Page 2 of 12 FirstFirst 123456 ... LastLast
Results 21 to 40 of 225

Thread: Tuning MAF and VVE at the Same Time

  1. #21
    Tuner in Training
    Join Date
    Jun 2022
    Posts
    18
    Yeah my VVE and MAF tables are all over the place with this method as well.

    With the MAF I'm 1%-2% off in a region then the next cell up is 13% off.

    VVE will have anywhere from -5% and up to 7% in the same zone.

    Stock tune seems like it should be better, so I'm likely doing something incorrect.
    Attached Files Attached Files
    2019 Chevrolet Colorado 3.6L LGZ

  2. #22
    Advanced Tuner Cringer's Avatar
    Join Date
    Aug 2021
    Location
    Somewhere smoothing your VVE table
    Posts
    481
    Quote Originally Posted by Nanserbe View Post
    Yeah my VVE and MAF tables are all over the place with this method as well.

    With the MAF I'm 1%-2% off in a region then the next cell up is 13% off.

    VVE will have anywhere from -5% and up to 7% in the same zone.

    Stock tune seems like it should be better, so I'm likely doing something incorrect.
    Can you post the before/after tunes/logs?
    A standard approach will give you standard results.

    My Tuning Software:

    VVE Assistant [update for v1.5]
    MAF Assistant
    EOIT Assistant

  3. #23
    Quote Originally Posted by Cringer View Post
    Can you post the before and after tune as well as the log file?
    I can post a log and/or screenshots in the near future, after I record another log. FYI I wasn't using your tool, but just commenting on the idea that the VVE may be in effect at steady state. As mentioned I adjusted only the VVE in this particular zone and left the MAF alone, so after the next log I'll see how much of an impact the change had.

    Just took a look at this log in question though and I see MAF airflow, dynamic airflow and VE airflow all near equal at the 1200 RPM column (0.45 - 0.60 pressure ratio); at 1400 RPM (same pressure ratio) the MAF airflow and dynamic airflow are near equal but the VE airflow is about 4 - 5% higher; in all case my lambda was 1.07 - 1.08. Also as mentioned, when the MAF was reading about the same airflow/frequency as in these areas but in a different zone my lambda was much closer to 1.00, with all three airflows reading about the same. Anyway, I'll update when I record another log.

  4. #24
    Potential Tuner
    Join Date
    Apr 2022
    Location
    Michigan
    Posts
    9
    Use Filters and yes this MAF method will work with just LTFT and STFT (math). Filter out the transients
    E-85 Flex Fuel, Texas Speed 2'' primary ceramic Long Tube Headers, High Flow Cats, LT2 Intake Manifold, LT5 Soler Performance Throttle Body, Rotofab, ZL1 1LE wheels are tires, Silvers coil overs, BMR strut brace, Black lugs, Recaros, Red belts and kick panels, HP Tuners, AEM Wideband, Sunroof and Bose, Elite Catch Can, Diode Dynamics smoked LED sidemarkers, NPP exhaust, Graphite black ground effects, 3 inch Drop, T93 Unlocked

    https://www.youtube.com/@ericknickerson
    Please subscribe

  5. #25
    Update from my last post:

    Did another scan and sure enough the zone in question (1000 - 1500 RPM; 0.45 - 0.60 pressure ratio) that I increased by 7 - 8% was still out by 7 - 8%, but this time the VE airflow that was near equal to the MAF and dynamic airflow (1200 RPM) was now a few % greater (didn't get any 1400 RPM driving where the VE airflow was already off but I suspect it would be even further off). Going to change the VVE back to where it was at 1200 and decrease at 1400 RPM so that VE airflow matches MAF.

  6. #26
    Advanced Tuner Cringer's Avatar
    Join Date
    Aug 2021
    Location
    Somewhere smoothing your VVE table
    Posts
    481
    All, just to make you aware, ttz06vette had messaged me stating that this was going great in CL, but PE was giving him fits. That is when I realized I made a mistake in my PE formulas. I have revised the formulas in post #5. So if you had issues thus far in PE, please update.

    If you are still having issues, the only way I can offer assistance is to get the before and after tunes, as well as the before and after logs. I will be happy to take a look. PM me if you don't have my email (in case you don't want to publicly post your secret sauce tune).
    A standard approach will give you standard results.

    My Tuning Software:

    VVE Assistant [update for v1.5]
    MAF Assistant
    EOIT Assistant

  7. #27
    Tuner in Training
    Join Date
    Jun 2022
    Posts
    18
    Quote Originally Posted by Cringer View Post
    Can you post the before/after tunes/logs?
    Yep, I will do a before and after log and post them and the tunes here.
    2019 Chevrolet Colorado 3.6L LGZ

  8. #28
    Advanced Tuner ttz06vette's Avatar
    Join Date
    Sep 2008
    Location
    Charlotte, NC.
    Posts
    242
    Quote Originally Posted by Cringer View Post
    All, just to make you aware, ttz06vette had messaged me stating that this was going great in CL, but PE was giving him fits. That is when I realized I made a mistake in my PE formulas. I have revised the formulas in post #5. So if you had issues thus far in PE, please update.

    If you are still having issues, the only way I can offer assistance is to get the before and after tunes, as well as the before and after logs. I will be happy to take a look. PM me if you don't have my email (in case you don't want to publicly post your secret sauce tune).
    PE is working now. I seem to be having good luck with these formulas. If I compare the results to Smokeshows results as well as just using trims for CL and wideband error, the results are close. The only places I am staying away from are the outliers such as low MAF Hz areas where I don’t have MAF enabled. I am getting large deltas in these cells where VE should be weighted the most. The VE CL tables look good but MAF has me pulling a ton of fuel. I hesitate to adjust that according to the results even though Dynamic Air validates the delta. Overall the car is running good. Still some variation log to log but nothing major.

  9. #29
    Advanced Tuner Cringer's Avatar
    Join Date
    Aug 2021
    Location
    Somewhere smoothing your VVE table
    Posts
    481
    Quote Originally Posted by ttz06vette View Post
    PE is working now. I seem to be having good luck with these formulas. If I compare the results to Smokeshows results as well as just using trims for CL and wideband error, the results are close. The only places I am staying away from are the outliers such as low MAF Hz areas where I don’t have MAF enabled. I am getting large deltas in these cells where VE should be weighted the most. The VE CL tables look good but MAF has me pulling a ton of fuel. I hesitate to adjust that according to the results even though Dynamic Air validates the delta. Overall the car is running good. Still some variation log to log but nothing major.
    Awesome! What I found is the Smokeshow/GMVE method seemed to really cause large (artificial) fluctuations in the histograms. Meaning it would show really rich, or really lean, which never really agreed with the actual fuel trims which were really mild. My only thought is that the GMVE formula is using the calculated Manifold Air Temp which seems to fluctuate a lot during different engine loads. I have always meant to circle back to this and pop in the ambient air temp or IAT PIDs for comparison sake.

    Since you are turbo'd I am curious if you are see the same thing with the GMVE_CL and GMVE_PE formulas?
    A standard approach will give you standard results.

    My Tuning Software:

    VVE Assistant [update for v1.5]
    MAF Assistant
    EOIT Assistant

  10. #30
    Advanced Tuner ttz06vette's Avatar
    Join Date
    Sep 2008
    Location
    Charlotte, NC.
    Posts
    242
    Yes except in steady state cruise. I get the same fluctuations using IAT though outside steady state even with filters and DFCO off. I think there is just a lot of noise going on with Min Airflow and short pulse. For me as long as Steady state and MAF PE boost are dialed in and safe I am happy. I run Meth so that's another factor to deal with in open loop boost. I literally can change my AFR's via the meth controller in a certain RPM/MAP range with the meth curve by having it come in sooner and faster.

  11. #31
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by Cringer View Post
    Awesome! What I found is the Smokeshow/GMVE method seemed to really cause large (artificial) fluctuations in the histograms. Meaning it would show really rich, or really lean, which never really agreed with the actual fuel trims which were really mild. My only thought is that the GMVE formula is using the calculated Manifold Air Temp which seems to fluctuate a lot during different engine loads. I have always meant to circle back to this and pop in the ambient air temp or IAT PIDs for comparison sake.

    Since you are turbo'd I am curious if you are see the same thing with the GMVE_CL and GMVE_PE formulas?
    That sounds more like a filtering issue and/or lack of steady state data collection. The simultaneous method is just a dumb equation...it still needs you to make sure the data it receives is accurate. Ensuring steady state will account for the 'moving target' MAT issue but it will not correct for a steady state offset in the modeled charge temp that may be present due to hardware changes versus the stock configuration. If you could get the problem vehicle on a loaded dyno and hold a sustained mass flow to monitor for unexpected variation in fueling feedback, it would be possible to back-calculate updated bias and filter tables for the model. I haven't done a write-up on it though...you and maybe 5 other people would be the only users lol. I attached a simulink model to one of my recent posts that provides visualization of MAT as it changes based on model inputs, if you're curious what that looks like.

  12. #32
    Advanced Tuner Cringer's Avatar
    Join Date
    Aug 2021
    Location
    Somewhere smoothing your VVE table
    Posts
    481
    Quote Originally Posted by smokeshow View Post
    That sounds more like a filtering issue and/or lack of steady state data collection. The simultaneous method is just a dumb equation...it still needs you to make sure the data it receives is accurate. Ensuring steady state will account for the 'moving target' MAT issue but it will not correct for a steady state offset in the modeled charge temp that may be present due to hardware changes versus the stock configuration. If you could get the problem vehicle on a loaded dyno and hold a sustained mass flow to monitor for unexpected variation in fueling feedback, it would be possible to back-calculate updated bias and filter tables for the model. I haven't done a write-up on it though...you and maybe 5 other people would be the only users lol. I attached a simulink model to one of my recent posts that provides visualization of MAT as it changes based on model inputs, if you're curious what that looks like.
    Smokeshow, I am open to any suggestions and pointers you may have. This example is a scenario that I saw time and time again when I logged on this car. A strange lean spot in the histogram at 3200 rpm. Then when I look at the Chart vs Time, things get really crazy. You can see the ECM is removing fuel (-3.18%) while the GMVE_CL calculation is showing that I need to add fuel (7.53%).

    Then when we look at the MAF, VVE, and Dynamic Airflows, you can see VVE is 64.41 g/s (which is higher than both the MAF and Dynamic Airflow). So something isn't right. If you see anything funky that I have done wrong, please correct me.

    This is the math I have for GMVE_CL:
    Code:
    (([16.71.avg(500)]*(1+(.01*([6.156.avg(500)]+[7.156.avg(500)])+.01*([8.156.avg(500)]+[9.156.avg(500)]))/2)-[50070.56.avg(500)]/60*4*[11.92.avg(500)]*[2312.avg(500)]/[2126.240.avg(500)])/([50070.56.avg(500)]/60*4*[11.92.avg(500)]*[2312.avg(500)]/[2126.240.avg(500)]))*100
    This is the filter I am using in the Scanner:
    Code:
    [2517.161.avg(1500)]=0 and [2517.161.avg(-200)]=0 and (abs([50090.156.slope(1500)])+abs([50090.156.slope(-500)]))<2 and ([6310]=9 OR [6310]=14 OR [6310]=15)=0
    MZoze4q.jpg

    jmJXQVV.jpg
    A standard approach will give you standard results.

    My Tuning Software:

    VVE Assistant [update for v1.5]
    MAF Assistant
    EOIT Assistant

  13. #33
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    First thing I'd suggest is ensuring the LTFTs aren't populated with poor data. I disable them on every vehicle during fueling calibration and only re-enable them if I'm certain they aren't going to 'wind up' in operating conditions where the OEM hardware would normally produce reliable O2 feedback. An example would be long tube exhaust scavenging with a larger camshaft. The short term will come and go as needed, but long terms can learn in a correction that may be erroneous. Only the mildest mods I've found allow leaving LTFTs enabled, and even then, they're best used with limited authority over fueling error. In short, the fuel feedback algorithms are designed to manage a stock engine...so the control strategy itself can be insufficient to manage certain mods, regardless of the calibration effort.

    That may or may not be involved in the discrepancy between the result of GMVE_CL and fuel trims. In any case, there looks to be a transition happening there. Not a transition that violates steady state airflow criteria, but something in the volumetric efficiency calculation. Can't tell much based on those snips alone. Would need to see TPS, MAP, temps, engine speed and the VVE value.

  14. #34
    Advanced Tuner Cringer's Avatar
    Join Date
    Aug 2021
    Location
    Somewhere smoothing your VVE table
    Posts
    481
    Quote Originally Posted by smokeshow View Post
    First thing I'd suggest is ensuring the LTFTs aren't populated with poor data. I disable them on every vehicle during fueling calibration and only re-enable them if I'm certain they aren't going to 'wind up' in operating conditions where the OEM hardware would normally produce reliable O2 feedback. An example would be long tube exhaust scavenging with a larger camshaft. The short term will come and go as needed, but long terms can learn in a correction that may be erroneous. Only the mildest mods I've found allow leaving LTFTs enabled, and even then, they're best used with limited authority over fueling error. In short, the fuel feedback algorithms are designed to manage a stock engine...so the control strategy itself can be insufficient to manage certain mods, regardless of the calibration effort.

    That may or may not be involved in the discrepancy between the result of GMVE_CL and fuel trims. In any case, there looks to be a transition happening there. Not a transition that violates steady state airflow criteria, but something in the volumetric efficiency calculation. Can't tell much based on those snips alone. Would need to see TPS, MAP, temps, engine speed and the VVE value.
    Long terms were disabled during this logging as well as countless others. I also saw the same reported lean spot during easy/smooth and/or quick throttle transitions. Even after adding a major spike to the VVE table (which made for a terrible transition between boundaries) it did not solve the issue. Yet when I disabled the MAF and went to pure SD there was no lean spot. I think GMVE is a good staring point, but it is flawed in some way.
    A standard approach will give you standard results.

    My Tuning Software:

    VVE Assistant [update for v1.5]
    MAF Assistant
    EOIT Assistant

  15. #35
    Advanced Tuner
    Join Date
    Aug 2010
    Location
    Detroit, MI
    Posts
    926
    Quote Originally Posted by Cringer View Post
    Long terms were disabled during this logging as well as countless others. I also saw the same reported lean spot during easy/smooth and/or quick throttle transitions. Even after adding a major spike to the VVE table (which made for a terrible transition between boundaries) it did not solve the issue. Yet when I disabled the MAF and went to pure SD there was no lean spot. I think GMVE is a good staring point, but it is flawed in some way.
    Can't really follow along with you without data.

  16. #36
    Advanced Tuner
    Join Date
    Jun 2014
    Location
    Boulder, CO
    Posts
    201
    I've had some success with adjusting values in the Dynamic Airflow Corrected Gain table. When I see a transient I'd like to work on, I find out what zone of the VVE table it maps to and try adjusting that zone in the corrected gain table. Higher values help fill in lean transients, lower numbers help avoid going rich. Look at a stock ZL1 tune for ideas.

  17. #37
    Tuner in Training
    Join Date
    Oct 2021
    Posts
    12
    Ive been trying hard with the GMVE formula as well. I was fighting something similar where the short term and the gmve were seperating. spent the time and redid the gmve formulas and it seemed to clean it self up. now the closer i get the gmve fuel trim the closer the stft is.

    Trying also to use the gmve set with the wide bands for pe mode. which is close to Maf Eq.

  18. #38
    Advanced Tuner Cringer's Avatar
    Join Date
    Aug 2021
    Location
    Somewhere smoothing your VVE table
    Posts
    481
    Quote Originally Posted by Capt Chewy View Post
    Ive been trying hard with the GMVE formula as well. I was fighting something similar where the short term and the gmve were seperating. spent the time and redid the gmve formulas and it seemed to clean it self up. now the closer i get the gmve fuel trim the closer the stft is.

    Trying also to use the gmve set with the wide bands for pe mode. which is close to Maf Eq.
    Can you post your GMVE_CL formula so I can compare to mine?
    A standard approach will give you standard results.

    My Tuning Software:

    VVE Assistant [update for v1.5]
    MAF Assistant
    EOIT Assistant

  19. #39
    Advanced Tuner ttz06vette's Avatar
    Join Date
    Sep 2008
    Location
    Charlotte, NC.
    Posts
    242
    I ignore these outliers that clearly point to something else going on and focus on a visually smooth graph and transition in the area of the anomaly. If the area in question is not repeatable on the next log I usually discount it. Also I have found that increasing the cell count sometimes gets rid of them. Obviously the exception is if it is in multiple cells and signifies a pattern.

  20. #40
    Tuner in Training
    Join Date
    Oct 2021
    Posts
    12
    Cringer sent pm with formula
    Last edited by Capt Chewy; 05-05-2023 at 06:23 PM. Reason: typo