HELP O2 sensor removal and Y pipe

Discussion in 'GS Boxers' started by CaptRick, Jun 29, 2007.

  1. CaptRick

    CaptRick On dry land 4 now.....

    Joined:
    May 14, 2007
    Oddometer:
    7,150
    Location:
    Santa Fe NM
    OK I am trying to install a Y pipe on my 1150GS. I got the muffler off, no problem, dropped the cat but I cant get the dam O2 sensor loose for the life of me!!!
    Before I resort to violence, am I missing something really simple????:gun2
    Any one else have a problem with the dam thing?:baldy :baldy
    It looks like a simple nut to turn and loosen????
    Help please I want to ride tomorrow AM early.....
    #1
  2. GB

    GB . Administrator

    Joined:
    Aug 16, 2002
    Oddometer:
    61,699
    If memory serves me :gerg, you have to trace the O2 sensor wire to it's source, which I believe is under the fuel tank :scratch and unplug it from there, then spray the O2 sensor in the cat with WD40 or Liquid Wrench and let it soak in and then you shold find it'll come out with ease. Install it in the Y pipe, etc..

    You could take the easier way and try to get the O2 sensor without unplugging it first, but the twisting of the wire might break it.
    #2
  3. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>If you have the sensor unplugged, slip a 6 or 12 point box-end wrench on there, tip the catalyst onto it's side, and give the wrench a whack with a hard metal-head hammer. Just remember, lefty loosey :D

    Reinstall the sensor into the y-pipe with any flavor of anti-seize. Any flavor will work because at max torque, the exhaust temp isn't ever much more than 600°C right at the cylinder port. And is cooler by the time it reaches the sensor. After the catalyst it warms up again.
    <BR><BR>
    #3
  4. CaptRick

    CaptRick On dry land 4 now.....

    Joined:
    May 14, 2007
    Oddometer:
    7,150
    Location:
    Santa Fe NM
    *#%$@#%^&*&^%()
    Blast dam O2 sensor!!!!
    Arghhh walk the plank yer scum dog
    Ill nail yer tits to the table bitch
    happy place happy place.......
    anger managment.....
    Ok I took offf the gas tank unpluged the sensor and willlet it soak over night with some liquid wrench, if in the AM I cant get it loose its over to the dealer for a new sensor. Hope its a stock item......
    On the plus side the bike is getting a good dusting and cob web removal!
    Oh the joys of a BMW, as the germans say YOU WILL DO IT MY WAY!!!
    #4
  5. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>It'll likely be a stock item at the BMW store. But there's another option.

    If the sensor cannot be removed, you do not need to buy the OEM BMW replacement part. It's a 4 wire heated O2 sensor that's identical to several hundred million installed in vehicles since time began. The only difference is the electrical connector. (And I think the sensor body comes in 2 different physical sizes)

    Auto parts stores have 4 wire and 5 wire UNIVERSAL replacement O2 sensors with a wire splice kit included. Cut the OEM cable and connector from the old sensor, and splice in the new sensor.

    To avoid the blank stare at the parts store when you say "Its for a BMW motorcycle", tell them its for a 2002 BMW 320. Not that I know that that sensor is the right physical size mind you, but rather it avoids triggering the reflexive response of "We don't have motorcycle parts, sir"

    Bring the catalyst with you and physically match up one of the universal O2 sensors. Really there's only like two different flavors.

    The 4 wire colors are White-White-Black-Gray. The whites are the heater, not polarity sensitive. Gray is signal ground, and Black is the signal.
    <BR><BR>
    #5
  6. CaptRick

    CaptRick On dry land 4 now.....

    Joined:
    May 14, 2007
    Oddometer:
    7,150
    Location:
    Santa Fe NM
    Thanks poolside! :thumb
    If my cycle dealer is out I will hit the auto zone. I had no idea it was a universal thing. I would also think the universal will be a little less coin since it wont come over the magic parts counter. It's magioc cause it turns penny parts into gold plated parts!
    #6
  7. BobFV1

    BobFV1 Long timer

    Joined:
    Dec 31, 2005
    Oddometer:
    23,807
    :huh Oh my!

    At least give the WD40 a try:evil
    #7
  8. GB

    GB . Administrator

    Joined:
    Aug 16, 2002
    Oddometer:
    61,699
    WD40 or Liquid wrench should do the trick, no need for violence :lol3 Just be sure to use a good size wrench and the cat is held in place so it doesn't move.. get a helper! The hardest part is getting the tank off to unplug it. When reinstalling it the sensor, you might want to use some anti seize compound on its threads just in case you have to remove it again without resorting to threats of violence! :lol2
    #8
  9. CaptRick

    CaptRick On dry land 4 now.....

    Joined:
    May 14, 2007
    Oddometer:
    7,150
    Location:
    Santa Fe NM
    NAPA to the rescue!
    Naturaly the dealer didnt have a sensor in stock and when I saw what he wanted for it,($ 186.00) I was glad he didnt.

    Went down to NAPA found a bosch 4 wire sensor for $ 92.00 and rewired the plug to the GS plug.
    Seems to work just fine! The old sensor is still welded in place, I used every lubricant known to mankind , nothing worked, Even had a helper and a breaker bar, no luck. Any one need a cat with a built in O2 sensor!

    Any way back is back together, and Im changing all the fluids, Engine castrol synthetic, trans and FD BMW synthetic.

    Tomorrow it's time for a ride! :thumb
    #9
  10. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>Glad it worked out for ya maytey.



    Something important about using penetrating oil to remove an O2 sensor.

    The backshell of the sensor has an vent that's open to the atmosphere. The location of the vent 'hole' may not be obvious. The vent allows atmospheric oxygen to reach the backside of the sense element.

    If, via spraying oil through the vent, the sensor element is covered with penetrating oil well, that is a bad thing.

    Blocking the atmospheric oxygen from the O2 sense element more or less 'inverts' the sensor output. You know, the sensor 'senses' the difference in levels between the oxygen in the air, and the oxygen in the exhaust.

    Another way to say that is, it's comparing a high oxygen environment on the outside, to a low oxygen environment on the inside.

    Blocking the O2 from the back side of the sensor 'inverts' the comparison. Instead of making a comparison to 'oxygen', the sensor is making a comparison to 'no-oxygen'.
    <BR><BR>
    #10
  11. bykemike

    bykemike "ready to navigate"

    Joined:
    Apr 30, 2006
    Oddometer:
    493
    Location:
    St. Augustine FL
    Is the sensor really needed? I was told ,at some point, that if the sensor is removed the ECU goes to a midpoint default setting and the operation of the bike is unaffected. Any thoughts on this??
    Mike
    #11
  12. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>
    Yea pretty much largely unaffected Mike.

    The ECU oscillates the mixture between 'either side of stoichiometric', and normally uses the O2 sensor feedback to do it. This is true for the 11xx/1200

    If the ECU does not have O2 sensor input, it uses real-time crank speed feedback to get to the same place.

    The mixture 'swings' accomplished by the ECU using rotation velocity feedback end up being the same as those accomplished with O2 sensor feedback.

    The difference is, on the load/rpm grid map, the the number of load/rpm points where closed-loop O2 sensor feedback is effective, is larger than the area where closed-loop crank speed feedback is effective.

    Said another way, using the O2 sensor for feedback, the motor is operated in closed-loop over a larger number of loads and rpms. Crank speed feedback is otherwise the same, except for a smaller closed loop area.


    Some mythology: Rich mixture 'limp-home' mode.
    <BR><BR>
    #12
  13. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>
    Check your PM inbox
    <BR><BR>
    #13
  14. CaptRick

    CaptRick On dry land 4 now.....

    Joined:
    May 14, 2007
    Oddometer:
    7,150
    Location:
    Santa Fe NM
    Dam Poolside, That was mouth full. Im still scrathing my head to figure out what you said!:loaded :loaded

    In any case thanks for your suggestion about a different O2 sensor. The NAPA guy was a big help also. He just went through the stock untill he found a 4 wire Bosch for me. When he said how $ it was I laughed and told him he was half of what the dealer wanted!
    I am gonna hang on to the old cat just for shits and giggles in case I ever want to sell the Y pipe.
    #14
  15. pdoege

    pdoege Been here awhile

    Joined:
    Jan 17, 2006
    Oddometer:
    399
    Location:
    Orlando, FL
    We need part numbers for the O2 sensor Capt!

    Please?

    Peter
    #15
  16. Anorak

    Anorak Woolf Barnato

    Joined:
    Jul 23, 2004
    Oddometer:
    39,138
    Location:
    OAK
    Jim, using the rotational velocity to read mixture, is the crank speed change in the course of one revolution at set load being sensed?
    #16
  17. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>
    I don't know if I understand the question exactly, if I do, yes. At any given load, the crank velocity does change as the injector pusewith is adjusted above and below stoichiometric.

    With or without an O2 sensor, the ECU does not operate in closed loop (i.e., adjust the mixture back and forth across stoichiometric) over the entire load map.

    And as you can imagine, without an O2 sensor, the mixture isn't read directly. Instead, crossing the 'stoichiometric threshold' is inferred via crank velocity.

    You prolly know how the ECU does that using an O2 sensor. It's done in about the same way using crank velocity.

    The difference is, instead of the ECU adjusting the injector pulsewidth up or down for each combustion cycle while looking at the O2 sensor output, instead ECU looks at portions of the crank rotation time. One too many insteads there.

    Adjusting the pulsewidth at any given load to find either combustion limit (maybe call it blocking) takes no more than a few dozen combustion cycles.

    Say the ECU finds the 'low limit' of combustion by shortening the injector pulse until the percentage of irregular combustion events reaches 15%. That's three out of 20. Irregular doesn't mean combustion did not happen. It means that for 15% of combustion events, the crank speed between TDC and BDC is lower than the other 85%.

    If I know you, I'm betting you can work it out from there :D
    <BR><BR>
    #17
  18. Anorak

    Anorak Woolf Barnato

    Joined:
    Jul 23, 2004
    Oddometer:
    39,138
    Location:
    OAK
    Thanks Jim. You understood my question. As I was trying to phrase the question I was able to visualize the process so I started to understand it. Next question, how much resolution does the motronic on the 1150 have. Can it read less 360 degrees?
    #18
  19. johnjen

    johnjen I've Been Resigned

    Joined:
    Nov 28, 2001
    Oddometer:
    16,695
    Location:
    Seattle
    Ahhhh the beginnings of another technical article with some gusto.
    How does the moronic brain REALLY work…
    Inquiring minds need to know…

    Carry on… :deal

    JJ
    #19
  20. Poolside

    Poolside Syndicated

    Joined:
    Apr 1, 2003
    Oddometer:
    11,887
    Location:
    Silicon Beach, CA
    <BR>
    Sure thing. At a minimum, microsecond resolution. And yes, sub-degree resolution.


    But adjusting the mixture using the crank velocity, like we're talking about, is usually a pure timing operation. Meaning that a conversion of time to degrees or rpm isn't usually done in order to do the task of sensing crank 'travel time'.

    Here's some FYI . . The 2 Hall effect sensors give 4 measuring points, or marks say, around the 360 circle. 0° - 135° - 180° - 315° As you know, the crank pulley trips the Hall sensors. The pulley has a 45° 'window' in it that passes over the Hall sensors. The 'window' is fixed to the crankshaft, and spans the distance between TDC and 45°.

    The 45° slices are a reasonably close guess using a degree wheel and a meter on the Hall sensor output. The mark locations are not related to the ultimate degree resolution capability of the processor. The additional marks are needed for other reasons, scheduling the ignition event for one.

    The Motronic times out fuel injector pulses with 1µs resolution, that's 1MHz. To calculate and time an event with microsecond resolution, while doing other tasks, requires the processor to operate at a much higher speed. I do not remember the speed of the crystal running the Siemens processor, if I remember it was 20 or 32MHz. That seems like a reasonable minimum speed to carry out ECU tasks with enough headroom to spare.

    As the injector time is counted with 1µs resolution, it's reasonable to consider that, at a minimum, 1us is also the resolution of the crank timing. Though it's easily possible that rather than time a complete 360° crank revolution, that instead the 0°-135° slice is timed at a much higher rate. Perhaps at a resolution as much 16 times finer than 1µs.

    I realize I just said I'm guessing, but my guesses and min/max estimates are in the middle ground of usual practice.

    Use the known injector timing resolution of 1µs, and consider 1µs to also be the resolution of the crank velocity timer. With microsecond resolution at the crank, at 7700rpm the crank rotates .0462° in 1µs. So that's pretty good.
    <BR><BR>
    #20