k/kpa) pid, use that to calculate your VE airflow.

[Dave_In_VA]

You still do not have the correct MAP sensor data. You are using an LS3 sensor, you need to input LS3 sensor data.
Your LS3 MAF sensor has an IAT sensor and unless you are using a breakout harness you need to use LS3 MAF sensor IAT data. The data you entered seemed to match the breakpoints for an LS3 sensor, but the resolution was reduced significantly, for whatever reason.
You had the MAF fail HI set to 10hz I believe it was.
You did not copy all of the tables blindsquirrel provided you.

I made these changes in the attached file along with some others that I think should be made. Give it a try or at least copy over the IAT, MAP, and remaining injector data.

Also, some will say to not log both MAF airflow and VE airflow at the same time as it could skew data. So, remove VE airflow and log the VE (mg k/kpa) pid, use that to calculate your VE airflow.

@eX03901

I believe stock LS3 injectors (ACDelco 12576341) have a stock minimum injector pulsewidth value of 0.797. I have verified this in numerous stock LS3 tuner files.

The suggested tune shows 0.344.

This is the value an internet tuner suggested for me.

Should I use 0.344 or 0.797?

[blindsquirrel]

That 0.344 was what was in your original files. Min PW is not really that important... it's just a limit it won't allow less than. Min PW is different in different files across different platforms, even when they all use the exact same injector#.

[Dave_In_VA]

That 0.344 was what was in your original files. Min PW is not really that important... it's just a limit it won't allow less than. Min PW is different in different files across different platforms, even when they all use the exact same injector#.

Thank you

[Dave_In_VA]

You still do not have the correct MAP sensor data. You are using an LS3 sensor, you need to input LS3 sensor data.
Your LS3 MAF sensor has an IAT sensor and unless you are using a breakout harness you need to use LS3 MAF sensor IAT data. The data you entered seemed to match the breakpoints for an LS3 sensor, but the resolution was reduced significantly, for whatever reason.
You had the MAF fail HI set to 10hz I believe it was.
You did not copy all of the tables blindsquirrel provided you.

I made these changes in the attached file along with some others that I think should be made. Give it a try or at least copy over the IAT, MAP, and remaining injector data.

Also, some will say to not log both MAF airflow and VE airflow at the same time as it could skew data. So, remove VE airflow and log the VE (mg k/kpa) pid, use that to calculate your VE airflow.

@eX03901

I loaded and idled for a few minutes using the suggest tune, exactly as suggested.

After a couple of minutes the car became really lean > 17:1, so I shut it down.

Also Mass airflow is significantly higher than VE airflow. I find it runs much better when VE and Mass are close.

For the first time since mods, it idled pretty stable lower mid 800 RPM range. Before it would only idle close to 1000 RPM. Kinda of rough, but that can be refined once VE and Mass airflow are dialed in.

Should I add more VE or try to tune for Mass?

Attached is a log of the idle.

To all helping, THANK YOU!

[eXo3901]

Add to MAF. I attached another file that added to MAF and increase idle air a little, it assumes you used the last file I posted in that log. No need to mess with idle air yet if you do not want to, its better to get the MAF/VE corrected first.

Not sure why the VE airflow PID shows so low. Again, could be an issue logging both MAF and VE airflow at the same time. I calculated VE airflow based on the VE(mgk/kPa) PID and am showing the direct opposite. Calculating GMVE values from the MAF and Dynamic airflows show lower GMVE values than the VE table. Once corrected for the AFR error (was about 16% increase to MAF) the GMVE from MAF was closer to the VE table. So, based on that if you put this thing into SD mode and failed the MAF, I would expect this same log to show a little rich.

[Dave_In_VA]

Add to MAF. I attached another file that added to MAF and increase idle air a little, it assumes you used the last file I posted in that log. No need to mess with idle air yet if you do not want to, its better to get the MAF/VE corrected first.

Not sure why the VE airflow PID shows so low. Again, could be an issue logging both MAF and VE airflow at the same time. I calculated VE airflow based on the VE(mgk/kPa) PID and am showing the direct opposite. Calculating GMVE values from the MAF and Dynamic airflows show lower GMVE values than the VE table. Once corrected for the AFR error (was about 16% increase to MAF) the GMVE from MAF was closer to the VE table. So, based on that if you put this thing into SD mode and failed the MAF, I would expect this same log to show a little rich.

@eXo3901 Thank You, I'll give this a try tonight.

[Dave_In_VA]

Ok, made good progress on tune this weekend. See attached screenshot.

The MAF low and high errors are all within 2% except for up high where they are a little richer.

Assuming I can trust my VE table, there're ally pretty decent except at 800 RPM where its a little rich and on deceleration which will be handled by deceleration settings once they are turned on.

Can I trust this VE table? The car is in MAF mode.


I would guess my next step is either SD or GMVE tuning to get the VE table a little cleaner.

Alternatively, the idle is a little rough, so maybe I could focus on that for a while.

Thoughts anybody?

[eXo3901]

Can I trust this VE table? The car is in MAF mode.

I would guess my next step is either SD or GMVE tuning to get the VE table a little cleaner.

Alternatively, the idle is a little rough, so maybe I could focus on that for a while.

Thoughts anybody?

Not above 3,600, yet. That VE table was just a starting point to get it smoothed out from that previous log/tune.
Based on the attached log and tune, my suggested VE table adjustments would be, in order.(I am looking at the table with kPa as columns):

1. Highlight the entire 400rpm row - reduce by 15%.
2. Highlight part of the 800rpm to 1,200rpm rows from 15kPa to(and including) 71kPa - reduce by 10%
3. Highlight part of the 1,200rpm to 3,200 rpm rows from 94kPa to(and including) 105kPa - increase by 3%.
4. Highlight part of the 2,400rpm to 2,800rpm rows from 26kPa to(and including) 88kPa - increase by 3%.
5. Highlight the entire 3,600rpm row - increase by 7%.
6. Highlight the entire 4,000rpm row - increase by 8%.
7. Highlight the entire 4,400rpm row - increase by 9%.
8. Highlight the entire 4,800rpm row - increase by 10%.
9. Highlight the entire 5,200rpm row - increase by 11%.
10. Highlight the entire 5,600rpm row - increase by 12%.
11. Highlight the entire 6,000rpm row - increase by 13%.
12. Highlight the entire 6,400rpm row - increase by 14%.
13. Highlight the entire 6,800rpm row - increase by 15%.
14. Highlight the entire table and click smooth vertical bounds two times.
15. Highlight the entire table and click smooth horizontal bounds two times.

800rpm at 54kPa should end up around 1,335-1,355 when done.

That end result will be a relatively smooth table and should be a good starting point if you want to switch to SD only tuning or use the MAF/VVE method. You will need to address your IAT temps at some point. At idle they climb pretty fast and it could be effecting fueling and/or timing. IAT advance table is basically always pulling timing with the current setup.