Results 1 to 14 of 14

Thread: MAF tuning issues - Running lean on decel/light throttle causing bad data, need help

  1. #1
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18

    MAF tuning issues - Running lean on decel/light throttle causing bad data, need help

    Hey everyone,

    I'm running into an issue with trying to tune the MAF on my setup but getting bad data from my histogram. This seems to be caused by the engine running lean under decel and very light throttle but I can't figure out what's causing it to go lean to give the bad data. This is causing me to not really be able to effectively tune the MAF.

    Also, something that's werid is I'm seeing the fuel trim cell be constantly switching between 21 and 22.... I would think it should stay in 20 (O/L) but I may be mistaken.

    To set up to tune the MAF, I:
    • Disabled the COT
    • Disabled the DFCO by increasing enable temp to 284?
    • Disabled LTFT and STFT's (O/L idle only)
    • Disabled dynamic air by setting high RPM disable to 100 rpm
    • Tried changing a few other DFCO settings to make sure it's disabled


    Here's my current tune and a log file of the most recent drive:
    10_18 second drive.hpl
    LS6 Working Copy - MAFStep6.5 - trying to kill DFCO for good - retry.hpt

    If anyone has any ideas that would point me in the right direction let me know! Thanks!

  2. #2
    did you check for a vacuum leak?

  3. #3
    Senior Tuner Lakegoat's Avatar
    Join Date
    Jul 2012
    Location
    Florida
    Posts
    1,454
    Anytime you are driving and let off the gas, you are going to get some lean spikes. There are some threads on here with formulas to eliminate the lean spikes if you want to do that. If not, try steady driving in a lower gear to hit a lot of cells, and before you are ready to let off the gas, hit the spacebar and shut off the scanner. Works for me, but I only tune my own car.
    2000 Camaro SS 2015 L83 port injected, Whipple 3.0, 4L80E, 8.8 Ford
    2013 Silverado 5.3, 6L80k 8.8

  4. #4
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    jsfrederick, I did check and can't find any apparent leaks at idle. May add some hose clamps to the nipple caps that I've got blocking off the driver side vac ports I'm not using just in case I have a leak under higher vacuum

  5. #5
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    Lakegoat, I was expecting some lean spikes but it seems like it's getting leaner than it should for having DFCO off. I do think I need to tune for the lean spikes that I'm having regardless though, I have some pretty substantial bucking/jerking under very light throttle loads and on/off throttle that seem to be a sort of lean surge condition. I'm going to play with some formulas/filters off of TPS % and rate of change to try to clean up the data to see how that works. Have you ever had to deal with lean surge/bucking at all while tuning your car?

  6. #6
    Senior Tuner Lakegoat's Avatar
    Join Date
    Jul 2012
    Location
    Florida
    Posts
    1,454
    Problem with tuning for lean spikes is they don't show up in the same place every time. Why do you want to tune decel spikes? Smooth your ve table---or just put the stock one back in there. smooth the spark table, its jumping all over the place. After you smooth those, an easy to get it driving smoother, quick is to turn back on stft's and see what they tell you about the fuel. I don't have any bucking .
    2000 Camaro SS 2015 L83 port injected, Whipple 3.0, 4L80E, 8.8 Ford
    2013 Silverado 5.3, 6L80k 8.8

  7. #7
    Senior Tuner mbray01's Avatar
    Join Date
    Dec 2013
    Location
    Slidell, La.
    Posts
    1,015
    honestly your going to be better off tuning with narrowbands than the wideband while cruising
    let the ecm tell you what it wants, and adjust using long terms and short terms. the ecm already has some filters built in to ignore these,
    very rarely will you tune with a wideband, then turn the fuel correction back on and it not be all over the place.
    save yourself some headache and tune normal driving with the trims
    Michael Bray
    Rusty Knuckle Garage
    Slidell, Louisiana
    20yr Master Tech.
    Advanced Level Specialist
    Custom Car Fabrication, Customization, High Performance.
    GM World Class Technician
    Shop Owner

  8. #8
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    Lakegoat, I was starting to notice that which was throwing me for a loop. I want to tune the decel spikes because I'm having some really rough on/off throttle transitions and bucking at low loads and highway cruising. That's interesting though - the VE table and the spark table I actually haven't touched at all, I was going to start with the MAF and didn't get further. Do you think they need that much work? I think (might be wrong since I don't 100% know the history of this motor/ecu) these are the stock tables so that strikes me as odd.

    mbray01, Interesting I was starting to wonder if it would be smart to just use the NBs instead. That's great to know though, I suppose I had the "you need a wideband to tune" rhetoric thrown at me so much I really took it to heart. Do you think trying to tune with the wideband has been getting me further from a good tune? I do kind of feel like when I turned fuel corrections back on it was running a little worse than before.

  9. #9
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    Lakegoat, I do see what you're talking about with the spark table though, it does look far from smooth honestly looking at the graph.... Could that really be the stock table?! Better go to the repository and look at a few stock tunes....

  10. #10
    Senior Tuner Lakegoat's Avatar
    Join Date
    Jul 2012
    Location
    Florida
    Posts
    1,454
    Spark is not stock, VE is not stock.
    You might want to post your mods also.
    2000 Camaro SS 2015 L83 port injected, Whipple 3.0, 4L80E, 8.8 Ford
    2013 Silverado 5.3, 6L80k 8.8

  11. #11
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    I pulled up a few stock CTS-V tunes and you're totally right, it seems like mine aren't stock at all. I guess that figures, I got the engine out of a gently crashed CTS-V and I wasn't sure what all of the mods were that were done to it and if it had been tuned or messed with before. After reading on here of all the messed up tunes people have gotten from shops and whatnot, I should have checked but it didn't even cross my mind to be honest.

    I should probably add this to my sig too but it's got a C6 LS7 intake with LS3 card style MAF and Kooks long tube headers to 2.5" dual/x-pipe exhaust.

    Maybe I should try to use the VE and spark tables from a stock tune and see where that gets me? I would think that I'll definitely need to tune MAF and VE due to the mods that I do have but maybe that'll get me closer to doing so.

  12. #12
    Potential Tuner
    Join Date
    Jul 2020
    Posts
    9
    I applied a filter to my VE and MAF calibration log

    [50091.156]>0

    basically when the apps is zero it stops logging. Im running a manual so my afr fluctuates as a decell depending if im still in gear or not so my data was all over the place. this helped clear that up.

  13. #13
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    cgmrdc, I wish I could do that type of filter - I did a DBC conversion which I'm starting to see some of the potential downsides of. Using TPS to try to filter has been less than ideal so I'm back to the drawing board on that one. That's good to know that that's normal behavior though.

    I did look through my whole tune to try to flush out everything that I could find that wasn't stock to be able to start back as close to a fresh slat as possible. I found that it looks like this pcm has previously been given the half-ass tune treatment by messing with the injector flow rate table and everything. I'm pretty sure the injectors are stock. One thing that I found that was mildly terrifying is that the misfire detection min ECT was set to be maxed out... WTF??? I set most of it back to stock but left my MAF/IAT tables and some other stuff alone and am going to try that out to see how it runs.

    EDIT:

    I double checked and the injectors are at least stock equivalents - it has Bosch 0280155931 injectors which allegedly cross reference to the AC delco 12561462 injectors. I'm feeling like mostly stock settings to start out with are going to be the best move here... We'll see when I have some daylight to drive the car.
    Last edited by armcaleer; 10-23-2020 at 05:34 PM.
    1997 Mazda Miata - LS6 (2004 CTS-V), C6 LS7 Intake, Kooks longtubes, T56 Magnum-F, Getrag G80 3.23, DBW to DBC conversion

  14. #14
    Tuner in Training armcaleer's Avatar
    Join Date
    Jan 2020
    Location
    Knoxville, TN
    Posts
    18
    Well, got a good update. With stock numbers put back in all of the tables that were jacked up, it runs SO MUCH BETTER!! The on/off throttle jerkiness is almost gone unless you're in too high of a gear and just overall it runs so much better. Looking at the log, the fueling is much closer to correct through every scenario that I put it through. Thanks for the catch that I didn't have stock tables!!

    Here's a log and my tune with all of the jacked up tables set back to stock:
    LS6 Working Copy - MAFStepX - With Repository Stock Tables.hptc7 race.hpl

    I was out playing with it on the highway to feel the cruising throttle response and a C7 surprised me by running up behind me, and [in my best Jeremy Clarkson voice] a bit of a race broke out....

    Now that I've taken care of that and a few exhaust leaks (dang flex pipes) I think I might finally have some better success tuning the MAF and VE which might not be that far off anyway. Runs much better and then AFR readings are way closer to target and much less erratic.
    1997 Mazda Miata - LS6 (2004 CTS-V), C6 LS7 Intake, Kooks longtubes, T56 Magnum-F, Getrag G80 3.23, DBW to DBC conversion