KTM 990 Idle Speed Control Stepper Motor Operation

Discussion in 'Dakar champion (950/990)' started by BillyD, Jul 6, 2010.

  1. mousitsas

    mousitsas Long timer

    Joined:
    Oct 10, 2008
    Oddometer:
    3,992
    Location:
    Athens, Greece
    However, between lets say 0C and 50C the resistance of copper (and the corresponding drop in voltage) is increased by 10%.
    So it is not unrealistic to drop out of this "safe" tps range between hot and cold.
  2. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    I had to think about this one a bit, but your story makes perfect sense. I would just let vehicles relearn idle in my stall, without resetting idle learn adaptations as it will on almost all (automotive, but it's the same damn system, on this bike) vehicles and it would do this very thing often.

    In your case what probably happened is the TPS was out of "idle parameters" based on the ECU's currently learned adaptations through many learn cycles. Thing is, there's a safety zone of voltage built in to each side of the actual TPS voltage that the ECU will interpret as being at idle, but there is also the base point that it has to learn to put the "idle parameter" buffer around, with voltage buffers on each side as long as you don't get completely crazy voltage wise, as I have no idea what max allowable idle voltage is, in the 990s ECU algorithms. In your case the TPS was high enough the ECU thought you were on the throttle and the stepper motor chased (more in a second) a perceived open throttle. The algorithms are such as to let them see the RPM change was from the stepper, and so it would drop it back down, it bounces back and forth as it learns but they have probably identified that feedback loop frequency as well and the ECU starts to move the learned "idle parameter" box after watching feedback loops that let it know, it is off. This is partially due to the stepper motor chasing the throttle a little bit like Salzig posted on the first page for reasons that would be (my guesses):

    1 - To ease off of massive deceleration a clapped shut throttle, a big torquey motor like this one, wallops you with, when you are spinning at anything over say 7,000 RPM. You clap the throttle shut in a turn where you were pushing the limits and you are going to light up a tire, or low side, you can lose control pretty quickly (at least this is part of my theory for stepper motor chasing the throttle) so this programming is likely viewed as a throttle "shock absorber" by catching the butterflies early, and easing them off at a speed TBD. In most conditions a slowly closing throttle is safer though it could still be really fast...maybe under a second even for a 2K RPM change. I would like to see it with my own eyes moving. My guess is, it is a safety feature and additional rider comfort. Not everybody...including me, can always harness the raw power of these motors, so it really is a decent thought/addition to the manageability of the motor. YRMV

    2- A stepper motor chasing the butterflies is also a way to let you get back on the throttle without as much lurch, as the plates are held open slightly so you aren't truly going from nothing to cracking them open in a dicey situation where you need throttle finesse, whether in the dirt or on the pavement, they are trying to make the throttle a little less choppy when you get back on the go juice. I have no idea how well it works, until I remove my stepper shaft which I intend to, but I will run it first WITH the stepper (haven't ridden in 4 months had knee surgery one week ago :) ) and then see if I can feel a big increase in deceleration after removal.

    Fun little motorhead geek test!

    :bmwrider

    To your question about putting a resistor in place of the TPS, not sure what you would be trying to chase, and or diagnose, with that method, but it doesn't seem to have much merit in my mind initially (no offense ;) ) for diagnosis. Maybe not such a great idea, but maybe I just don't understand what you are trying to accomplish. ;)

    :hmmmmm

    The problem with doing additional adjustments to these bikes by chasing the voltage on the TPS while sitting on the stepper (not how the factory manual recomends) is a touch scary to me because when you adjust the boxes after the fact, like people have been doing, you are hoping that you know more then the Engineers who are way freaking smarter than I am, in knowing what the tolerances of moving them around are, compared to where they engineered the buffers on each side to be, as we don't even know if the "idle parameter" buffers are the same size on each side, of the learned base idle, which would be easy to write into the algorithms. I try not to be smarter then the guys that put this intricate system together and know ALL the variables we have no way of knowing. Dem is some smart peeps.

    Cheers!

    :beer

    Todd

    Edit: I need a factory service manual, for an 2011 990. I thought I had one but I just have the owners manual. Link anywhere?
    Benduro likes this.
  3. mousitsas

    mousitsas Long timer

    Joined:
    Oct 10, 2008
    Oddometer:
    3,992
    Location:
    Athens, Greece

    Well, bare with me electronics are not my forte!
    Lets say you mechanicaly disconnect the tps from the throttle bodies but it is kept electrically connected.
    Would it be possible in this way, on an bike haunted with idling hickups, to learn something about the s/m?
    Perhaps nothing changes by keeping the tps voltage steady, or maybe something does happen as you turn it to a different steady voltage and maybe this voltage is the voltage you have to set it for a good idle.
    The main idea is to take out of the equation the tps change as the idle oscilates.
    chevtech likes this.
  4. mousitsas

    mousitsas Long timer

    Joined:
    Oct 10, 2008
    Oddometer:
    3,992
    Location:
    Athens, Greece
    Also, has it been established which inputs are fed to the s/m to do what it is supposed to?
  5. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    It doesn't really work that way, because if you change the TPS off the bike it will add (or subtract) fuel and fuck up the idle anyway, and the base idle voltage is ALWAYS learned regardless. Small changes to the TPS will just make it learn a new postion, and do not neccesarily correspond to the "best" position. The best position ( 99.9% of the time) is having the buffers exactly where the engineers intended as they have all the variables, and can work what the best position is given the heat changes etc, while designing the system. We will never be privy to the actual algorithm matrices that control this system.


    SMs are notoriously intermittent (as in; it's not going to fail when you test it) and notoriously hard to diagnose. You have to have a labscope hooked up to the SM ECU inputs and know that the motor didn't move when commanded. It's virtually impossible to diagnose an intermittent SM. In my years of wrenching I have replaced many SMs and never a failed ECU for stepper woes. It is the ONE part on a vehicle that if it is intermittent, I will throw the part on first, diagnose (if necessary) later.

    The best test (that I have found...I would love to be enlightened, but I think they are just hard) of an intermittent or completely failed stepper motor is to command it through full range with a scan tool.


    :beer


    Todd
    Benduro and mousitsas like this.
  6. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    No, but we would need to know the voltages the SM runs on as well, to try and test them manually. Not even sure it can be done, without some kind of signal generating tester...I have forgotten the electrical inputs that make these work, and am not even sure if they are all the same.

    :dunno


    Todd
    Benduro and mousitsas like this.
  7. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    The "safe range" they give us, is not the actual range of identifiable idle voltage. It is exactly that, a safe range for consumers/mechanics to shoot for, that has much larger buffers around it to take into account ACTUAL TPS voltage "idle parameters". Engineers now how to build in a true safety range that takes into account heat expansion, and heat based resistance increase.

    We on the other hand, don't know diddly about what those actual parameters are, so best to just shoot for the middle value of what they give us!


    ;)


    Todd
    mousitsas likes this.
  8. mousitsas

    mousitsas Long timer

    Joined:
    Oct 10, 2008
    Oddometer:
    3,992
    Location:
    Athens, Greece
    Todd, you have exhausted me lol
    And I thought I had something to add in this thread :imaposer
    Back in my cave for now and thank god I ride a carburated katoom!
    chevtech likes this.
  9. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    Hahaha
    Hahaha! Surgery last week dude! What more do I have to do besides dump the random misfiring contents of my brain on a thread that I find intersting!

    LMAO!

    :y0!

    Cheers!

    :beer


    Todd
  10. Themastermike

    Themastermike Think you caught me in a coma Supporter

    Joined:
    Jan 20, 2011
    Oddometer:
    2,911
    Location:
    So Ill
    You can work on my bike in your spare time

    Sent from my SCH-I535 using Tapatalk
    chevtech likes this.
  11. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    The service manual I have doesn't even have a section 7-51 WTH?

    :baldy
  12. chevtech

    chevtech Long timer

    Joined:
    Jun 14, 2016
    Oddometer:
    1,176
    Location:
    Hood River OR
    I have so much to do on my bike and I can't even work on it. :(

    Hopefully soon.
  13. WrldRiderWC

    WrldRiderWC Adventurer

    Joined:
    Oct 7, 2007
    Oddometer:
    53
    Location:
    Reno, NV

    Stepper motors from my experience are motors that run in open loop. PWM power is sent to the motor, a pulse make the motor step or increment a certain degree and direction. They are generally 3-phase. It can also hold a position when the power is not modulating. They are incremental as opposed to absolute. So if it stalls or on power up they they need to homed.
    speedy 1 and chevtech like this.
  14. Stmad

    Stmad Adventurer

    Joined:
    Sep 22, 2010
    Oddometer:
    10
    Location:
    Athens , Greece
    It is in 2003 - 2007 Repair manual .
  15. speedy 1

    speedy 1 Wizard

    Joined:
    Jan 26, 2013
    Oddometer:
    566
    Location:
    Dark Coast Southern Oregon
    This is my understanding as well.
    Typical robotic application has a limit switch the motor hit during the start up phase to remind the stepper where it is at.
    These bikes don't seem to have that limit switch.
    So the question is : for the KTM, what is the procedure for reminding the stepper where "0" is ?
    chevtech and WrldRiderWC like this.
  16. Salzig

    Salzig Long timer

    Joined:
    Oct 17, 2008
    Oddometer:
    1,249
    Location:
    Cremona, Italy
    Three simple letters: TPS
    chevtech and WrldRiderWC like this.
  17. Themastermike

    Themastermike Think you caught me in a coma Supporter

    Joined:
    Jan 20, 2011
    Oddometer:
    2,911
    Location:
    So Ill
    Simple but confounding.......

    Sent from my SCH-I535 using Tapatalk
  18. Salzig

    Salzig Long timer

    Joined:
    Oct 17, 2008
    Oddometer:
    1,249
    Location:
    Cremona, Italy
    The letters are simple.
    The knowledge behind... not so much
    chevtech likes this.
  19. speedy 1

    speedy 1 Wizard

    Joined:
    Jan 26, 2013
    Oddometer:
    566
    Location:
    Dark Coast Southern Oregon
    Good call ...
    With robotics, during start up procedure the controller will see a "zero" which then sets absolute home position.
    With the KTM "TPS" there is never a zero set point. The voltage is set between .56 ~ .64
    I'd speculate if the controller saw a zero it would turn the injectors off causing the engine to stall.
    chevtech likes this.
  20. StevenD

    StevenD Hmmmm, dirt!

    Joined:
    Aug 3, 2005
    Oddometer:
    1,505
    Location:
    Holland, Amserdammed
    The IS a 0 point that cycles at every initial power on.
    The zero for the stepper is the stopping bolts at full closed minus a hair of the plates. That's why their position is what determines the tps ref. voltage for the ECU.

    Verstuurd vanaf mijn SM-G935F met Tapatalk
    chevtech likes this.