2004 R1150RT Wideband O2 Sensor Project (and AF-XIED for BMW)

Discussion in 'Vendors' started by roger 04 rt, Dec 13, 2011.

  1. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    Terry is in the midst of a several hundred mile trip into the mountains on his 2011 R1200GSA with dual LC-1s and has sent me hours of very impressive data that I've been going through. I'll add some of it to the thread later today or tomorrow. A few things stand out:

    1) The BMS-K spends much more time in closed loop operation than the Motronic MA 2.4. After acceleration or deceleration it returns to closed loop rapidly.

    2) After the Adaptation Values are reset there are very obvious imbalances between the AFR of the Left and Right cylinders. It seems to take quite a bit of time and driving for the BMS-K's full adaptation matrix to be populated. During that time the L/R imbalances persist.

    3) After hours and a hundred miles or more of riding, the L/R cylinders seem nearly identically matched. The BMS-K has a very potent ability to balance the two cylinders through it Adaptation Strategy.

    4) The quality and consistency if the data is high and the BMS-K is not issuing any error codes. The BMS-K is fully functioning and fueling the motorcycle at the programmed lambda of 0.93 (as Terry has programmed it).

    Overall I'm very impressed by effectiveness of dual LC-1s on the R1200 and also the power of the BMS-K to manage fueling and to balance the L/R cylinders.

    There are some implications: 1) if you change the intake or exhaust, the BMS-K will fuel to the set-point of the O2 sensor; eventually adapt to it fully. There could still be different exhaust resonances for aftermarket systems which would result in richness or leanness at specific engine RPMs and loads. 2) Simple attempts to alter fueling will be negated by the BMS-K over time. However, reprogramming the value of lambda through the O2 sensor are not negated. In fact the BMS-K is fully engaged in the process of shifting fueling. 3) The BMS-K with adaptive fueling and adaptive spark is so powerful that Closed Loop should not be disabled without a lot of consideration. Disconnecting the O2 sensor, as with a Power Commander, will leave the motorcycle without its most important cylinder balancing tool.

    RB
  2. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    Here is Terry's 2011 R1200GSA bike with dual LC-1s and Notebook PC connected for datalogging.
    [​IMG]


    Next, a plot of about 80 minutes of data logged from his LC-1s. The important thing to see is how closely the bike holds to his 13.65:1 lambda target. The spikes on the chart are his accelerations and decelerations. There are actually two plots, one on top of the other for the left and right cylinders. The LC-1s and BMS-K have worked perfectly together to pinpoint Terry's fueling at 13.65:1. The small inset boxes show the data distribution of the fueling pulse AFRs. These are great, statistically normal, AFR distributions.

    [​IMG]

    The next plot shows, in very fine detail, just how well the two LC-1s and BMS-K work together to deliver highly syncronized fueling. The LC-1s provide the fueling shift from 14.7 to 13.65 (lambda 0.93) and the BMS-K has used it's strong adaptive capability to nearly perfectly match the cylinder AFRs.
    [​IMG]
  3. EKinOR

    EKinOR Been here awhile

    Joined:
    Mar 25, 2012
    Oddometer:
    274
    Location:
    Central Oregon
    Any updates on the plug and play LC1 (almost) equivalent? I'm either going to buy an LC1 or your new plug and play device, but I'd rather buy something plug and play.

    Any updates are appreciated!
  4. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    Back a couple of pages ago I posted a short update, here. The Proto 2 photo is below, a product for all R1150s, R1100S and K-bikes as well--any BMW that is fueled by a Motronic MA 2.4.

    The key pieces are the two BMW OEM connectors for which we now have a supply, you can see them in the photo. The connectors are what make it plug 'n play. Our manufacturing subcontractor has begun production on the first 25 units. Then after we've built the first couple dozen, we will gear production for a higher volume. Sometime in the next month I'll launch a web site that will allow the product to be delivered, the web site will be simple at first.

    Because I also have the LC-1 installed that the same time I can compare Proto 2 to the gold-standard LC-1. For the past week, I've continued to ride Proto 2 which works very well. This weekend I'll be removing Proto 2 from my bike and sending it to a beta-rider.

    I've also built an R1100, Motronic MA 2.2 proto which I'll post later this morning.

    So progress is good, I'm evaluating thoroughly so that when we turn on production buyers will be able to expect prompt delivery.

    [​IMG]
  5. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    Although I'm posting this in the Oilhead forum, I am doing so because it further shows the benefits of installing something like an LC-1 on a another model of motorcycle. In this case it was a friend's 2007 F800S. We began discussing the project about a month ago and it is now complete. Most of the installation took place over about a week.

    The F800S has a BMS-K the same as Terry's installation last month on the R1200GSA. The same software settings worked but rather than a single load resistor to satisfy the BMS-K that an O2 sensor was installed, seven 1/4 watt 1000 ohm resistors were installed in parallel which you can see in the box below(he's a very neat craftsman) . His reasoning was that this would save space and slightly warm the box insides to keep it dry.

    There are also two relays added inside the box. The first was a relay to switch the power, rather than use the O2 sensor Heater+. The second was a relay controlled by a panel switch to allow selection between the LC-1's two analog outputs. He has one setting at lambda 1.00 (14.7:1, stock) and a second setting at lambda 0.94 (13.8:1, 6% richer). This was a clever idea and he can switch mixture on the fly. (He comments on below.)

    [​IMG]

    Hi Roger,

    I managed to finish the installation work on the bike - only a faulty USB cable and a missing driver for the USB to Serial converter took some time to figure out.

    And the first road tests (100 km today) have been really promising.

    First, I'm happy and proud that everything worked in the first attempt.
    --Function is easy to check with out lifting the seat to see the status led - what you do is to ride the bike at a steady speed, fix the throttle, read the current fuel consumption in the extended dashboard display, and flick the switch on. Two seconds later, the fuel consumption has stabilized it self at a slightly higher level.

    --Your calceulations were spot on: 4,4 l/100 km became 4,7 l and 3,3 l/100 km became 3,5 l when the richer mode was switched on. I know the dashboard reading is not 100% accurate and there's only one digit after the comma, but both tests are indicating a 6% increase in fuel consumption - or something close to that on a spot measurement.

    --So I'm not seeing a reduction in fuel consumption like you do, but it may be related to the non-ethanol fuel we got over here? (Editor's Note: I believe I get about the same fuel mileage overall at 6% richer due to being in one higher gear much of the time.)

    The F800 was never surging badly at constant speed, but with the richer mixture, it's definitely running smoother and feels stronger in all closed loop conditions. Quite funny actually, because now you can really feel how hard the lean burning engine was [struggling] before.

    I would like to take the bike to a Dyno at some point to have a performance run on both settings to compare them. My butt feeling says that the engine is a couple of horses stronger all the way from 3500 RPM to redline, and the torque dip around 4000 RPM seems to be gone. But a dyno run will tell if my butt is right….

    So it seems that your recommendation of the LC-1 complimenting my BoosterPlug installation very nicely—Open & Closed Loop. It's a real pleasure to ride the bike now with balanced richer fueling.

    This setup is a keeper and is staying on my bike, and I'm happy that I have the capability to chose between fuel saving mode for cruising and power/fun mode for backroads and mountain riding.

    I'm taking the bike to a rally tomorrow that will involve 70-80 km backroad riding, so I will get a good chance to test the LC-1 further.

    Please accept my compliments for a bright idea and thanks for sharing it with me.

    All the best.
  6. FlipJr

    FlipJr n00b

    Joined:
    Oct 8, 2012
    Oddometer:
    3
    Location:
    Seattle, WA
    I've been following your threads extensively regarding the wideband 02 sensor installation on your R1150RT. All the data and discussion you posted in the forum really got me interested so I went ahead and got the LC-1 kit for my '02 R1150GS.

    Installed it on Sunday, everything worked as planned. I put the 14.2 AFR tune you had posted on the installation guide in, adjusted the TB's for idle and balance again (idle was at like 1500 after the bike warmed up as it should). Today I took it out for an extended ride, about 250 miles.

    WHAT A DIFFERENCE.

    Acceleration across the board is better. Even starting out in first gear, the bike doesn't feel like it wants to stall below 1700 RPM. I can easily use one higher gear in all my driving, and there's no more pinging if I crack the throttle at low RPM.

    Cruising at 75 mph was so much more enjoyable as well. The buzzing in the hand grips was drastically reduced, and the throttle seems a lot more forgiving. Before the mod, the throttle was so sensitive it was like a balancing act between accelerating and decelerating, very hard to maintain a set speed (a lot of people call this surging, maybe, or lack of throttle finesse. It's hard to keep steady even with everything setup perfect ). After the mod, I feel that the throttle is more forgiving, much easier to control. Much more like my old R65 used to handle.

    I just wanted to say thanks for all your hard work and dedication to this project! It made the install a breeze.
  7. terryckdbf

    terryckdbf Bumbling BackRoad Riders™

    Joined:
    Jun 1, 2006
    Oddometer:
    4,595
    Location:
    Perry, Ga
    FlipJr

    Congratulations on the install and result. All of us are indebted to Roger for this.

    My bike is performing flawlessly with the changes, just as you describe.

    After you run 14.2 for a while see what you think of 13.8.

    Again, well done.

    Terry
  8. FlipJr

    FlipJr n00b

    Joined:
    Oct 8, 2012
    Oddometer:
    3
    Location:
    Seattle, WA
    Took about a 2 hour ride with the new tune of 13.8 AFR. Much of the same as before, even smoother than before at cruise. It's a new bike. I can romp on the throttle in third and get the front wheel off the ground.

    I data logged for the whole trip as well. AFR latches onto 13.8 and fluctuates between 13.7 and 14.0 in closed loop as expected.

    The log also informed me of a richer idle. After I performed the first tune of 14.25 AFR, and warmed it up, the idle was around 1400, so to compensate, I performed a TB balance with the Twinmax. Got the idle down to 1100, but the left hand BBAS is nearly fully closed, right side might be 1/2 to 3/4 turn open. No big deal though, did my run yesterday and today, performance was great, excellent low throttle response, great takeoff from lights blah blah blah. Anyway, turns out the AFR at idle goes to 11.0, and then after about 5 seconds, the motronic starts leaning it steady until it hits the 13.8 mark and then just fluctuates there between 13.7 and 14.0. Idle speed gradually goes up by maybe 50 RPM during this time. This process occurs over about 45 seconds total. If I blip the throttle, idle returns to 1100 and AFR goes back to 11.0 and starts marching back up again. Don't know if this is good or bad, but there is no operational issue with the bike running that rich of an idle. I feel that I have more control of the bike at lower RPM's now anyway, it definitely doesn't want to stall. No stumbling, no backfiring, who knows.

    I'll play with it more tomorrow. If someone has ideas of what to check out pm me. Another awesome day around Puget Sound!
  9. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    PM'd you.

    I think based on your idle AFR and idle speed reports, something else is going on with your bike that I'd like to help you get to the bottom of. As you can see the LC-1 is an excellent diagnostic tool. Here is how a normal warm start to idle looks (my setting of lambda at this time was 0.92, which is 13.5:1):

    [​IMG]
  10. FlipJr

    FlipJr n00b

    Joined:
    Oct 8, 2012
    Oddometer:
    3
    Location:
    Seattle, WA
    Got my idle issue sorted out. Completely unrelated to the installation of the LC-1. LH Idle Stop Screw Jamnut was loose and idle stop screw was adjusted incorrectly, too far in, making the butterfly valve more open that it should have been, explaining my high idle before. I knew everything was nominal before I started the LC-1 Modification with my GS-911. Checked the TPS this morning and saw that it was off! I never touched the TPS, paint was still in tact so I started looking elsewhere. That's when I found the Idle stop screw.

    Adjusted it back using the TPS adjustment tool on the GS-911 as a guide. TB balance performed after a 20 minute warm up, and I've got my idle back where it should be and I have way more adjustment left on the BBAS's. AFR still dips down to 12.2, but its better than 11.0 like before. Ride-ability still completely unaffected.
  11. cardoctor1

    cardoctor1 Been here awhile

    Joined:
    Nov 21, 2010
    Oddometer:
    638
    Location:
    horsham pa
    scribed
  12. terryckdbf

    terryckdbf Bumbling BackRoad Riders™

    Joined:
    Jun 1, 2006
    Oddometer:
    4,595
    Location:
    Perry, Ga
    I am borrowing Roger's thread to post the following:

    After following Roger’s wideband O2 project and researching many alternatives, I have completed an install of dual wideband controllers on my 2011 R 1200 GSA. I have logged many hours of data over a few thousand miles both in the mountains (HeleNBack 2013) and in the flatlands of Central Georgia. I feel I finally have enough information to comment as some have requested.<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:eek:ffice:eek:ffice" /><o:p></o:p>
    The install was completed a few days before HeleNBack. It went very well due to Roger’s previous work and tremendous help by e-mail as well as phone.<o:p></o:p>
    <o:p></o:p>
    We started with conservative settings, progressing in small increments based on data Roger was seeing and how I felt the bike was responding. Each day leading up to HeleNBack 2013 I would add 1% to 2% of fuel. I left for Helen with a setting of 0.93 Lambda (13.67/1 A/F R) in one switch position and 0.96 Lambda (14.1/1 A/F R) in the other. I logged data throughout the day, each night I would send Roger the data from the day to analyze and recommend changes. The bike performed very, very well.<o:p></o:p>
    <o:p></o:p>
    Would I do it again? Yes, without a doubt. Does it transform the Big Pig into an S 1000 RR? No. The Camhead 1200’s run pretty good out of the box, add some fuel and they are wonderful to ride. The throttle is smoother, the motor response to throttle change is more linear, both on accel and decel, less hesitation, very responsive to small inputs. Adding fuel results in a cooler exhaust, cooler valve face, longer lived motor. The biggest change is in the higher gears. I can stay there a little longer without lugging the motor. Or shift up a little sooner with pulling power available. Based on my research and some knowledge of motors I expected some of these changes, one I did not expect was the ease with which it starts. Cold, warm, or at temperature it fires right up where before it did not. Upon returning I continued the programming changes. I am now at 0.91 Lambda (13.38 A/F R). The roll on power in 6<SUP>th</SUP> gear from 60 to 95 is surprisingly much stronger. Although best power may be down around 12.8 +/- I am looking for best ride-ability, best throttle response, best feel for me, not racetrack performance. I am very pleased where we are now.<o:p></o:p>
    <o:p></o:p>
    Fuel mileage is still up in the air. I keep a record of every drop of fuel, I have information from Little Switzerland, HeleNBack 2012, and F.A.R.T. before the LC-1 install. Preliminary tabulations from HeleNBack 2013 indicate two tanks got less mileage than a year ago, 2 tanks got slightly more. Too many variables to reach an accurate conclusion. I will monitor it in the flatlands here and have a better idea but I gotta say it is a lot more fun to ride.<o:p></o:p>
    <o:p></o:p>
    Do I recommend it? It works through the entire RPM range, open loop and closed, plays very well with the BMSK ECU. For those wanting to log data, program precise control, and do not mind the wiring, it is ideal. My install and set up went extremely well because Roger knows what he is doing. The fact we spoke the same language with regard to the data and the motor helped the process proceed swiftly. Roger’s groundwork has resulted in bikes running very well quickly. Most settings have been tested, the 1150, 1200, and now an F 800 are pretty much good to go. If you want all of the benefits without the wiring involvement I would recommend the device Roger is working on. It will do everything the LC-1 will except log data, not really necessary now. I have owned a Techlusion, I have a lot of experience with the Power Commander, I have a Boosterplug device, the O2 sensor device is the correct way to go about it. Tell the ECU what you want and let it accomplish the task.

    Terry
  13. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    The plug and play kits will be for three different products:
    1) a device that gets plugged in series with a stock narrowband sensor. I've logged dozens of hours with it running in parallel with the LC-1 and it is being beta tested on R1100 and R1150 (R1200 and F800 will be added soon) as I write this. We'll have some info from them soon. All units will preserve all Closed Loop Adaptation capabilities of their respective ECUs. You can read about it here:

    -http://www.advrider.com/forums/showpost.php?p=21060343&postcount=168

    2) a dual-AFR fixed-point Wideband sensor and controller built into a cable. I've run it for tens of hours and it is off being beta tested on an R1150GS single-spark this week. Here is a photo link:

    -http://www.advrider.com/forums/showpost.php?p=21030089&postcount=159

    3) Innovate LC-1 and MTX-L. Kits to make them plug and play, including the detailed software settings that best match each motorcycle. These are in process. Here is the main thread of my research and the installation overview. Near the end of the thread you can see the report of a well-know Danish F800S rider:

    -F800S LC-1 Installation Report
    -LC-1 Installation Overview for
    -Wideband O2 Research Thread

    Here's a shot of the R1150 Narrowband shifting prototype. I believe we will be able to get to a beta unit for the F800S/GS in a couple months:

    [​IMG]
  14. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    Although I've written extensively about this in the Oilhead forum, I'm adding it here to showthe benefits of installing something like an LC-1 on a another model of motorcycle. In this case it was a friend's 2007 F800S. We began discussing the project about a month ago and it is now complete. Most of the installation took place over about a week.

    The F800S has a BMS-K the same as Terry's installation last month on his 2011 R1200GSA. The same software settings worked but rather than a single load resistor to satisfy the BMS-K that an O2 sensor was installed, seven 1/4 watt 1000 ohm resistors were installed in parallel which you can see in the box below(he's a very neat craftsman) . His reasoning was that this would save space and slightly warm the box insides to keep it dry.

    There are also two relays added inside the box. The first was a relay to switch the power, rather than use the O2 sensor Heater+. The second was a relay controlled by a panel switch to allow selection between the LC-1's two analog outputs. He has one setting at lambda 1.00 (14.7:1, stock) and a second setting at lambda 0.94 (13.8:1, 6% richer). This was a clever idea and he can switch mixture on the fly. (He comments on below.)

    [​IMG]

    Hi Roger,

    I managed to finish the installation work on the bike - only a faulty USB cable and a missing driver for the USB to Serial converter took some time to figure out.

    And the first road tests (100 km today) have been really promising.

    First, I'm happy and proud that everything worked in the first attempt.
    --Function is easy to check with out lifting the seat to see the status led - what you do is to ride the bike at a steady speed, fix the throttle, read the current fuel consumption in the extended dashboard display, and flick the switch on. Two seconds later, the fuel consumption has stabilized it self at a slightly higher level.

    --Your calceulations were spot on: 4,4 l/100 km became 4,7 l and 3,3 l/100 km became 3,5 l when the richer mode was switched on. I know the dashboard reading is not 100% accurate and there's only one digit after the comma, but both tests are indicating a 6% increase in fuel consumption - or something close to that on a spot measurement.

    --So I'm not seeing a reduction in fuel consumption like you do, but it may be related to the non-ethanol fuel we got over here? (Editor's Note: I believe I get about the same fuel mileage overall at 6% richer due to being in one higher gear much of the time.)

    The F800 was never surging badly at constant speed, but with the richer mixture, it's definitely running smoother and feels stronger in all closed loop conditions. Quite funny actually, because now you can really feel how hard the lean burning engine was [struggling] before.

    I would like to take the bike to a Dyno at some point to have a performance run on both settings to compare them. My butt feeling says that the engine is a couple of horses stronger all the way from 3500 RPM to redline, and the torque dip around 4000 RPM seems to be gone. But a dyno run will tell if my butt is right….

    So it seems that your recommendation of the LC-1 complimenting my BoosterPlug installation very nicely—Open & Closed Loop. It's a real pleasure to ride the bike now with balanced richer fueling.

    This setup is a keeper and is staying on my bike, and I'm happy that I have the capability to chose between fuel saving mode for cruising and power/fun mode for backroads and mountain riding.

    I'm taking the bike to a rally tomorrow that will involve 70-80 km backroad riding, so I will get a good chance to test the LC-1 further.

    Please accept my compliments for a bright idea and thanks for sharing it with me.

    All the best.
  15. jscottyk

    jscottyk Been here awhile

    Joined:
    Apr 14, 2007
    Oddometer:
    387
    Location:
    Austin, Texas
    Snipped from here.
    Roger, does running the motor in the Best Power mixture range have any impact on catalytic converter function? Or, would it be best to run these sort of mods on headers with no catalytic converter?
  16. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    You can leave the catalytic converter in place but it will not operate very efficiently below lambda of 0.96 or 0.97. Because the extra fuel remove all the remain oxygen, there is no O2 to recharge the catalyst and as a result there is little converter oxidation and pollution levels are higher (these types of mods are only for the racetrack and not street legal, check the Dynojet site). As a result the converter runs cooler than stock.

    Now if you run an engine that's misfiring you are sending fuel AND oxygen to the converter and it will glow cherry red and melt the guts of the converter. But that's not way any of the mods do.

    My recommendation is that riders inch down the lambda from 0.96 to 0.94 and use something in that range. Most of us play with 0.90 to 0.94 when we've got nothing better to do, but my R1150RT runs great at 0.96.
  17. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    We've been running some tests with the LC-1 and my GS-911 on Wally's bike. I've learned about several differences on the R1150GS single-spark Motronic. The important differences are that the ECU's microprocessor seems slower, it's data acquisition seems less frequent, and the GS locates its O2 sensor in a place that seems to have less air flow. As a result, at regular idle, the Closed Loop period is a VERY slow 4 to 5 seconds. One to two seconds would be the norm.

    Other than that, the LC-1 and single-spark Motronic MA 2.4 play very well together. Here is a beautifully adapted cold start sequence with some notations.

    RB

    [​IMG]
  18. jscottyk

    jscottyk Been here awhile

    Joined:
    Apr 14, 2007
    Oddometer:
    387
    Location:
    Austin, Texas
    Do Secondary Air Systems have any impact on these mods? It would se that they would considering how they operate.
  19. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    Before I could answer that question with certainty, I'd like to read a good description of the function on the F800GS. can you point me to one?
    RB
  20. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,456
    Location:
    Massachusetts
    This is the only F800 I've supported with an LC-1 so far but there have been many Oilheads. Here is a link to an R1200GSA installation: Terry's R1200GSA Report.

    Below is a mail from the F800S owner who just finished a track day.
    RB

    Hi Roger,
    ...

    I'm just back from a very good track day (we got there by the bridge BTW). I had the L-1 in rich mode all day, and it was a real joy to ride the bike. Throttle control in higher RPM (above 4000) was so much smoother and precise. Especially when you just closed the throttle briefly to adjust the sped slightly before a corner (at 5000-7000 RPM), the control of the bike was much better than before.

    The bike was pulling strong all the way to the redline, and I actually had a guy on a K1200S saying that he had a hard time following me in the acceleration out of corners (we later discovered that he was riding in 3rd gear where I was in second, but the F800 did well anyway.

    We was divided in 9 groups from the fastest riders to the slowest, and I was in the fastest group with a BMW K1200S, a Honda CBR1100XX, Suzuki GSXR 750, Yamaha Fazer 1000, Honda CBR1000F, Honda VFR750F. I was not able to catch up with the guy on the CBR1100XX - he was really fast, but I could follow everyone else in my group. Not too bad for a 85HP F800S with a few tricks done :-)


    It's been a long day, and I'm off to bed :-)