Close

Results 1 to 7 of 7
  1. #1
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    7,980
    Rep Points
    8,872.8
    Mentioned
    626 Post(s)
    Rep Power
    89


    Reputation: Yes | No

    JB3 2.0 boost control

    Hey guys,

    One of our vendors had made a similar thread on another forum, and I figured I’d modify it a bit and repost it here for the BimmerBoost crowd. Lots of good information here and this can be a “home base” thread for discussing any 2.0 boost control questions you may have.

    The key to JB3 2.0 boost control is working with the ECU for boost control rather than trying to replacing it. Those of you who have driven good flash tunes like the GIAC know how safe, effective, and responsive the factory boost control system is when properly tuned. The JB3 retains everything the factory system does well, such as its heavily integrated safety systems and precise control, scales it up to handle higher boost targets, and then augments certain aspects such as response to improve your driving experience. This is all done through the existing integrated factory safety systems. For discussion purposes the JB3 boost control is considered integrated to contrast other systems which would be considered stand alone or hybrid depending on their details.

    With the JB3 the TMAP boost sensor output remains directly correlated to actual boost and conveys the tuned boost target to the ECU. It is never altered to be “what the ECU wants to see” or to avoid throttle closure. It is dynamically altered based on current boost levels and JB3 mapping to precisely reflect the tuned boost target. The TMAP sensor itself is scanned at ~200hz and replicated back to the ECU at ~200hz. When it comes to precise boost control there simply is no substitution for the raw speed of hard wired sensors. That is why BMW uses hardwired sensors for boost control. Replication based on CANbus data at 20-30hz simply isn’t fast enough.

    From a safety perspective with this arrangement the moment you go over your tuned target you’ll experience throttle closure. If you go too far over target you’ll experience a 30FE limp. If you never get up to target you’ll experience a 30FF limp. These systems are cornerstone to the safety BMW engineered in to the N54. It’s true it would be much easier to cheat and rescale or eliminate these systems. But it would not be responsible. Reduction of throttle closure must come from better boost control and keeping boost on target, not from redefining the rules and increasing the room for error. If a system properly controls boost there is no need to reduce throttle closure when over target as you’ll rarely go over target.

    With the JB3 the wastegates remain directly connected to the ECU and are tapped through load resistors. This ensures that the ECU is always aware of the actual duty cycle less what the JB3 is adding, and prevents the JB3 from “deciding” to run 100% DC and potentially damaging the motor in the process. This makes the nightmare scenario of run away boost or sustained 20+psi boost levels almost impossible at the HARDWARE level. It would be just too easy for a mapping mistake, software crash, harness problem, corrupted firmware, or some other mishap to allow the piggyback to run uncontrolled boost levels.

    So with all these safety nannies in place how is JB3 2.0 delivering lighting fast boost response with almost no throttle closure? The key is working within those systems to precisely control boost within the proper target window. We’ve developed algorithms to dynamically model what stock boost levels should be allowing the JB3 to ensure the ECUs boost control is rescaled to precisely hit these new targets. The beauty of this system is the modeling is done in PSI rather than actual duty cycle so changes in vehicle conditions, wastegate slop, bolt on modifications, air density, and even larger compressors are all automatically compensated for.

    Here are a couple logs to help visualize how the system works. The first is from an automatic car and the second from a manual. These are the examples Mike had in his post and I’m just going to run with them. I think they are from the older 1/24 firmware.

    These are the key elements for boost control:

    1) PID PSI (scaled to the left, 0-20psi), the light pink line, which represents the minimum allowable boost for these conditions.
    2) ECU PSI (scaled to the left, 0-20psi), the yellow line, which represents how much boost the ECU is observing.
    3) PWM, the dark pink line (scaled to the left, 0-20 = 0-100%), which represents the amount of direct wastegate control the JB3 is doing to keep the ECU properly ranged. As the yellow line dips below the pink line the JB3 must increase the duty cycle and physically increase boost. When the yellow line is above the pink line the ECU remains in full control and minimal JB3 wastegate intervention is required.

    Notice how in both cases, the yellow and light pink lines converge and closely follow each other, with PWM increasing or decreasing as needed to keep the error rate within a fraction of a psi. In the second log, from a manual car, you can see how quickly boost responds after a shift with minimal PWM input. This is with no bogfix in place, notice how PWM drops to only 10% during the shift. Under different circumstances, with different mods or weather, you may see PWM spike up to meet the target. It’s all completely dynamic now.

    Several variables are available to custom tune here also. Including how aggressive the JB3 is in targeting, how quickly it responds to requests for additional boost via throttle, minimum prespool/lagfix levels, etc. As we move further alone and out of beta we’ll publish a tuning guide and suggestions for those who find the out of the box settings too mild or wild for their taste.
    Attached Images Attached Images   

  2. #2
    Join Date
    Jan 2010
    Location
    SoCal
    Posts
    117,183
    Rep Points
    31,312.0
    Mentioned
    2056 Post(s)
    Rep Power
    314


    Reputation: Yes | No
    Ok, just wondering, the safety aspects you mention, pretty much all the tuning solutions leave the factory N54 ECU safety parameters in place, correct?

  3. #3
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    7,980
    Rep Points
    8,872.8
    Mentioned
    626 Post(s)
    Rep Power
    89



    Reputation: Yes | No
    That used to be true but much is changing. Other systems are now running stand alone boost control, sort of like an electronic boost controller. One of them reads the factory target via CANbus and claims that makes theirs safe but from my perspective has the same flaws as any other stand alone system. Names of systems have been intentionally omitted to prevent flame wars. Click here to enlarge

  4. #4
    Join Date
    Jan 2010
    Location
    SoCal
    Posts
    117,183
    Rep Points
    31,312.0
    Mentioned
    2056 Post(s)
    Rep Power
    314


    Reputation: Yes | No
    Click here to enlarge Originally Posted by Terry@BMS Click here to enlarge
    That used to be true but much is changing. Other systems are now running stand alone boost control, sort of like an electronic boost controller. One of them reads the factory target via CANbus and claims that makes theirs safe but from my perspective has the same flaws as any other stand alone system. Names of systems have been intentionally omitted to prevent flame wars. Click here to enlarge
    Even by omitting the name your implication alone could start the flame war Click here to enlarge I'm not going to argue, I don't know enough about this subject.

  5. #5
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    7,980
    Rep Points
    8,872.8
    Mentioned
    626 Post(s)
    Rep Power
    89



    Reputation: Yes | No
    Another random custom map 7 log, DCI only. Holding right around 15psi down to 13psi at the shift. In some cases customers may find 13psi at the shift too aggressive for them and can easily reduce that by increasing taper in the tuning interface.
    Attached Images Attached Images  

  6. #6
    Join Date
    Feb 2010
    Posts
    324
    Rep Points
    467.6
    Mentioned
    8 Post(s)
    Rep Power
    5


    Reputation: Yes | No
    Could you reproduce the above graph with spark advance plotted please?

  7. #7
    Join Date
    Nov 2009
    Location
    Simi Valley, CA
    Posts
    7,980
    Rep Points
    8,872.8
    Mentioned
    626 Post(s)
    Rep Power
    89



    Reputation: Yes | No
    I could post other ignition advance graphs, but that particular customer was logging out of the JB3. For ignition advance he would need to also be recording with the BT cable.

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may post replies
  • You may not post attachments
  • You may not edit your posts
  •