Results 1 to 17 of 17

Thread: DFCO Not Working

  1. #1
    Tuner in Training
    Join Date
    Feb 2019
    Posts
    21

    DFCO Not Working

    This is a GMPP LS376/480 controller (E67). The cam, heads, and intake have been upgraded since the the original LS376/480 went in the car (a Miata).

    I'm trying to get the DFCO to work. This is mostly a street car, so I want the best mileage out of it I can get.

    I think the best place to look at the attached log is 42.3 seconds in ... according to what I think I have in my tune file, the DFCO should be active. What I suspect is that maybe my understanding of where to set the TPS thresholds is wrong. There are a lot of ways to read the TPS, I'm not sure which is supposed to be used in the DFCO control tables. I did try setting the DFCO enable value to 100 (of 200 max), and even that didn't work. Maybe the throttle has to transition from 101->99 for this to work, not sure. But I'm not using anywhere near that much throttle around here that time of year on mostly wet pavement with a grossly overpowered Miata.

    I attached the log and the tune. I have a screenshot of the most obvious area of the log where DFCO isn't activating... 42.3 seconds in.

    dfco.JPG
    Attached Files Attached Files

  2. #2
    Senior Tuner
    Join Date
    Jan 2014
    Location
    Australia
    Posts
    1,970
    make sure all the perameters are met, airmass, tps, speed will be one of them that may have shifted a little from mods and camshaft prob the airmass

  3. #3
    Tuner in Training
    Join Date
    Feb 2019
    Posts
    21
    From looking at the log, it seems that everything is met except TPS. I knew that before I posted.

    I was hoping somebody would be able to tell me what flavor of TPS is the one that's used in the table. Commanded throttle position? Relative? Actual ETC setting? The table takes values from 0 to 200.

  4. #4
    Senior Tuner
    Join Date
    Jan 2014
    Location
    Australia
    Posts
    1,970
    just lift it up until it stays in dfco, i only use pedal 0-100% and throttle body 0-100% dosnt really matter which one u use just raise it up so its like 1% over so any throttle input cancels dfco

  5. #5
    Tuner in Training
    Join Date
    Feb 2019
    Posts
    21
    So here you go. This is a capture of the log, and a screenshot of my DFCO set up.

    I'm in 6th gear, 66 mph, so that should enable DFCO. I'm at .088 cyl air at 1833 rpm, so that should enable DFCO. I'm at 2% relative throttle, and 13.7% SAE throttle, so either one of those is less than the 90% (!!!!!!) I set the TPS threshold to in order to try and take TPS entirely out of the picture. You can see I have ZERO delay for fuel cut.

    Still no DFCO. Does anybody know if the GMPP LS376/480 controller even supports it, or did they turn it off for the aftermarket (GMPP) controllers?

    YES, the temperature conditions and rpm conditions are also met.

    Log_Capture.JPG

    dfco_setup.jpg
    Last edited by grubinski; 12-30-2020 at 05:17 PM.

  6. #6
    Advanced Tuner
    Join Date
    Jun 2014
    Location
    Boulder, CO
    Posts
    201
    DFCO in my E67 from the LS376/430 crate engine wiring harness works fine. Looking at your tune and log I'm as confused by it as you are! Could you PM me where your MPG graph data come from?

  7. #7
    Tuner in Training
    Join Date
    Feb 2019
    Posts
    21
    I had posted in another thread and got some feedback that looked interesting. https://forum.hptuners.com/showthrea...l=1#post636430

    However, after altering the VSS setup in my tune from 17 teeth/rev and 3.73 rear end ratio to 40 teeth/rev and 1.59 rear end ratio as suggested by posts in that thread, the DFCO is still not engaging, even with the DFCO TPS enable set much higher than I see in other tunes (factory or modified). All of the other DFCO entry parameters are met.

    I think for now I may just set the high vacuum areas where I know the car is off throttle to be lean and call it the best I can do.

  8. #8
    Tuning Addict blindsquirrel's Avatar
    Join Date
    Apr 2017
    Location
    Meridian MS
    Posts
    7,451
    I'm just speculating, so don't take this as anything other than that.

    Compared to the initial stock file (via the change logs), you've got PCM Output Calibration at 126 per mile, versus stock of 111,600. Could that be screwing up some other internal calculation it's using for the DFCO?

  9. #9
    Tuner in Training
    Join Date
    Feb 2019
    Posts
    21
    I appreciate your willingness to speculate. With my old eyes, I read that as 126,xxx something. I'm doing some wiring on the car right now, so can't drive it for a few days. I'll look into this and try it out then. Thanks for the sharp eyes.

  10. #10
    Potential Tuner
    Join Date
    Dec 2008
    Posts
    9
    Subscribed
    I have a GMPP E67 as well.
    Same deal,
    Every year or so I go back to it trying things.
    LS7 T56
    Will be interested to see hat you come up with.

  11. #11
    Tuner in Training
    Join Date
    Jul 2018
    Location
    DENVER CO
    Posts
    20
    [ECM] 12264 - DFCO Entry Spark Blend Time. is what the delay is. reduce those numbers significantly and you can return the rest of your settings back to stock or corvette ls3 settings.

  12. #12
    Tuner in Training
    Join Date
    Apr 2020
    Location
    Germany
    Posts
    38
    Im also dealing with this issue and I need DFCO kicking in imediately to pass a emissions test.

    LS3 stock besides shortys and CAI (also stock converters)

    The DFCO entry spark blend time just defines the time between DFCO is (!!) activated and spark has been ramped down to the set value of the entry table. So lowering this time doenst change the time between leting of the pedal and DFCO kicking in. It just reduces the time between DFCO begins and entry spark is achieved.

    I can clearly see how the upstream O2s are getting pig rich. With the typical time difference for flooding the catalyser with HCs and travelling to the post O2s also they get rich and as end effect I see massive spikes in CO and HC values on the emission tester.

    Disabling DFCO also makes it rich, since the injectors will be driven with defined minimum inj pulse width. In my understanding DFCO overrides the inj pulse even if we dont see it on the scanner, at least O2 values are telling this story. The injectors seem to stay fully closed even if a pulse with and duty cycle is read. If DFCO is deactivated the O2s tell me that fuel is still added by getting rich after another (pre to post O2s).

    I need a solution to get rid of the 1,5s I see between getting of the pedal and entering DFCO. Once DFCO is running the signals are maxed out lean.

    I dont know if there is some hidden table or other value to delay the DFCO entry. I looked through tons of logs to check if it could be something like spark (in the second letting of the pedal), clutch status (depressed or not), TPS (what is an entry criteria, but somehow never met and even if this is raised to meet on entry time it seems to make no difference) and many others. I running out of ideas now but there seems to be very few confirmed knoledge about this topic.

  13. #13
    Tuner in Training
    Join Date
    Jul 2018
    Location
    DENVER CO
    Posts
    20
    It will be easier to help if you post a tune .but my first question would be about [ECM] 18698 - Minimum RPM Rate to Disable DFCO . Did you adjust this RPM?

  14. #14
    Tuning Addict blindsquirrel's Avatar
    Join Date
    Apr 2017
    Location
    Meridian MS
    Posts
    7,451
    The issue with the GMPP ECMs is, you change some setting and nothing happens. Normal rules do not apply, and there's no documentation on which changes are allowed and which ones get ignored.

  15. #15
    Tuner in Training
    Join Date
    Feb 2019
    Posts
    21
    Quote Originally Posted by blindsquirrel View Post
    I'm just speculating, so don't take this as anything other than that.

    Compared to the initial stock file (via the change logs), you've got PCM Output Calibration at 126 per mile, versus stock of 111,600. Could that be screwing up some other internal calculation it's using for the DFCO?
    Changed that, and it had no effect. I left it changed anyway because what I had was obviously wrong.

    Back to working on DFCO, the car is running really well driveability-wise other than that. DFCO is a "nice to have", not a "must have".

    I suspect as others have said that the GMPP E67 has this turned off. Also, my car has no clutch switch or brake switch going to the PCM, so maybe that is a problem ... though I'm not sure this PCM is even wired to accept those signals. I'll have to check.

    Thanks to everyone for all the input.

  16. #16
    RE: OP's question on which throttle parameter: I believe that the DFCO tables use the "Accelerator Pedal Position" parameter. That one should go to 0.0% when you aren't touching the pedal regardless of throttle blade position. May be worth a check that it is definitely at zero (have seen a couple pedals physically "sticking" just before the top of travel) could make idle and/or coast mode transition not happen as it should.

    Dang, coincidentally I just noticed earlier this afternoon the (E38) vehicle im working on isnt entering DFCO with all enablers hit anymore either!?! It was just a couple days ago on the same tune file. Every time I talk about something it acts up
    If it's some weird anomaly on mine I'll report back just in case it may be something new to look at.
    Last edited by CaudleDynamicsLLC; 08-29-2022 at 08:06 PM. Reason: meant to quote question...

  17. #17
    One other thing I just vaguely remembered about the last couple connect and cruise E67's I worked on, and actually saw in an LS1B ECM in a swap: even if the vehicle speed that the ECM is reporting is accurate, if the inputs used in the calculations in getting to that speed are out in left field, i've had things like mysterious traction control appear out of nowhere, add-on cruise control work intermittently, and would imagine speed based things like DFCO could too.
    Round ambiguous numbers for simplicity sake; example: if you used a trans output shaft speed sensor pulse per rev of 40, rear gear ratio of 3.50, and tire diameter of 30" you could have an accurate speed reported. OR if the speed sensor pulse per rev was way off, say 80 you could also have a crazy tire diameter input of 60" and it would still report the correct vehicle speed. But, if that raw trans pulse per rev value is used in some other speed based calcs it may be out of range. I dont know exactly what was going on behind the scenes or recall exactly which parameters it was, but just know ive had weird issues come up that finally solved when i realized someone fudged those kinds of values. Just throwing another idea out