Page 6 of 7 FirstFirst ... 234567 LastLast
Results 101 to 120 of 126

Thread: E39 ECM running standalone on LDK | having trouble getting it running

  1. #101
    I think I've figured out the root cause. My ground wire for the sensor is sent to ground and not the ECM. Apparently the E39 has a pull up circuit on the O2 ground wire that is 0.9-1.1v (reference: https://forum.hptuners.com/showthrea...p-harness-Help)! That explains why I'm having to provide +1v to get a reading on the ECM. I have a new pigtail for the O2 sensor on the way that will go to the ECM and should solve this closed loop issue.

    With that out of the way I will indeed need to get a wideband. I see that innovative has a wide band setup that has a tunable analog output. Could I not send this output into the ECM signal and log the voltage? If I set the output to provide the same voltage at stoichiometric AFR that the normal O2 is providing then couldn't I just use the wideband for everything?

    I realize that it's a little ridiculous to work this long to find the root cause for the O2 sensor not working just to turn around and try to replace it with a wide-band, but I really wanted to make sure everything was working properly before going to the next step.

    This leads me to the next question: what is the conversion from O2 voltage to lambda? I must be blind because I can't find any AFR, equivalence ratio, or lambda function to log (other than equivalence ratio command). I also can't find a VE or target lambda table to adjust. What am I missing?

    Daniel

  2. #102
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    last i checked e39 had a narrowband so it wont produce a lambda sweep i believe .680 is stoich. these are off the top of my head though so likely not accurate.

    if you want to record in the scanner lambda then you can search for
    commanded lambda
    commanded afr
    or just keyword search either commanded, lambda, afr as individual words.
    Last edited by cobaltssoverbooster; 08-13-2018 at 07:01 PM.
    2000 Ford Mustang - Top Sportsman

  3. #103
    What does "commanded" mean? I found AFR and equivalence ratio commanded. Is this the conversion of the O2v to AFR/equivalence-ratio or is it the requested value from the ECM? If it's the requested value then how do we know what it is in real time?

    I agree an O2 sensor has an almost digital output. That's why I'm thinking I could use a wideband output in its place so long as the stoich voltage was the same. Does that make sense?

    Daniel

  4. #104
    Advanced Tuner
    Join Date
    Apr 2011
    Location
    Chicago, IL
    Posts
    763
    Commanded means the afr or eq ratio your commanding based off the pe ratio example - pe ratio of 1.250 is 11.76 afr -14.7/1.250 and eq ratio of .80 is a afr of 11.76 - 14.7*.80
    Only way to read wot fueling is with a wide band o2 sensor such as aem ...narrow bands 0-1v not accurate for anything other than stoich
    Wide band 0-5v or 10 afr to 20 afr

  5. #105
    Thanks for the reply. If commanded is what the ECM is asking for then what does it compare to? There must be an O2 voltage for stoich somewhere in the calibration....right? If I set the wideband output to produce this same voltage at stoich then would the E39 ECM be able to use the wideband for closed loop operation while still allowing me to log the full load AFR?

    Daniel

  6. #106
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    equivalence ratio commanded is the channel for what the ecu is commanding in lambda.
    the O2 mV breakdown from teh ecu is only to target stoich because its a 1 volt range and is specifically designed to be accurate at the stoich fuel range.
    thats why everyone has a wideband. the only GDI Ecotec to have a factory wideband is the LNF a20nht LDK range which utilized the e69.

    the wideband is only useful when used on a 5 volt channel so if you use its 1 volt channel then you are limiting its calibration accuracy. tried it in the past on other ecu's and it worked fine for targeting stoich but never really worked great in pe; too much drift.
    2000 Ford Mustang - Top Sportsman

  7. #107
    The lack of resolution makes sense. Does anyone know how high the voltage can go on the O2 signal before the ECM is damaged?

    Are there any other input channels on the E39 ECM that the wideband signal could be sent to?

    Is there any consensus on what are the best wideband options? I like that the innovative system allows you to adjust the analog output, but the AEM unit looks nice too. Does anyone know why the 4-channel AEM unit doesn't require open air calibration, but the x-series seem to still need this?

    Daniel

  8. #108
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    aem does it on startup...but you can force a free air calibration if you read the instructions.

    i use daytona and innovate sensors. they are expensive but i use these because they have dual bank units that i like for the v8 applications i currently work with.
    i did use aem sensors in the past when i owned my lnf cobalt and liked it too. for single bank applications i tend to go aem.
    2000 Ford Mustang - Top Sportsman

  9. #109
    I have closed loop! Sending the O2 sensor ground to the ECM is required. My LTFT is already at -25% at idle. It seems I was still running rich.

    My high pressure fuel pump is making a clicking sound at idle. Is that normal? I assume this is just the sound of the cam contacting the pump as it does sound like it's half engine speed.

    Having removed the balance shafts I've already discovered a few bolts that have vibrated loose. I'm having to double check everything now, but overall it seems to run well.

    Daniel

  10. #110
    Senior Tuner cobaltssoverbooster's Avatar
    Join Date
    Dec 2008
    Location
    Nevada
    Posts
    4,452
    the hpfp has a bleed and damper that make noise. sounds like a slow click. completely normal.
    2000 Ford Mustang - Top Sportsman

  11. #111
    Holly crap is it hard to keep my foot out of the throttle. I?ll post the log later, but I peaked at ~16.8psi map and 5000rpm and I thought it was already fast. I do appear to have a boost control issue though becaus the pre-throttle pressure was 25psi-a! This is one time DBW probably saved my ass because it was limiting map even though I was making mad boost. Clearly the WG isn?t opening.

    Wooohhooooo.

    Daniel

    *edit* pre-throttle pressure was 25psi-a
    Last edited by TurboWood; 08-15-2018 at 05:00 PM.

  12. #112
    log attached. Any thoughts?

    Daniel
    Attached Files Attached Files

  13. #113
    Tuner
    Join Date
    May 2014
    Location
    Metro Detroit
    Posts
    119
    congrats on getting it running!
    from glancing at the file it actually looks pretty good. the only problem is the bad fuel trims, but i think that's to be expected with the MAF being in a different pipe and all the tuning changes you've been making trying to get it started.

    I know why you want to replace your narrow band sensor with an aftermarket wideband, but as cobaltssoverboost has explained, it will not work.
    E39 was meant for a narrow band, the controls are designed for a narrow band, and a large number of calibrations that HPTuners does not give you access to are all set up for a narrow band and you can't change them. so the controller can not make use of data from an aftermarket wideband. If you don't have a HPTuners pro, and can't log your wideband, then some people have wired in the wideband in the AC pressure pin, but again, that's just for logging purposes.

    The pre throttle pressure sensor is seeing pressures higher than the desired boost. in areas where desired boost is below 160kPa-ish (9-ish psi above atmosphere) that is normal. Since the spring is pushing the the wastegate closed, you need about 160kpa-ish pressure before the boost overcomes the wastegate spring and starts opening the wastegate more. you can tell the E39 is still in control, because the "desired boost" -what it wants- is matching the manifold absolute pressure - what it measures - by controlling the throttle, this is normal. above 160kpa-ish pressures then desired boost should match the pre-throttle pressure sensor. so you might have overboost problems once you start pushing it harder but from this file it might be too early to tell. the one event at t=1:20:16 is the only spot where "boost pressure" signal is a little concerning. did you mess with the spring preload on the wastegate?

    if you think its fast now at 0.7g/cyl airmass, just wait until it's pushing 1.2g/cyl airmass . cylinder airmass is a good signal to start getting used to, just about everything is based on cylinder airmass, spark maps, cam phaser maps, torque model, just about everything.

    So in the end... the harness manufacturer slowed you down again? glad you got it running.

  14. #114
    Tuner
    Join Date
    May 2014
    Location
    Metro Detroit
    Posts
    119
    EDIT: forgot you had a different turbo. So ignore the ?160kpa -ish? you have to figure out your own spring pressure and wastegate cals.

  15. #115
    I really cannot say who was at fault here, the harness manufacturer or the vendor that sells the system. The harness manufacturer can claim they made it to the vendor spec. For me it?s odd that these two companies, that both claim to have made/sold harnesses for standalone 2L turbo w/E39, didn?t know about a few basic requirements, namely the need for pre-throttle P/T sensor or O2 grounding. These seem fundamental. I suppose you might be able to tune around these, but that would be a really screwy cal. Running in open loop all the time is believable, but not having manifold T sounds scary.

    Anyway, I do have to say that they both tried and remained supportive. Neither company gave up on me. Certainly it?s strange to have had these issues, but it?s still a good product that saved me a ton of time and allowed me to avoid the E69 ECM. I can?t imagine the troubles I would have faced trying to build my own harness and the quality was much better!

    If I can?t just use the wideband directly then having it logged in the A/C pin for tuning sounds like a good alternative. Do you know what pins I need to use?

    Thanks,
    Daniel

  16. #116
    Bringing my thread back from the dead!

    I have about 3k miles on the car now and had it tuned at Church Automotive in Long Beach back in April'19. The resulting dyno is attached.

    A few things I've noticed that bug me about the tune are:

    1) cold start and even warm idle speed. You can see in the attached log file that it ramps up over 3krpm at startup and never really gets below 1krpm. I logged the target idle speed and there is nothing out of place there (800rpm). I suspect this is linked to not having a VSS which causes the ECM to peg at 158mph constantly. I tried manipulating the external load tables and increasing the max VSS for startup idle, but couldn't detect any change. Is there anyway to improve the idle speed w/o adding a VSS?

    2) I think the car would be more drivable with a slightly different tq curve. Having peak tq at "low" rpm is making the car easy to spin up the tires and causes me to try modulating throttle (car will lose traction in 3rd w/o a passenger). I also can't tell how linear the relationship is between TQ and throttle position. When I talked with the team at Church it sounded like he may have set an excessively high tq target and then let power be controlled by the boost control. This makes me think that I'm hitting boost limit before WOT. Taking these two things into consideration I would like to try dropping peak tq a bit, raising boost at higher rpm, and figure out the tq vs. throttle relationship. Any guidance on how best to tackle that?

    My current tune and cold start log are attached. I tried uploading a log with driving in it, but it's almost 10MB and wouldn't load. Is there a limit?

    Thanks,
    Daniel
    Attached Images Attached Images
    Attached Files Attached Files

  17. #117
    Potential Tuner
    Join Date
    Apr 2021
    Location
    Cbus, Ohio
    Posts
    3
    Quote Originally Posted by TurboWood View Post
    1) cold start and even warm idle speed. You can see in the attached log file that it ramps up over 3krpm at startup and never really gets below 1krpm. I logged the target idle speed and there is nothing out of place there (800rpm). I suspect this is linked to not having a VSS which causes the ECM to peg at 158mph constantly. I tried manipulating the external load tables and increasing the max VSS for startup idle, but couldn't detect any change. Is there anyway to improve the idle speed w/o adding a VSS?
    Im currently working out the exact same issue.

    Finished installing an LNF with E39 ecu into my BMW E30 a few weeks ago. Had it tuned locally and the tuner was unable to get the idle and cold start working properly. Im working with ZZP right now and he's recommending getting the VSS operable. But then I talked to the guy that build my harness and he said of all the ones he's done no one has implemented VSS into their car that he's aware of. Im hoping to try and use the speedometer output from my BMW to input to the E39 ECU. Please email me at at durkyturk at gmail if you find a solution and Ill update here if I do as well.

  18. #118
    Potential Tuner
    Join Date
    Apr 2021
    Location
    Cbus, Ohio
    Posts
    3
    Just wanted to update this for anyone having similar issues in regards to cold start and idle speeds. I swapped to a ECU from a manual car. With the manual ECU and Regal VSS it showed zero mph. I then used a Dakota Digital SGI-100BT box to convert the 2 wire speed sensor from the solstice trans into a signal for the LHU ecu (pin 8 to the ECU, Output 4 on the DD box). This worked perfect and solved all my issues.

  19. #119
    Quote Originally Posted by cruizin01 View Post
    Just wanted to update this for anyone having similar issues in regards to cold start and idle speeds. I swapped to a ECU from a manual car. With the manual ECU and Regal VSS it showed zero mph. I then used a Dakota Digital SGI-100BT box to convert the 2 wire speed sensor from the solstice trans into a signal for the LHU ecu (pin 8 to the ECU, Output 4 on the DD box). This worked perfect and solved all my issues.

    I'm surprised by this. Can you double check that your calibration is still showing a manual transmission (in the general tab of the trans cal)? I've played with that some and when an automatic transmission is selected the VSS defaults to 0mph, but I found there were other drivability issues created when I did that.

    Was the change in ECU really needed? I would think feeding the speed signal would do the trick even if the default was 158mph. No?

    Thanks for the update!

    Daniel

  20. #120
    Potential Tuner
    Join Date
    Apr 2021
    Location
    Cbus, Ohio
    Posts
    3
    Quote Originally Posted by TurboWood View Post
    I'm surprised by this. Can you double check that your calibration is still showing a manual transmission (in the general tab of the trans cal)? I've played with that some and when an automatic transmission is selected the VSS defaults to 0mph, but I found there were other drivability issues created when I did that.

    Was the change in ECU really needed? I would think feeding the speed signal would do the trick even if the default was 158mph. No?

    Thanks for the update!

    Daniel
    My calibration is currently showing manual trans. With the old ECU it didn't matter what I selected it still showed 158mph.

    When working with the old ECU, it would not read the signal from the VSS. As soon as I switched to the manual ECU it registered zero.

    Now I will say I dont know if it registered zero WITHOUT the regal sensor hooked up. Its possible. I just know that without this manual ECU I couldn't do anything with the VSS.

    Both were the same OS btw.