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

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

  1. LAFS

    LAFS Long timer

    Joined:
    Dec 15, 2016
    Oddometer:
    1,697
    I lied.

    I get what you are saying you take it out of the loop.

    However your problem shows up after a period of time from what you explain.

    And as I thought, and Rodger has confirmed it can not wet a O2 sensor enough to kill it.

    So as it stands you have a good O2 sensor after applying heat to it and burning some substance off of it. You have another that did not wake up after the same heat treatment and is dead.

    What happened to the original/factory O2?

    It just fascinates me that 1 O2 played dead so to speak, and another is. I do not think I ever had a O2 sensor on anything go out. I am sure fascinate is not the word you would use.

    I can tell you on my wide band there is a fresh air calibration where I remove the O2 from the bung and hang them in a fresh clean air stream, hold some buttons down on the PC V, and it calibrates them.

    If it is not gas, then it has to be oil, if contamination is the real problem?

    Again I am rooting for you.
  2. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Hi Lee, you and I are on the same page now. I am very skeptical of an O2 sensor that "heals". I also know that it damn sure wasn't working when I took it out. Outside of electrolytic caps, I don't know of any electronic components that recover from failure. AND I always replace electrolytic caps even if they do recover. I take a dim view of things that fix themselves. I realize it's quite possible I burned off some contamination when bench testing but it wasn't really caked. It was just black.

    The original factory O2 sensor is still around. It does work but it's wiring harness is now too short to use. I bench tested it last night and got a good response. As I said, I changed it out as a precaution, it didn't fail. When I installed the AX-IED I wanted a known good sensor so I wouldn't have confounding problems form a slow old sensor. Sort of trouble shooting in advance. I didn't know then how to bench test an O2 sensor and I don't think it was too stupid a precaution. I got the bike in Sept and have found lots of signs of deferred and poorly done maintenance. The rear suspension was totally and dangerously mal-adjusted. Clutch cable broke from lack of basic lube at the clutch lever. There was a small oil seep due to improperly torqued bolts on the front engine/alternator mount. Bad throttle position sensor adjustment. All fixed now. I don't like the idea of bench testing O2 sensors anyway because it's so hard to control the heat and damage seem possible.

    O2 sensor #1 is now in the bike. At first I left the AX-IED unplugged and jumpered it's harness so that everything was as stock. It took kind of a long warm up (28° F for a high here today, but it did start responding with a voltage that dithered up and down from ~420mV. Then I plugged in the AX-IED on setting 4 and watched the voltages on the AX-IED and the O2 output do just what they should. Then I set the AX-IED to 7 and everything still worked as advertised. I can't imagine a setting of 7 causing any problems.

    The 1/2 mi of dirt road between me and pavement is too snowy for me to test ride today. It's gonna be that way for a while so I won't be able to do any testing for several days.

    One thing I've wondered about:

    When these O2 sensors first started failing they put out a few hundred mV right after start and then slowly the voltage would count down to ~15mV over a period of a couple/few minutes. Eventually, they wouldn't put out any voltage.

    A (perhaps) coincidence is that I did a throttle balance around the time of failure for both these sensors. The first one was adjusting the right side throttle stop after getting the TPS properly adjusted (thanks Rodger) and the second one was after the new injectors went in. I'm careful about how long I let the bike idle because of heat, but I wonder of longer idle times were a contributing factor or just another red herring.

    I'll post again after getting some riding in. Perhaps I will be allowed to exit the twilight zone.

    Thanks to both of you for taking an interest and the time to post some suggestions. It's much appreciated.

    Spence
    LAFS likes this.
  3. LAFS

    LAFS Long timer

    Joined:
    Dec 15, 2016
    Oddometer:
    1,697
    I hate odd stuff that you cant proof out, good or bad. That self heal, is a self fail waiting again.

    My thing is that exhaust gas is very hot running over that O2 sensor. It should be enough to have the sensor white or just slightly tan. A idle would just make that exhaust gas hotter

    Where are your plugs in all this? New? Inspected?

    What is the quality of this current tank of gas? Can you dump it and use it in a car and put "new" in? Maybe just a bad tank, it happens.

    If not the injectors, or the O2, or the AF-XIED, could you have a bad plug/coil that is fouling?

    I do know about flowed/matched injectors as I have had a set done so I understand it and would agree they should not be the issue.

    I mean suck, squeeze, bang, blow just how any motor works.

    Just a bunch more stuff to do it now.

    Keep at it I hope you can nail it down.
  4. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Good stuff Lee. Here's some responses:

    New plugs. Inspected at about 500mi when I did a valve adjust. Clean, beautiful.

    Current tank of gas is Chevron supreme from a station I always use. It'd have to be pretty bad gas, I think, to take out an O2 sensor, wouldn't it? But I'll consider dumping it in the car. Or better the mower, no O2 sensor on the mower.

    Brand new coil and plug wires. I replaced them just before this happened. Same day.

    I'm in the process of taking care of all the things I can think of or find to make this a solid bike. It really runs wonderfully when I can keep the O2 sensor working. No complaints.

    You're correct of course. Induction, fuel, ignition and exhaust - not too many places to look. I'm just gonna ride it and watch the O2 sensor and see where it ends up. I'm confident it will eventually reveal itself. I'm kind of a pit bull when I get involved in a problem. But it's up to the bike now to make the next move.
  5. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Here's a picture of sensor #2 removed today. It did have about 30mi or so on it after it stopped responding. Doesn't really look that bad does it? This is the one that failed it's bench test and is toast.

    IMG_4874.JPG
  6. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,430
    Location:
    Massachusetts
    You're doing all the right things Spence, I think you will get to the bottom of this but it could take a while. The O2 sensor looks ok to me ... maybe a little more carbon than I'd expect but I can't say with certainty. Maybe I can pull one of mine out of the R1200 and have a look at it, I'm running setting 8.

    If you set the AF-XIED to setting 1, the O2 sensor should clean up pretty quickly.

    The R1200 injectors do add more fuel due to faster turn on time, but that gets adapted away pretty quickly.
  7. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Well, yesterday and today were warmer and the snow melted. I got in some riding in the rain on both days. Only 30 to 40mi each day before I got soaked. The O2 sensor has worked like a champ with the AX-IED set on 7. I'll keep it on 7 for a good while even though I prefer 8.

    It would be amusing, perhaps, to see what the O2 sensor looks like but it's a pain on this bike to remove and install and I don't want to stress the sensor harness removing and re-installing. I've reached the mental state that while it's working I don't want to upset it.

    Thanks again, Rodger and Lee for all the ideas and especially the sanity checks.

    spence
  8. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Well, drat.... I rode my bike today to do some errands. 10mi of twisties to get to town and 3 stops in town. On the way home it was obvious to me that the bike wasn't fueling quite right. When I got home, I started it up with the AX-IED visible. Ten seconds after the yellow light blinked the setting, the green light started blinking. I got out my Fluke 87 and used the neg battery terminal for the ground ref.

    Measuring the voltage at the blue/wht wire (raw output of the O2 sensor) I got -600 to -700 mV at idle. Yes, negative. When I reved the engine slightly the voltage came up to about 35mV. I then unhooked the black wire from the O2 sensor from both the AX-IED and the motronic and measured a few mV at idle and no change with a rev.

    Now those are pretty goofy results. But they don't immediately speak to me. Since I don't have a schematic for the AX-IED, I don't know what pin of what op amp or micro that wire blue/wht wire is connected to and therefore whether it would be normal for that negative reading with an essentially zero output from the O2 sensor. I can think of lots of pins on lots of IC's that would do that. I'm guessing that the O2 sensor can't really source or sink much current when dead and therefore would probably read close to what ever it was connected to when it wasn't putting out any voltage.

    I watched the AX-IED for a minute on startup before my ride today and never saw the blinking green light following the yellow setting indicator. That blinking green light has always accompanied the O2 failure. I didn't get out the DMM then since there was no indication of a fault. So a 10 mi twisty ride and stop at the gas station and two other stops and on the way home it had all gone to schmidt.

    So.... problem with the AX-IED? I'm ruling out contamination because, though I can imagine a very unlikely scenario that *might* have caused contamination the first time I had a failure, there hasn't been anything but good gas through this motor today or anytime that I know of. The tank was inspected and cleaned in Nov. with hoses checked and fuel filter changed. The mechanic found the tank liner deteriorating and removed it. No coughing or sputtering that would accompany water in the fuel. No rain today so no water on the bike.

    It's hard to believe I have an AX-IED failure because when it all went back together this last time I checked the output of the AX-IED and found it varying at idle between 200mV and 800mV while the output of the O2 sensor was rock steady in the 800mV range. And now with no output from the O2 sensor the AX-IED blue wire measures a continuous ~450mV, which I believe is also what it should do. So the AX-IED does what it's advertised to do but my O2 sensor dies. If a diode or a transistor on that input blu/wht input was damaged in a way that resulted in very low input impedance I can see it discouraging the O2 sensor over time as the O2 sensor worked hard to come up with the current to drive the low impedance. That is one way the AX-IED could be causing the O2 to die without being itself totally hosed. Roger, do you know the output impedance of an O2 sensor? I don't know, it's all guesswork from here.

    It's not good enough to get a new O2 sensor and hook it up directly to the motronic and forget the AX-IED. Once you go rich, you can't go back. I'm starting to think of getting an LC-2. I don't mind the increased complexity of installation. Rodger, will the LC-2 log data without carrying a laptop on the bike while riding? There is surprisingly little documentation on the website about data logging. And data logging is an area I haven't explored.

    I will say the bike gets abominable gas mileage when O2 sensor dies. My fill up today showed ~35mi/gal during the time the fueling has been on the fritz.

    Tomorrow I'll remove the 02 sensor and take some pics and see if it will wake up again with a bench test.
  9. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,430
    Location:
    Massachusetts
    Spence, The -600 mV reading is "impossible". I'm not saying you didn't measure it but the AF-XIED has no mechanism to bring the voltage below "ground".

    Since you've had that implausible reading, before doing anything else, you need to get to the bottom of it. The most likely reason for a negative reading is a grounding problem on your motorcycle.

    The way to figure that out is to establish what is "ground". For the Motronic, and all other electronics on your bike, ground is defined as the case of of your engine. However, for convenience, the AF-XIED is grounded at the battery negative terminal because that is normally within 10 mV of "ground" and it's easy to attach to.

    What could create a negative reading is for a high resistance connection between the engine case and the battery negative terminal.

    To begin with, when you measured the white/blue wire, where did you attach the negative lead of your DVM?
  10. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Thanks Roger,

    I completely agree. I've been spending a lot of time with circuits with both pos and neg rails and it didn't click that there wasn't a neg rail here.

    As I said, I used the negative battery terminal as my ground ref. (Buried in my loooong post). I'll check the impedance between engine and negative battery terminal. I have a bench meter that can make a 4 wire (kelvin) resistance measurement. I'll also look at the connections.

    I can also measure the blu/wht wire again with the AX-IED using the engine as ground.

    Edit: I don't have time, this morning, for a full exploration but I did make a measurement with the DMM in "relative" mode between the negative battery terminal and the screw that connects the ground wire at the left throttle body - 0.0 ohms. I also measured between the negative battery terminal and grey wire going into the stock O2 connector with nothing but the DMM and the motronic connected to the wire - 0.2 ohms. Considering that this wire is not copper and has to go through the connector I don't find that measurement too alarming. This was also done with the meter probe just pressed against the wire, not as good a connection as with a clip sometimes. A 4 wire measurement will be more definitive. (Why does the left throttle body have a ground wire and not the right throttle body?)

    Even if this 0.2 ohm reading is perfect, it would require a current of 3 amps to drop 600mV across 0.2 ohms.

    Edit #2: should I start a new thread so as not to hijack this one, or is it relevant enough?

    spence
  11. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,430
    Location:
    Massachusetts
    4-wire kelvin, that’s pretty impressive. The resistance seems fine, I guess you’ll have to look further for the -600mV.

    Probably a new thread makes more sense. It seems very likely you have something bike specific going on.
  12. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    From memory and verified this morning, a failed O2 sensor can put out a negative voltage on a bench test. I saw -150mV on my failed O2 sensor this morning and up to about -800mV in earlier attempts.
  13. Gimble88

    Gimble88 Hooligan

    Joined:
    Dec 1, 2018
    Oddometer:
    29
    Location:
    Oregon
    Well, I like an entertaining problem as well as the next guy, but... When I got back to the problem today, I started the bike with everything hooked up and it all worked flawlessly at idle. The voltages on the blue/wht wire were correct and once warm the blue wire showed the 200mV-800mV variation, this was at setting 7. I changed to setting 2 and the AX-IED when rapidly closing the throttle I got the yellow light showing very lean. I even got surging when I tried to hold 3000rpm. Back to setting 7 and surging gone and an indication of richer operation from the AX-IED.

    I don't have much experience with Posi-Tite connectors. I always solder or crimp. I am very careful to make sure there is contact when I put Posi-Tite connectors together and test them to be sure the wire is captured. I can't believe I can start a ride with good connections and loose the operation of the O2 senor/AX-IED during a short ride. But operator error is looming as the cause here, unless some ground connection is randomly intermittent.

    Thanks for all the help,

    I'm going for a ride....
  14. bmw_rider

    bmw_rider Been here awhile

    Joined:
    Mar 6, 2016
    Oddometer:
    110
    Location:
    Australia
    Hello roger 04 rt

    I put this in a post on another forum about the AF-XIED and i think you wanted it posted here so you could respond to it.

    As there are 2 controlling units , with two O2 adjusting pots, my thoughts are that getting the LHS & RHS cylinders to be running the same AFR will be near impossible due to the tolerances within the AF-XIED unit components and manually setting the dials (by eye). How much of a difference and whether this will in application make a difference i have no idea. Two cylinders both running the same AFR, working in synergy, must be better than having a 'lazy' cylinder.
    I have seen an article where when tuning a 1200 with a Power Commander and the oxygen sensors disconnected, there was a marked difference in the stock mixture between the left & right cylinders. As they adjusted the fueling so the AFR was the same for both, the engine got smoother & smoother and ran slightly cooler.

    My thoughts are that once the AF-XIED unis have settled down with some use and the fuel trims have stabilized, it would be worthwhile getting the bike dyno'ed and the trim pots adjusted so the 2 cylinders are running the same mix, or reading the same O2 output. I presume this would best be done by doing some closed-loop runs at constant revs (3k, 4K & 5K rpm) in the region most likely to be riding at steady throttle.

    I would appreciate comment from roger & others as to whether this is not an issue OR worth worrying about, and i am being a bit OCD here. :hmmmmm

    And roger might be able to advise whether balancing the AFR's can be done thru monitoring with a GS-911 tool.
  15. leafman60

    leafman60 Long timer

    Joined:
    Aug 14, 2005
    Oddometer:
    3,801
    I probably have over 2000 miles on my AFX (at setting 6) with my '16 R1200GS.

    Everyday riding, with an occasional high speed jaunt is averaging 40-41 MPG.

    T'aint bad. It's about like stock. The AFX helps on low end, off-throttle take-offs and clutch work off-road.

    .
  16. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,430
    Location:
    Massachusetts
    Thanks. I will reply in the next day or two.
  17. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,430
    Location:
    Massachusetts
    These are very good questions that I've addressed at other times, and issues that I've wondered about too. Along the way, I learned that the R1200s, including the LC bikes, have a powerful set of Short Term and Long Term trims that not only align each cylinder to the AFR target of the O2 sensor (including AFR shift by an AFXIED) but also are compared to each other. If one O2 is disconnected, that cylinder's AFR is regulated by the other working/available O2, with surprisingly good results (you can find this earlier in this thread).

    Regarding your questions:

    --Can two separate AF-XIEDs achieve sufficient balance due to tolerance differences. The answer to this seems to be yes. Technically, the tolerance chain includes the O2 sensor, the temperature of each O2 sensor, the BMSK itself, and the AF-XIED. A rider with the handle Mistacat on the UKGSER.com forum installed AF-XIEDs and LC-1s so that he could measure AFR, from memory, he didn't mention any L/R AFR differences, which means the tolerances of the pieces must be pretty good. A deeper reason for not worrying about AFR differences between the two sides is that the richer you make the mixture, the less sensitive power output is to AFR. At what's called the best power mixture (13.0-13.2), all the oxygen is consumed and more fuel added doesn't add power. The slope of the power output versus fuel flow is very shallow at that point, meaning that a little more or a little less fuel doesn't matter. At around 13.8:1, where many operate their AF-XIEDs, the slope is a little less shallow but still not too sensitive to fuel flow differences. By enriching the mixture away from 14.7:1 (stock), you make the bike's two cylinders less sensitive to AFR imbalance. Here's a good article to read: http://www.gami.com/articles/bttf.php

    --The worst way to try and measure final AFR would be on a Dyno. The reason is that they measure AFR after the catalytic converter and near the tail pipe, two bad conditions. The cat alters the AFR and at low rpms, measuring near the tailpipe can easily have air reversion errors. The best way to measure AFR is to install second fittings adjacent to the stock O2 sensor locations and use Innovate Motorsports Wideband equipment to measure AFR. Several of us have done that and get excellent readings that are very different from what is seen on Dynos. With an Innovate Wideband set up you can make riding logs of AFR. If you read the earlier parts of this thread you will find dozens of examples using this method on R1150s and R1200s

    --There is no GS-911 data which gives any indication of actual AFR, but the Lambda Control Factor Data and the Multiplicative and Additive Trim data can show the BMSX's convergence and trim development process, which is interesting but doesn't tell you AFR, it only shows how much trimming has had to be done to reach target AFR.

    The bottom line is that the empirical and measured results suggest that the AFRs on an R1200 with AF-XIED get close enough that any differences aren't an issue.
  18. bmw_rider

    bmw_rider Been here awhile

    Joined:
    Mar 6, 2016
    Oddometer:
    110
    Location:
    Australia
    WOW thanks Roger. They were simply the wondering's of an inquiring mind. Well answered by the way.

    Interestingly, the 1250 now has sensor downstream of the cat and apparently the two sensors per pipe side compare to check the cat is working OK.

    FWIW, the 1250 fueling imo is better than my 2017 1200GS (which was pretty good anyway imo) especially at slower speeds being smoother, less surge at small constant throttle settings & nice off-on-off throttle control.
    But, like with my 1200, the 1250 seems to be adapting up it's own orifice and the overall tune of the engine goes off after a while. Small, constant & gradual throttle action i am sure is contributing to the issue as there is so much closed-loop running with low fuel demands. As well, after about 2,500kms the bike began the not start first time. The sound and reaction of the failed start is not nice to behold. Second try & starts immediately.
    I did (supposedly) a reset by disconnecting the battery and the tune seemed to come back (like when new from memory) and the not-starting 1st time went away.
    Some 1500kms later the not starting issue has returned but the tune still seems OK or is acceptable for how i am starting to ride the bike more now (slower & cruisey).
    Bike computer is showing 4.2 L/100kms over 5,000 kms and it is ridden hard & fast occasionally. About 0.3 L/100 kms less than the 1200 over 42,000 kms.

    My BMW service manager wants a video of a non-start so he can send to BMW. I want to do another reset and see if the non-start 1st time goes away. If it does go away it is probable that the issue is a result of the ECO/O2 adaption process.

    I don't really need an AF-XIED to correct fueling/running issues pre-adaption of long-term fuel trims.
    I am interested in an AF-XIED in the hope it will manage the long-term fuel trims in the longer-term and prevent or reduce or delay the effects & impacts that i believe the 'adaption' process brings about.
    Will it negate the need for periodic ECU fuel trim value resets?

    cheers & thanks again
  19. roger 04 rt

    roger 04 rt Long timer

    Joined:
    Nov 9, 2011
    Oddometer:
    3,430
    Location:
    Massachusetts
    The concept of lambda-shifting negates the need for ECU resets because it tunes the fuel to adaptation targets that it reaches after the adaptation period. In other words, it uses adaptation to reach its AFR goals.

    That said, there have not been any tests on the R1250 and in order to know what error codes the 1250 might or might not throw because of the downstream O2 sensor, the AF-XIED would have to be tested on that bike series. A gs-911 is needed for the testing, and I haven't heard whether or not the GS-911 is working yet on the 1250.
  20. bmw_rider

    bmw_rider Been here awhile

    Joined:
    Mar 6, 2016
    Oddometer:
    110
    Location:
    Australia
    Thanks Roger.
    Where i thought i may have found the answer to the adaption issue with the AF-XiED, I had not considered with the 1250 the sensor downstream of the CAT. This will confuse the heck out of the ECU.
    Dare say it will not work.
    I hope you and the other guys involved in developing the AF-XiED can sort this out asap. Likely another component to be added to the AF-XiED increasing its cost and complication.
    And just when you thought you had sorted the AF-XiED after the BMW issue of revising their software etc to cancel spoofer effects.


    More thoughts.

    I added a Dobeck fuel controller to my former Honda VFR800. I had to use O2 eliminators so that the changes made by the controller were not negated. The O2 eliminator has a resister which i understand sent a return signal value to the ECU so the ECU thought the O2 sensors were there & working and did not throw an error code.
    Not sure if the O2 eliminator also sent a 'nominal' O2 sensor AFR value to the ECU or in effect nothing and fuelling was only based solely on AIT, throttle position etc

    While the AF-XiED allows the perceived value seen by the ECU to be adjusted, curious why something like the simple O2 bypass plug used with the Dobeck, Power Commander etc can not be used with the BMW ECU to send a fixed Lambda value ? (of your choosing) to the ECU and let the ECU know the O2 sensor is there & working and no error code.
    Or eliminate the O2 sensors with no error codes and revert back to base mapping with no O2 input.
    This works with Dobecks & Power Commanders but maybe the BMW ECU can outsmart this too.


    I have read much of your thoughts on AIT spoofers and their output being sensed by the BMW ECU and adapted out over time, negating the effectiveness of the AIT spoofer. I see your logic in this.
    An ECU reset should fix this and make the AIT spoofer effective again to a reducing value, till adapted out again.
    Repeat reset.
    The fuelling on my 1200 and 1250 are good straight after an ECU reset so this is something i can and will do periodically. I am contemplating installing a +ve or -ve in-line switch under the seat so i can simply chop the battery power to the ECU for a while and do a reset. Not sure if this type of reset actually works and to what extent it resets but i do notice a difference after disconnecting the -ve lead.
    For example, with the 1250 the second appearance of not starting 1st time from a cold start happened 5 consecutive starts. I disconnected the -ve battery lead and next cold start, engine fired up 1st attempt immediately. Bike ran smoother and stronger too. BMW need to sort this issue out.


    In respect to the ECU sensing the AIT spoofer and its output, one point i have never seen mentioned is the fact the bikes have 2 air temp sensors, external ambient air at the front of the bike and the AIT itself measuring intake air temp inside the air box. The air temp within the air box would have to be higher (hotter) as the air box itself has a lot of heat soak from the engine etc
    IF the ECU is monitoring both the ambient external air temp and the air temp within the air box, with the AIT spoofers lowering the AIT value the ECU sees by 20 degrees, the ECU would likely see the engine intake air temp lower than the outside air temp and assume there is a sensor error and start 'adapting' to correct this variation.

    If i am correct in assuming the intake air within the air box is hotter than outside ambient air temp, what about relocating the internal air box AIT sensor to the front of the bike where it will 'see' cooler air naturally than it would register within the air box (same as the external air temp sensor so no gross discrepancy) and enrich the fuel mix somewhat.
    But then again, i suppose the O2 sensors would sense the richer mix and adapt it back to 14.7:1

    The AIT spoofer i tried on a F800GT and the 2017 1200GS connected to the temp sensor inside the air box. I noticed no improvement on either bike. Maybe a bit too rich. Never left it on long enough to see if any ECU adaption change.