Found the reason for my intermittent crappy idle..

[Don_Eilenberger]

My bike has had an intermittently crappy idle problem almost since it was new. Sometimes idle is fine, other times it resembles a Harley (but side to side rocking rather then front to back..) Last night on the way home from work it was particularly annoying, and I got home at a reasonable hour, so I decided to take a look with my TwinMax and GS-911 to see if I could find any possible reason.

 

The TB sync - was as expected - dead nutz on above idle (about 1,800 RPM and up) - Twinmax at full sensitivity (wish they had a 10 turn ZERO pot..) - it never moved from dead center. (And I've never adjusted the TBS..)

 

At idle was a different story.. It would intermittently center itself, then drift off to one side (same side) about a division (which is really tiny on full sensitivity). This was after using the GS-911 to lock the steppers for sync'ing. When it moved off to one side - the idle quality got worse. At times it almost stalled (since the steppers were locked they couldn't manage the idle speed.)

 

No answers..

 

But.. then... I decided to look at real time values. I immediately noticed while one O2 sensor was showing a fluctuating voltage (as expected) - the other O2 sensor (cylinder-1) was sitting at a steady state. Luckily - the GS-911 allows for plotting the two sensors.

 

VOILA! Sensor #1 was sluggish and intermittent. While it would start responding at higher RPM's (over about 2K), at idle RPMs it alternated between sitting there dead as a door-nail.. and working. When it was sitting there dead, the idle quality went to poop. When it started responding, idle quality immediately improved.

 

I checked the TB sync at idle with the TwinMax - and when the O2 sensor stopped responding, balance wandered off center. As soon as it recovered - balance was restored. The engine runs fine above idle - which is when the sensor works consistantly. The GS-911 showed the O2 heater voltage being switched on/off - at first I thought only O2 sensor#1 was being switched, but watching it for a while, both are, so it's not a control thing.

 

So - question now - try to get this covered under my Contego extended warranty? Or just replace with a Bosch universal heated O2 sensor? (BMW one is silly price - around $275..) Or look for a good low miles used one? Or take this one out and eyeball it for any deposits?

 

 

[NewBlue]

Interesting about the HO2S heater cycling on and off. They used to use a PTC resistor for the heater, temp goes up resistance goes up, working like a thermostat. Things change.

 

For $hits and giggles, swap the sensors and see if it follows.

[notacop]

That is rather superlative trouble shooting. Are O2 sensors really that much...Gack. Try Beemer bonehead.They may have used for a good price.

New blue had a good recommendation though. Swapping would insure your finding were accurate.

[RoSPA_man]

Interesting about the HO2S heater cycling on and off. They used to use a PTC resistor for the heater, temp goes up resistance goes up, working like a thermostat. Things change.

 

For $hits and giggles, swap the sensors and see if it follows.

 

For what it's worth, I had my O2 sensors out recently whilst swapping header pipes and they were really clean after 52k miles- I was expecting to see them caked in carbon.

[Hall Vince]

My RT has the occasional rough idle, it was doing some of this yesterday. Comes and goes.

 

I still have a very small flat spot off idle, which seems worse some days than others. Quite annoying when I'm trying to get a smooth up/down shift with no clutch slip and I release the lever to find I need to wind on a little more throttle to stop me nodding. It's the intermitent bit I hate and dealer states nothings wrong.

 

I was beginning to think I was imagining it or I'm just cr@p at riding the bike!

But I wonder if I should get the sensors checked out.

\v/

[dirtrider]

Afternoon Don

 

Yes the BMS-K uses the o2 sensors as part of the independent idle control. The early 1200’s seems to be better as I (think) they also used cylinder head temperature in conjunction with o2 input. The later 1200’s went to a single cylinder head temperature sender and to me they seem to idle worse that the early 1200’s.

 

On your o2 problem. I don’t know if they are the same (they sure look the same) but if so try swapping the o2 sensors side to side to see if the problem switches sides. You might very easily have a bad o2 sensor but maybe a possible dripping fuel injector or something similar and the o2 is just reporting the combustion oxygen that is sees.

 

With you o2 operating at above idle but not on extended idle it sure points to an o2 heater not working. Have you ohmed out the actual o2 heater on each side and compared? Maybe o2 heater circuit is working but the actual o2 has a bad internal heater, that wouldn’t show on just a functional voltage supply test.

 

(but) A dripping injector at idle would probably also be less troublesome as the RPMs come up.

 

On the universal o2 replacement. I have used some on the 1100/1150 with great results but your stating the o2 heater circuit is pulsed or pulse width modulated kind of puts a red flag up for me.

 

Some of the late model automobiles have gone this way and seem to use a 6 volt type heater at a pulsed 12 volts to get quicker warm up o2 heating. Before buying a universal o2 maybe ohm out your good side o2 sensor heater then you can make an educated guess if the heater on the universal o2 will work on your pulsed system. Narrow band O2’s are all pretty similar but the heaters can have different resistance. Also mind the o2 length that goes into the exhaust as the BMW pipes are small diameter, you don’t want the o2 to extend in too far.

 

 

Let us know what you finally find as this type of problem teaches all of us something.

 

 

[Nesbocaj]

Would unplugging the O2 sensors be a diagnostic aid?

[Don_Eilenberger]

Afternoon Don

 

Yes the BMS-K uses the o2 sensors as part of the independent idle control. The early 1200’s seems to be better as I (think) they also used cylinder head temperature in conjunction with o2 input. The later 1200’s went to a single cylinder head temperature sender and to me they seem to idle worse that the early 1200’s.

The '07 R12R appears to still have two.

On your o2 problem. I don’t know if they are the same (they sure look the same) but if so try swapping the o2 sensors side to side to see if the problem switches sides. You might very easily have a bad o2 sensor but maybe a possible dripping fuel injector or something similar and the o2 is just reporting the combustion oxygen that is sees.

Possibly, but the O2 voltage was low (around 80mV).. if the O2 sensor had reported a rich mixture then I'd suspect an injector maybe, but in this case it was reporting lean, which probably drives the BMS-K to try to richen the mixture, making the idle crappy. So far - it has all the symptoms of a sluggish sensor.. they are identical PN's so.. swappable if I get ambitious.

With you o2 operating at above idle but not on extended idle it sure points to an o2 heater not working. Have you ohmed out the actual o2 heater on each side and compared? Maybe o2 heater circuit is working but the actual o2 has a bad internal heater, that wouldn’t show on just a functional voltage supply test.

Possibly.. might do that.

(but) A dripping injector at idle would probably also be less troublesome as the RPMs come up.

But it also should drive the O2 sensor to a higher voltage (~ 1V or more), and that's exactly the opposite of what I'm seeing.

On the universal o2 replacement. I have used some on the 1100/1150 with great results but your stating the o2 heater circuit is pulsed or pulse width modulated kind of puts a red flag up for me.

 

Some of the late model automobiles have gone this way and seem to use a 6 volt type heater at a pulsed 12 volts to get quicker warm up o2 heating. Before buying a universal o2 maybe ohm out your good side o2 sensor heater then you can make an educated guess if the heater on the universal o2 will work on your pulsed system. Narrow band O2’s are all pretty similar but the heaters can have different resistance. Also mind the o2 length that goes into the exhaust as the BMW pipes are small diameter, you don’t want the o2 to extend in too far.

I was reading up on spec's for O2 sensors, and one discussion was on BMW sensors vs the rest of the world (natch!) They gave values of 6-12 Ohms for the heaters on most sensors, and 2 Ohms on BMW sensors since BMW want them to heat up quickly. So.. if the bike ones are designed like the car ones (and no reason to suspect not..) I'll be looking for a sensor spec'd for BMW use. Found some on Ebay for $25..

Let us know what you finally find as this type of problem teaches all of us something.

Will do. Will update as I do more troubleshooting. Problem is - it only annoys me at idle, and I try to idle as little as possible. Once off idle it seems to work fine - although I did get a feeling of a bit of surging at small throttle openings at speed - and a faulty/intermittent O2 sensor would explain that also. My response was to just keep it at larger throttle openings.

 

Maybe tomorrow - although the riding weather is supposed to be FABULOUS in NJ tomorrow, so playing with the bike may be restricted to actually riding it.

[Don_Eilenberger]

Would unplugging the O2 sensors be a diagnostic aid?

Not really. It would probably put the computer into a default map mode - which is typically a bit of a rich mixture.

[dirtrider]

Morning Don

 

Sounds like you have a good plan of attack and done enough homework to get it quickly figured out.

 

If/when you find a replacement o2 sensor that matches heater resistance and probe length please let us know what works on your 1200.

 

I haven’t had to put a replacement o2 in the hexhead yet but the hexheads are quickly coming out of warranty coverage so that day is quickly approaching a lot of us.

 

Thanks for keeping us informed on your findings.

 

[Nesbocaj]

Yes,I understand, I was thinking that if the bike idles smoothly without the lambda installed that this may indicate that the sensor is the problem, or maybe just the mapping. Maybe you could borrow a Techlusion.

 

I also ride a Husky. On them you pull the lambda, toss a 2.2K ohm resistor across the the two sensor wires and presto you now have a race map running on your ECU. Just so simple.

 

Looks like I suspect the map more than the sensor. Coming from an oh so smooth '04 1150 to an '08 1200 that runs lots less smooth, even after an exhaustive tuning, I have my bias reasons.

[dirtrider]

Morning Nesbocaj

 

Removing the o2 sensor as a test works on many bikes including the 1100/1150 BMW boxers.

 

Unfortunately the BMW 1200 heaxhead has two o2 sensors (one on each side) AND an electronic’s controlled idle speed controlled by separate idle air stepper motors one for each cylinder.

 

About anything you would do on the heaxhead like disconnect an o2 sensor could upset the side to side balance and cause a rough or uneven idle.

Once you drive one of those stepper motors to add or remove idle air all bets are off on how much more fuel the fueling computer will add to keep it running. In one respect the fueling computer tries to keep it idling but on the other hand it’s primary concern at idle in catalytic converter protection.

 

[Nesbocaj]

Howdy DirtRider,

 

Dang, the 1200 motor is new to me. More stuff to figure out on this puppy! If I understand things; the steppers respond through the ECU magic, taking into account all of the sensors involved and adjust the idle. Does that mean that any sensor in the program; air, temp, and O2 can give a crappy idle? So then diagnostically (wow, I spelled that right!) without a fault being present one must play the parts replacement game.

 

I think if it were me I'd still try the Techlusion or power commander if I could borrow one, or park the steppers with a GS911 to try to eliminate their involvement, as well as any upstream sensors.

 

Thanks for the explanation, making sense, slowly!

[dirtrider]

Afternoon Nesbocaj

 

Yes the fueling computer magic does control the idle speed, idle cross side balance, and idle fueling all at the same time.

 

While the engine sensors do have some cumulative effect on idle qualities things like the 02 sensors are the major players with things like air (AIT) having little effect.

 

Even adding a Techlusion, or Wunderlich fuel controller doesn’t do much for the hot base idle as they both default to the o2 sensor control for hot curb idle fueling for cat. converter protection.

Now the power commander is different as it has the stock o2 sensors removed and fuels open loop. Well unless the single wide band o2 is added along with the dynamic learning add on box then it sort of adjusts the system as you ride to the parameters you ask it to use.

 

I have installed a couple of the Techlusion 1334st fuel control boxes as well as I am currently playing with the Wunderlich unit (same as the 1334st only slightly different internal programming) on the 1200 hexheads and that does make quite a difference in the ridability of the hexhead, they (the fuel controllers) even seem to improve shifting as the RPM drop off and return at shifting is much smoother and linear. The 1200 with the 1334st box on it makes the 1200 hex run like you would think a $20,000.00+ motorcycle should. The down side is at the expense of fuel economy. A v-e-r-y thin line on the 1200 hex add on fuel controllers between running lots better and a significant drop off in fuel economy. At light throttle slow riding not so bad as a well adjusted fuel controller will allow a higher gear to be pulled at a lower engine RPM.

 

[Don_Eilenberger]

Well.. some good news I think. And it's proven again, Eilenberger's Law of Parts.. (Law#2 - http://www.eilenberger.net/laws.htm)

 

Went for a very nice ride today. Bike was running exceptionally well, except for the usual crappy idle. Last night, just before falling asleep I had an "ah-ha!" moment. Back when I first bought the R12R I was keeping an eye out on Ebay for any bargains in used parts.. one of the "spares" I picked up (hoping to invoke Eilenberger's Law#2 of Spare Parts) - was a wiring harness for around $50. It came to me last night that I sort of remembered it came with O2 sensors. (I hadn't even looked at it in years - it's been sitting in a box on a top shelf in the garage.)

 

So - got home, looked up at the box - it said right on it (in my handwriting - go figure) "R1200R wiring harness & O2 sensors".. Zowie. That $50 was starting to look like one of the better deals I bought on Ebay.

 

So - I got the box down, sure enough, tossed in it with the wiring harness were two O2 sensors with a tag on them that said "'07 R1200R". I went in and got my netbook and GS-911 and hooked things up. I decided to try to determine which cylinder BMW considered #1 and which was #2 by unplugging the O2 sensor on the left side cylinder. Turns out - when I started the engine (still warm from my ride) - cylinder #1 was doing it's thing (voltage varying from 0.2V-0.9V at about 1 second intervals) and #2 cylinder (with unplugged O2 sensor) was plotting a straight line at ~0.4V (which is what an O2 sensor is supposed to put out when the mixture is "perfect".) Soooooo... #1 is on the right, #2 is on the left (mebbe the right is more forward? Have to look..)

 

It appears that the BMS-K, when it sees no signal from an O2 sensor, defaults it to ~0.4V telling the system to use whatever mixture would be used at that simulated voltage (and the bike idled rather nicely actually.) At least that's my guess. From my experience the other night, it might be the system is smart enough to know when a sensor is unplugged (perhaps checking for heater resistance, or any signal from the sensor..)

 

So - just for giggles - since #2 (the left cylinder) was the one that was intermittent the other night, I plugged it back in. And it was working - just fine. I ran it another 10 minutes or so - and it kept running fine.

 

Eilenberger's Law just came into effect. Apparently having two used O2 sensors within inches of the recalcitrant one was enough to scare it back into action. The bike was idling just fine. And it kept idling just fine.

 

I'll be keeping an eye on it, now that I know what cylinder was acting up (left side) and since it looks like it might take all of 5 minutes to swap out a sensor. If it starts doing it's bad idle deal again, I'll swap in a replacement sensor.

 

It's possible the intermittent condition was cleared up by unplugging and plugging in the sensor. That sort of thing has cleared up lots of old K bike problems in low voltage circuits in the past. IF I had changed the sensor without plugging the old one back in - I would have been here reporting how great the new sensor is. I'm guessing the old sensor is actually fine, just had a not so good connection.

 

FWIW - there is a clip thingie on the sensor plug that holds it to the bottom plug plastic cover. It's very easy to break the little clip (DAMHIK, I just do..) Luckily, a narrow tie-wrap works just fine to hold the connector to the plug cover, so not a big deal. And you do have to release it from the cover to get the plug disconnected, the catch on the plug is on the inside facing the plug cover. So if you're thinking of doing a disconnect, reconnect take some care doing it.

 

I'll let'cha know if the crappy idle returns - and if so - if a new sensor fixes it. It's a case where parts swapping won't cost me a thing since I own the parts already.

[Lineareagle]

I recall helping a pro-football player friend of mine, great player but not the brightest penny, figure out a home wiring problem.

I came in for dinner and he asked me to come downstairs and look at his wiring.

 

Down we go, I look up and he had torn open the ceiling, tracing the wire which he was sure was bad.

 

"What did you do?"

 

" I came down the other day and the light was out, so I got a new bulb and screwed it in and it still didn't work so I started tracing the wires."

 

"This the new bulb?"

 

"Yes"

 

I take the new bulb over to a lamp that is working, unscrew the working bulb, screw in the new bulb. No workie.

 

Take the hot, ouch, ouch bulb over and screw it in, voila, the light she works!

 

All this to say, I would be careful with swapping used stuff, new stuff for that matter, in and out to check a problem, just sayin.

 

With your penchant for 'issues' . . . .

 

 

 

[Don_Eilenberger]

All this to say, I would be careful with swapping used stuff, new stuff for that matter, in and out to check a problem, just sayin.

 

With your penchant for 'issues' . . . .

Certainly always a risk.. but since I havent' swapped anything yet (just screwed the same bulb back in.. and it lit up - to use your analogy..) I don't think I'm out much. And with new sensors going for > $250, it's worth 5 minutes of my time to see if a used sensor DOES work if I run into the problem again.

[Bradley_Gillie]

Thank you for figuring this out. I've had the same problem on my 07 RT for a while. I wish I wasn't out of town right now, or I would run into my garage to check it out.

[NewBlue]

This is an interesting thread.

I was a master tech for three manufacturers, MB, Volvo, and Hyundai. If you follow the diagnostic manual for fault tracing it ends up noting that you should test with a known good component. Don followed the fault tracing steps and the diagnostics pointed to a faulty O2 sensor. The gift you have with this is you are given a test sensor, the other cylinder, swapping the sensors would have been the verification if the sensor is to blame. The point it's at now is the fault is still a mystery.

 

As far as idle speed, it is done with the steppers.

All they do is bypass air from one side of the throttle plate to the other the same as opening the throttle plate. The O2 sensors don't control idle speed but they can/do control idle fuel trim along with the adaptives which can have an effect on idle quality. Quantity and Quality are two different things.

 

[windyducat]

I hope you guys don't mind me jumping onto this thread with yet another possible O2 sensor problem. My R1200ST ran really rough between 1,200 and 3,000 rpms. Slight changes in throttle at slow speeds (1st and 2nd gear) had no effect. For example rolling off throttle the power stayed the same then with less throttle it drops like a hammer. Same with gradually rolling on throttle...more like a light switch, than a rheostat. I synced the throttle bodies, adjusted the valves, changed the plugs, and went through the tps relearn procedure and amazing, it was like a new bike. The slightest throttle inputs and power changed smoothly. Great, I thought. Rode it 200 miles at varying speeds and acceleration, slowed to go through a little town, and dang the problem was there again. Old school, I'd say a carb had a flat spot just off idle. Seat of the pants - feels like the computer is abruptly increasing/reducing fuel to the injectors.

 

Definitely fuel calibration/delivery is out of sync with throttle opening changes. Appears that the computer is learning "wrong" - possibly a bad input. One theory (my guess) is the computer is getting bad inputs from a O2 sensor and throwing off the fuel block map. If I understand it right, if the O2 sensors are disconnected the computer will run in open mode (not sure if that is the right term, my experience is with ford EFI).

 

Interesting, on the road I disconnected the battery and crossed the positive and negative leads to clear memory, and the problem mostly went away, but it was not near as good as when I got it back from the shop/tech.

 

Might be time to break down and buy a GS-911.

 

Thanks for reading and any insights are appreciated!

 

Windy

[dirtrider]

Hi Windy

 

You seem to have a lot going on there. I really don’t think a bad o2 would make your bike act like it is. The first place a bad o2 usually shows up is at idle or light steady state throttle. As far as a bad o2 skewing the block learn? Maybe but it usually doesn’t take long for the fueling computer to figure out the o2 isn’t following the fuel request and disregard it.

Don’t totally rule out an errant o2 but that wouldn’t be my first guess.

 

Your problem is acting somewhat like a stick coil going out or low fuel pressure at low RPM’s (like a partially failing FPC). Could even be a bad spot in the TPS that allows the TPS to have a hole in it’s output.

 

At the point you are at now it sounds like a GS-911 would be prudent investment, either that or a lot of testing and parts substitutions are in your future.

 

[windyducat]

 

Could even be a bad spot in the TPS that allows the TPS to have a hole in it’s output.

 

 

Dirtrider, thanks. The O2 came to mind while riding today, but my first thought was a bad spot on the TPS. What is puzzling is why would it run great right after the relearn (have done the relearn 3 times and each time the same results, runs good at first but then after about 200 miles the problem reemerges) and then perform so poorly. I ruled out the a bad spot on the TPS for that reason. Same with the stick coils. Wouldn't the problem exist even at the on start of the relearn?

 

Does the relearn start with the open mapping?

 

One other variable..on each occasion I was in WI/IA and was forced to put in E90 gas. Has than been know to throw off the fuel mapping?

 

Cheers

[Don_Eilenberger]

This is an interesting thread.

I was a master tech for three manufacturers, MB, Volvo, and Hyundai. If you follow the diagnostic manual for fault tracing it ends up noting that you should test with a known good component. Don followed the fault tracing steps and the diagnostics pointed to a faulty O2 sensor. The gift you have with this is you are given a test sensor, the other cylinder, swapping the sensors would have been the verification if the sensor is to blame. The point it's at now is the fault is still a mystery.

It's still a mystery - but if what I suspect the problem was is true - swapping them would have still resulted in the same mystery.

 

I rode the bike about 4 hours and 150 miles today. Idle was perfect the entire time. That's the first time in months (or perhaps a few years actually) that I saw perfect idle over that period of time. And all I did was unplug - and plug back in - the O2 sensor I saw faulting earlier. My suspicion is - a bad connection at the connector that was cleaned up enough by the physical unplugging/plugging-back-in to no longer be a problem. If I had swapped O2 sensors - chances are good the results would have been the same, and we'd still be looking at the same conclusion.

 

So.. while following diagnostics manuals can lead to a resolution - it also can lead to a false (and costly) conclusion. If I'd started with a new O2 sensor (known good as you suggest) and replaced the one I thought was bad, likely the result would again be the same. The problem would have been gone. I would have been out the $270 or so for the new sensor, and we would have falsely concluded that the sensor was at fault. I have seen this scenerio play out quite a few times with car repairs and bike repairs. A coincidental fix - caused by the action of replacing the part, even when the part is good. I do try to think beyond the manual, which is why I plugged the sensor back in after disconnecting it. I was simply curious if anything changed by doing that - and it apparently had.

 

I'll be monitoring it over the next weeks and let'cha know if it remains behaving well.

[NewBlue]

I hope you guys don't mind me jumping onto this thread with yet another possible O2 sensor problem. My R1200ST ran really rough between 1,200 and 3,000 rpms. Slight changes in throttle at slow speeds (1st and 2nd gear) had no effect. For example rolling off throttle the power stayed the same then with less throttle it drops like a hammer. Same with gradually rolling on throttle...more like a light switch, than a rheostat. I synced the throttle bodies, adjusted the valves, changed the plugs, and went through the tps relearn procedure and amazing, it was like a new bike. The slightest throttle inputs and power changed smoothly. Great, I thought. Rode it 200 miles at varying speeds and acceleration, slowed to go through a little town, and dang the problem was there again. Old school, I'd say a carb had a flat spot just off idle. Seat of the pants - feels like the computer is abruptly increasing/reducing fuel to the injectors.

 

Definitely fuel calibration/delivery is out of sync with throttle opening changes. Appears that the computer is learning "wrong" - possibly a bad input. One theory (my guess) is the computer is getting bad inputs from a O2 sensor and throwing off the fuel block map. If I understand it right, if the O2 sensors are disconnected the computer will run in open mode (not sure if that is the right term, my experience is with ford EFI).

 

Interesting, on the road I disconnected the battery and crossed the positive and negative leads to clear memory, and the problem mostly went away, but it was not near as good as when I got it back from the shop/tech.

 

Might be time to break down and buy a GS-911.

 

Thanks for reading and any insights are appreciated!

 

Windy

 

What you describe sounds like an issue with the fuel shut off on decell. It's just what came to mind when I read this. Why you notice it to the extent you do, I don't know, but the GS911 will let you monitor the data in real time and should help isolate it.

[NewBlue]

This is an interesting thread.

I was a master tech for three manufacturers, MB, Volvo, and Hyundai. If you follow the diagnostic manual for fault tracing it ends up noting that you should test with a known good component. Don followed the fault tracing steps and the diagnostics pointed to a faulty O2 sensor. The gift you have with this is you are given a test sensor, the other cylinder, swapping the sensors would have been the verification if the sensor is to blame. The point it's at now is the fault is still a mystery.

It's still a mystery - but if what I suspect the problem was is true - swapping them would have still resulted in the same mystery.

 

If I had swapped O2 sensors - chances are good the results would have been the same, and we'd still be looking at the same conclusion.

 

But, if you swapped the sensors right to left and the idle fault resurfaced, when you checked the HO2S with the GS911 seeing if the faulty O2 signal changed sides would have let you know if it was in fact the sensor or to look somewhere else. I didn't read back in the thread but keep the purge valve in mind if the condition happens again. If the valve sticks open it could be a rich mixture initially but then the O2 sensor would show a lean condition.

It's just a chess game.

 

The manufacturers are smart, they don't say to replace with a "new" component but to "test" with a known good one. At a dealership known good ones were at hand. If they stated, in the diagnostic procedure, as fact to order and replace with a new one and the fault continued they would have to eat it not us.

[dirtrider]

Morning Don

 

Thanks for keeping us in the loop. We always learn something from others misfortunes.

 

I have seen many sensors (not just o2 sensors) act like what you are seeing. It can be slight oxidation on the terminals or a terminal with not enough connection tension, or even moisture in the connector.

 

In your case the problem at first doesn’t seem to be on the sensor output side as your sensor seemed to work fine above idle. You would think any problems on the sensor output side would stay with you through the above-idle test.

 

Your original tests seemed to indicate the o2 heater might have been the problem as an inop o2 heater or heater problem usually shows up as a lazy o2 at idle then working OK at higher RPM’s.

 

In the auto industry one of the things we commonly look at is connector terminal friction or terminal connection integrity. For that you need a new and proper male terminal that you can smoothly slide into the female side terminals one at a time then check the terminal drag on insertion and removal. A faulty female terminal with improper friction will show up quickly.

 

Moisture (even a little) in an o2 connector is always a problem as there is potential for both current transfer or ground transfer to the sensor output side. That usually shows up as erratic engine operation at many operating ranges not just idle.

 

I suppose there are other potential problem areas like a poor connection inside the o2 sensor itself that was changed when the wires were messed with or a problem with the o2 sensor getting it’s outside ambient comparative air into the sensor.

Not real sure on the hexhead but with those short o2 wires and low o2 mounting I would take an educated guess that it comes down through one or more of the o2 wire pigtails (right down between the wire strands) . I have seen problems in past in that area if an o2 sensor can’t access outside air, usually because someone greases the inside of o2 sensor connector or folds the o2 sensor wire bundle and tie straps it tightly.

 

Hopefully your problems are solved but if it comes back please update this thread. “Inquisitive minds want to know”

 

[dirtrider]

Morning Windy

 

We don’t really know if it the change in operation was due to the actual TPS relearn or just the power down of the fueling computer. A GS-911 would show real time active TPS values as well as other real time sensor values.

 

Next time it acts up you might try JUST removing the battery cable for a short time without doing the TPS relearn. That might show you a direction.

 

If the operational change is in fact due to the TPS relearn then there is more than likely something wrong with the TPS voltage stream that causes the BMS-K to become skewed while riding.

 

You could also have something wrong with one of your other engine sensors therefore skewing your fuel trim.

 

Even a stick coil acting up can make the hexheads act like a fueling issue.

 

You are at the point where you probably need to get some sort of test equipment on your system or will need to start trying to test each component with a meter or start swapping parts with a known good bike.

 

You can try unplugging both o2 sensors then riding it as that might point a finger but even that won’t be definitive.

 

[NewBlue]

Don,

 

Could be this, could be that, might be this, might be that,

Swap the O2's and use it as a diganostic tool, easy and free.

Find out what it isn't to help find out what it is.

 

windy,

 

Could be the TPS, could be the O2 sensor, could be another sensor, could be a lot of things. Unplugging sensors can mask the real problem. Beg, borrow, or buy a GS911 and procede with caution from there (or have a dealer look at it).

 

 

[Don_Eilenberger]

This is an interesting thread.

It's still a mystery - but if what I suspect the problem was is true - swapping them would have still resulted in the same mystery.

If I had swapped O2 sensors - chances are good the results would have been the same, and we'd still be looking at the same conclusion.

But, if you swapped the sensors right to left and the idle fault resurfaced, when you checked the HO2S with the GS911 seeing if the faulty O2 signal changed sides would have let you know if it was in fact the sensor or to look somewhere else.

 

It appears so far that wouldn't have happened, since the act of disconnecting and reconnecting appears to have "fixed" the problem. If I had done that - the probability is both sides would have read fine, and that would have resulted in either no conclusion, or the same one I have right now - that it was a connector problem.

 

I didn't read back in the thread but keep the purge valve in mind if the condition happens again. If the valve sticks open it could be a rich mixture initially but then the O2 sensor would show a lean condition.

I should have mentioned, my charcoal canister and purge system fell off the bike a long time ago. Never got around to putting them back on (they got in the way of adjusting my front shock..) Certainly given the symptoms that would be a possibility if the system was still connected (since it feeds into the left side TB - which is the side showing the bad O2 sensor behavior..)

 

It's just a chess game.

Absolutely agree.. that's what makes if fun or very frustrating at times (depends on whose bike it is..)

 

The manufacturers are smart, they don't say to replace with a "new" component but to "test" with a known good one. At a dealership known good ones were at hand. If they stated, in the diagnostic procedure, as fact to order and replace with a new one and the fault continued they would have to eat it not us.

True again - my point there is - Occam's razor http://en.wikipedia.org/wiki/Occam's_razor - says "usually the simplest explaination is the best.." My experience is - sometimes it's not.

 

The point I was trying (inadequately perhaps) to make - sometimes problems are "fixed" by component substitution when the actual problem wasn't the component, but the connection/installation of the component. That can lead one to conclude (via Occam's Razor) that the problem WAS the component, when actually it was something entirely different. I think in this case - it was something entirely different (a bad connection.. more ideas on this below..)

 

Time will tell if the problem is solved, or just in remission..

[Don_Eilenberger]

Morning Don

 

Thanks for keeping us in the loop. We always learn something from others misfortunes.

It's actually been a really interesting thread since it's going into HOW to logically troubleshoot, something that isn't often discussed (but I - along with Brian Curry - did give a talk on this at a national rally some years ago..)

I have seen many sensors (not just o2 sensors) act like what you are seeing. It can be slight oxidation on the terminals or a terminal with not enough connection tension, or even moisture in the connector.

Yup. It's called the "Laying on of Hands" - Brian is famous for being at one rally where someones K1100something wouldn't start. The first thing he did was disconnect and reconnect the main Motronic connector. The bike then started and ran fine. The connectors were apparently slightly oxidated, and the simple act of disconnect/reconnect (the connectors are "wiping" types) cleared up the problem. Given the somewhat harsh environment the connectors on a bike have to exist in - it's not a surprise that they may at times not work quite as designed given time and weather exposure.

In your case the problem at first doesn’t seem to be on the sensor output side as your sensor seemed to work fine above idle. You would think any problems on the sensor output side would stay with you through the above-idle test.

I agree and this IS a bit puzzling, but I came to somewhat of the same conclusion you did below..

Your original tests seemed to indicate the o2 heater might have been the problem as an inop o2 heater or heater problem usually shows up as a lazy o2 at idle then working OK at higher RPM’s.

Yup. This is still a possibility - if the heater wasn't working due to a bad connection. We'll leave this as a "maybe".. It would match the symptoms, and since we couldn't look at current draw on the heater we have no conclusive data to prove or disprove it. The only things I might have noticed (have to say "might" because I think I noticed it, but didn't think of "why" when I did) - was the good O2 sensor was showing the heater voltage switching on/off on the GS-911, and I think the bad sensor - the heater voltage stayed on. Maybe. I think. That could indicate there is some feedback from the actual sensor to the voltage switching control, and one was receiving the expected signal and went into switching mode, and the other wasn't - staying in constant-on mode. Depends on what it uses for the "expected signal" - and we'll probably never know that..

In the auto industry one of the things we commonly look at is connector terminal friction or terminal connection integrity. For that you need a new and proper male terminal that you can smoothly slide into the female side terminals one at a time then check the terminal drag on insertion and removal. A faulty female terminal with improper friction will show up quickly.

 

Moisture (even a little) in an o2 connector is always a problem as there is potential for both current transfer or ground transfer to the sensor output side. That usually shows up as erratic engine operation at many operating ranges not just idle.

And despite the location of these connectors (on the R12R - they are very exposed..) there was no sign of dampness when I took them apart. Might have been wet in the past - again, a maybe, but unproveable. Symptom match wasn't quite as good.

I suppose there are other potential problem areas like a poor connection inside the o2 sensor itself that was changed when the wires were messed with or a problem with the o2 sensor getting it’s outside ambient comparative air into the sensor.

Not real sure on the hexhead but with those short o2 wires and low o2 mounting I would take an educated guess that it comes down through one or more of the o2 wire pigtails (right down between the wire strands) . I have seen problems in past in that area if an o2 sensor can’t access outside air, usually because someone greases the inside of o2 sensor connector or folds the o2 sensor wire bundle and tie straps it tightly.

And this could be the answer. I came to the same conclusion you did on the location of the comparative air vent/intake. The upright position of the sensor in what can be a very wet location would seem to preclude using ones that vented through the O2 sensor itself. It would make more sense to use one that vents through the wire. The connector is well gasketed, and appears to be water-tight, so that means it has to have a vent specifically made to feed air to the wire connection that's used for the comparative air intake. If that vent was partly plugged it could account for the behavior I saw. At idle speed there wasn't enough vacuum drawn on the wire to pull comparitive air into the sensor. At higher speeds, the higher exhaust velocity past the O2 sensor created a greater vacuum in it that did pull air into it. It would be interesting to know the failure mechanism of "sluggish" sensors..

 

The thought did cross my mind that "gee.. mebbe some dielectric grease would be a good idea" - then I realized it would be a VERY BAD idea at this connector IF it uses the wire-vent scheme. The grease would be drawn into the comparative air poisoning the active element. So - advice - don't use ANY grease on this connector. Clean and dry is the operative mode for it.

Hopefully your problems are solved but if it comes back please update this thread. “Inquisitive minds want to know”

Will do. It's an interesting brain puzzle, luckily not one that disables the bike. BTW - it's not a Bosch sensor unless Bosch has taken to manufacturing their sensors in Japan (the sensor is labeled "Japan" - no manufacturer name I could see..)

 

Best,

[NewBlue]

[quote=Don_Eilenberger It's an interesting brain puzzle, luckily not one that disables the bike. BTW - it's not a Bosch sensor unless Bosch has taken to manufacturing their sensors in Japan (the sensor is labeled "Japan" - no manufacturer name I could see..)

Best,

 

If the fault moved to the other cyliner it would indicate an O2 failure, if it stayed with the same cylinder then, yes, the wiring, connector, etc. This would only help if the fault reared it's ugly head, if not all bets are off.

I'm done beating this poor horse to death.

 

Bosch used a zirconium dioxide sensor, it's operating range was from .1 to .9 volts. It used a PTC heating element with an operating temp. above 290C.

 

Bendix fuel systems, known as Regina RexI with Volvo, used a titanium dioxide sensor made by NTK a divison of NGK.

It's operating voltage was 0 to 1.2 volts. It also used a PTC heating element with an operating temp of 700C

The bendix system was a MAP system

 

Physically the Bosch unit had a large body, almost the size of the nut, The Bendix had a much slimmer body.

[KTsRidin]

Okay, for all us "non-techy" folks who are just learning about our R1200's, could someone provide a location and "how-to" on accessing to do the unhook/hook method? Pictures would be especially appreciated. While I'm riding an '09 R1200RT, my guess is the difference has more to do with the Tupperware than the sensor locations.

 

THANKS

[Don_Eilenberger]

The connector is located on the bottom of the cylinder. There is a plastic cover (held to the cylinder with 2 bolts) that covers the bottom spark-plug-coil. The connector is fastened to the outside of that cover. It locks in place, and since I broke the lock on mine, I can't really speak as to the official way to get it to release.. My way was to pull on it. That was effective, but did result in an undesireable (but easily fixable) result.

 

To seperate the connector, there is a long tab on the bit fixed to the plug-cover. You have to lift the tab a bit and then pull on the rubber/plastic connector from the O2 sensor. To reconnect - push them back together until the tab locks them.

 

(Fix for the broken retainer - a small tie-wrap around the coil-cover/connector after they're slid together. Works fine. Cost minimal. Try not to break yours.)

 

Any yes - I would guess (don't have one to look at) - an RT might require some tupperware removal to get at the connector.

[dirtrider]

Morning KT

 

To add to Don’s above info-

 

On the 1200RT you have to remove the upper and lower tupperware on the side you want to disconnect the o2 sensor on.

 

You don’t really have to unclip that long dual-ended connector piece from the back side of the lower spark plug cover. After removing the tupperwware, if you simply lie on your back you can see up behind the connector enough to easily use a very small screwdriver to release (and hold released) the little lock tab on the body of the long connector to allow the o2 sensor plug to be pulled free of that long dual-ended connector. Once that little locking tab is held open the o2 wires and plug will pull right out. If it won’t you probably don’t have the locking tab pried out far enough.

 

(or) If you have done it more than a few times you can remove the lower panel 2 front screws by the exhaust pipes and the 3 rearmost lower screws then hold the lower end of the panel out slightly, then working blind and with only one hand use a small screwdriver and pry the o2 plug locking tab open while using the remainder of your hand to work the o2 sensor connector free. I would strongly suggest you not try this method until you become very familiar with how that connector comes apart.

 

[NewBlue]

Okay, for all us "non-techy" folks who are just learning about our R1200's, could someone provide a location and "how-to" on accessing to do the unhook/hook method? Pictures would be especially appreciated. While I'm riding an '09 R1200RT, my guess is the difference has more to do with the Tupperware than the sensor locations.

 

THANKS

 

You were given information on the O2 sensor connector removal and it's always good to know but I would suggest you leave it be. It's not something I would add as a general maintenance procedure. You have an 09 with a 3/36 warranty so no real need to do the unhook/hook method. As noted connectors can be fragile to include the pins, wires, and seals.

If it ain't broke don't fix it.

[KTsRidin]

You have an 09 with a 3/36 warranty so no real need to do the unhook/hook method.

Well, it may be an '09, but I'm well out of warranty at over 47k.

 

If it ain't broke don't fix it.

The idle acts very much like described, so I'm wondering if that's not the issue for me. But, if there's concern that I'm too "fumble fingered" to get to the sensors without damage, the shop who services the RT (Morton's in Fredericksburg, VA) is outstanding and listens to my feedback concerning the bike.

 

Thanks to all for your guidance. In all likelihood, I'll listen to you all and get with a trained mechanic to show me the ropes since there's the potential for damage... ah well. So much for being self reliant

[NewBlue]

You have an 09 with a 3/36 warranty so no real need to do the unhook/hook method.

Well, it may be an '09, but I'm well out of warranty at over 47k.

 

If it ain't broke don't fix it.

The idle acts very much like described, so I'm wondering if that's not the issue for me. But, if there's concern that I'm too "fumble fingered" to get to the sensors without damage, the shop who services the RT (Morton's in Fredericksburg, VA) is outstanding and listens to my feedback concerning the bike.

 

Thanks to all for your guidance. In all likelihood, I'll listen to you all and get with a trained mechanic to show me the ropes since there's the potential for damage... ah well. So much for being self reliant

 

Sorry, I didn't know your mileage so I assumed it was below 36000 and, well you know, made an ass out of me.

 

You can try doing the plug/unplug but just be aware that it can be misleading. The system is designed to use a preset value when a sensor signal is faulty, limp home mode, so even unplugging a functioning, or not functioning, sensor might not cause that much of a change in engine performance. Pins can get pushed out of the connector, bent, etc. creating a new problem, not always the best way to perform diagnostics. Don was pointed to the O2 sensor, wether it's the sensor or wiring, by using a GS911 and watching the real time values. Much better to do this and get a starting point than guess.

 

I know people look for a similar problem and correction when surfing the net to save time and money but when possible it's much better to see a signal out of line and move forward from there than to guess and hope.

 

You can certainly do what you want to your bike I just wanted you to be aware of some things that can cause additional ptoblems or mask the original.

 

Anyway, Good Luck

[KROWBAR]

FWIW. I had a surging issue with my Power Stroke diesel in my F-250. Turned out to be the connector from the TPS to the wiring harness back to the ECU. Cleaned the contacts and surge went away.

[Don_Eilenberger]

Thought I'd post an update..

 

The problem reoccurred. Rather quickly actually, but not bad enough to really bother me. A day or so after the successful report I made above, I noticed the idle quality start to deteriorate.

 

Finally it became bad enough to annoy me, and since I had a full day ride planned for Sunday, this past Saturday night I swapped out the old O2 sensor for a "new" one. The ones I got with the wiring harness basically looked brand new. Where the old one had a slightly tarnished and discolored exterior - the ones in my box of spare parts was shiny and new looking. Anyway - I swapped them out, and checked with the GS-911.. the O2 sensor was again working fine, and the bike idled smoothly.

 

That lasted for about 100 miles yesterday. It was wonderful while it lasted. I'd forgotten just how smooth this bike is. The new O2 sensor did away with a surge condition at small throttle openings making it much easier to creep along in traffic in 1st gear. It seemed to smooth things out at higher RPMs also. The most significant change was the periodic drone/vibration I could feel in the footpegs when riding in 5th gear was completely gone. This has been with me since perhaps 12k miles on the bike - it was a buzz in the pegs at about a 1 second interval (BIG clue there!) at fixed speeds (other big clue) in 5th gear at highway speeds. Acceleration or decel - it would go away. I always had chalked it up to an odd harmonic, and since it never got worse, didn't worry about it. What I should have realized was the ~1 second interval indicated something with the mixture/O2 sensors, which typically have about a 1 second "cross-over" pattern from lean to rich. Plus the O2 sensors are ignored on accel/decel, with mixture being shut down on decel, and on the "map" on accel. The sensors really do their regulation primarily at constant throttle.

 

Anyway - hooked up the GS-911, and sure enough, #2 O2 sensor was dropping out again at idle. Rev'd the engine a bit to a steady 2.5K RPM, and watched - and after about 20 seconds, it would drop out again.

 

So.... by now it was obvious it was NOT the sensor that was the problem.

 

What to do.. what to do? Since the problem cleared up every time I unplugged and plugged the sensor in - that points to the connector to the wiring harness, or the harness itself. Wiggling the wiring going to the connector doesn't seem to change anything. The connector is quite firmly connected (with a watertight seal) when plugged together, so wiggling it wasn't possible.

 

So - what I tried - I cleaned the connectors with some contact cleaner (good stuff that leaves no residue), then hunted down my tiny bottle of Stabilant-22 (http://www.stabilant.com/appnt20h.htm) and treated the contacts, plugging them back together with the Stabilant still wet (how it is used..) Stabilant has fixed some wonky connectors like this in the past for me, BMW recommends it's use in their car manuals for any low voltage/current signal connectors, and other reports I've seen on it indicate it's good stuff.. http://www.ducati.ms/forums/80-hall-wisdom/96088-waterproofing-your-ducati.html

 

Started the bike, and it took about 15-20 seconds for the O2 sensors to start to respond. When they did - the new sensor (#2) showed slightly larger peaks on the voltage steps then the #1 sensor did (in the other cylinder - an original 44,000 mile sensor.) That might be due to a better connection, or it might be just a good connection and a younger O2 sensor.

 

I haven't taken the bike out for a day ride yet, did this last night, and this morning it was 34F.. so I drove to work. I'll update this when I see how this "fix" holds up.

 

One note on diagnostics - I checked the engine fault codes, and the BMS-K had stored a code for an "O2 heater failure, sensor #2" from when I unplugged the O2 sensor and ran the bike. I'm guessing that means that the heater sensor/voltage is OK on that sensor, and I'm looking at a signal path problem from the sensor. Makes sense since the O2 signal is low voltage and very low current.

 

 

[NewBlue]

A fault such as this can be maddening for sure. Hopefully cleaning the connector solved your problem.

You can remove the male terminal from the connector and check the drag on the female side. A loose pin can add to the frustration as well as a bad terminal to wire crimp. Ohm out the wiring from the O2 connector to the ECU connector (not my favorite, prefer using voltage drop tests, but OK as an initial test).

I don't know what sensor they are using but if it is Titanium Dioxide you could hard wire it, that style does not use an air reference, if it comes to that.

 

Best of Luck

[Don_Eilenberger]

The pins used are surprisingly small (especially for the O2 heater circuit IMHO).. and unfortunately solidly molded into the connectors.. both male and female.

 

I did clean inside the female connectors using a tiny touchup thingie I have that is made for micro-touchup work on car paint. It has a round fuzzy brush on the end of a toothpick sort of stick - the brush is smaller then the OD of the female connectors.

 

I can assume the male connectors are OK since they're on the O2 sensor itself, and got replaced when I swapped out the sensor.

 

If the problem persists, as much as I hate to do it - I'll probably end up swapping out the (female) connector on the wiring harness. I have a complete spare harness, but before hacking it up I'd see if BMW sells a repair connector - they probably do. I've been soldering things since I was about 8 years old, meaning well over 5 decades, and I believe I can still do an acceptable solder joint..

 

The sensor is the Japanese slim body one.. which I seem to recall is different from the Bosch ones. Have to look at the actual PN on the body and see if I can dig up any info on it.

[99Roadster]

Don,

 

Before you replace the connector is it possible to try soldering all the connections together to bypass the connector? You will essentially be doing that to at least one side if you use a repair connector.

 

It will at least eliminate the plausable faulty connector and send you looking elsewhere if it continues.

[Don_Eilenberger]

I could, but it would be a PITA since the soldering would have to be done laying on the ground basically under the cylinder head... and I don't really see I'd be much ahead. If I solder in a replacement connector - that would answer the question if it was the connector or someplace else in the harness. I've replaced the O2 sensor (so we can assume THAT connector is OK), then if I replace the harness connector - we could assume that connector is OK, and if the problem reoccurs, it is somewhere upstream.

 

Given the symptom, which has been consistant, of disconnect/reconnect the connector (with both the old and "new" O2 sensors) and the problem goes away for a while, it pretty much screams that the problem IS in the connector.

 

I'm really hoping the Stabilant puts an end to the problem. It has worked for me before when nothing much else would on wonky connectors. Their claim is the electrical conductivity of a connection done with Stabilant is as good or better then a soldered connection. The material actually creates a conductive bond when electricity flows through it.

 

We'll probably know if this worked - and holds up - next weekend. There are several riding events. All it will take is some half-way decent weather to get me out on them.

[jack d]

This is an interesting thread and maybe I missed something. Has an assumption ever been made that the O2 sensor is actually reporting things correctly and that there may be a mechanical problem. I apologize in advance if I missed that in the previous posts.

[dirtrider]

Afternoon Don

 

I would caution against soldering the 02 sensor wires as that is more than likely where the 02 sensor gets it’s ambient reference air from. You could probably solder the heater wires but that doesn’t sound like the side the problem is on.

 

Also remember that 02 connector to the engine harness has two sides to it, there is a connector on the engine side that almost mirrors the 02 sensor side with about the same type plug and pins. There is a small (maybe 2” long) inter-connector between the 02 plug and the engine harness so maybe your problem is on the harness side of that inter-connector.

 

Another thing to consider is the 02 sensor might be the messenger and is just reporting something like a dripping fuel injector or a misfiring stick coil. If the combustion isn’t complete on a cylinder the oxygen in the spent exhaust gas will reflect that.

 

[NewBlue]

This IS an interesting thread.

I believe the voltage reading was low indicating a lean mixture.

You can watch an O2 sensor to see if it reacts by providing a vacuum leak or give it a whiff of flammable vapors in an intake port.

 

Depending on the sensor type it could be hard wired but I wouldn't do that unless I was sure of it.

[Don_Eilenberger]

Afternoon Don

 

I would caution against soldering the 02 sensor wires as that is more than likely where the 02 sensor gets its ambient reference air from. You could probably solder the heater wires but that doesnt sound like the side the problem is on.

I had no intention of doing so.. any soldering would be on the upstream side of the wiring harness connector (and that only if absolutely desperate.)

Also remember that 02 connector to the engine harness has two sides to it, there is a connector on the engine side that almost mirrors the 02 sensor side with about the same type plug and pins. There is a small (maybe 2 long) inter-connector between the 02 plug and the engine harness so maybe your problem is on the harness side of that inter-connector.

Huh? There are two O2 sensors. Each is connected to the harness by a connector that goes directly into the harness just past the connector (it's quite a short - perhaps 2-3" - lead going to the plug. There is no further plug upstream of this that I can see.

Another thing to consider is the 02 sensor might be the messenger and is just reporting something like a dripping fuel injector or a misfiring stick coil. If the combustion isnt complete on a cylinder the oxygen in the spent exhaust gas will reflect that.

That might be a possibility - EXCEPT - there would be no explaination for the symptoms getting better when the O2 sensor is unplugged and plugged back in. AFAIK - that would have NO effect on a dripping injector, or a misfiring stick coil.

 

FWIW - I have a complete set of stick coils, and way back when the symptom first appeared (crappy idle), I swapped the coils out one by one. No difference in the crappy idle. While I could do that again, I still see no explaination for an O2 sensor plug connection fixing a bad stick coil (or injector.) EDIT - I also don't see having one of two coils on a cylinder resulting in a complete lack of output from the O2 sensor, and that's what is happening. I might expect it to go rich, but dead? Nahhh..

 

FWIW-II - I'm very familiar with "shooting the messenger" - ie - blaming the device showing the problem rather then finding what's causing the problem. I believe in this case - repeated disconnects/reconnect fixes have shown where the problem is. Now it's just a matter of fixing it.

[dirtrider]

Afternoon Don

 

No answer for you on why it acts differently after you un-plug it than plug it back in other than the fueling computer instantly goes to fixed open loop mapping on that side when it loses o2 input. Maybe (big maybe here) that also dumps any block learn fueling offset due to other causes.

 

As far as that o2 connector goes, look very carefully at the other end of that 2” long plastic inter connector (the one snapped into the lower coil cover. See if you find a right angled plug similar to the o2 sensor end. Can’t say for sure on your R bike but the RT’s have that other end plug.

 

 

[Don_Eilenberger]

Afternoon Don

 

No answer for you on why it acts differently after you un-plug it than plug it back in other than the fueling computer instantly goes to fixed open loop mapping on that side when it loses o2 input. Maybe (big maybe here) that also dumps any block learn fueling offset due to other causes.

The unpluggging and plugging back in is done with the engine/ignition off - so no learning or loss of learned offset would happen. It's the simple physical action - think wiping contacts - that seems to clear it up.

As far as that o2 connector goes, look very carefully at the other end of that 2” long plastic inter connector (the one snapped into the lower coil cover. See if you find a right angled plug similar to the o2 sensor end. Can’t say for sure on your R bike but the RT’s have that other end plug.

 

The Roadster doesn't have that connector.. the plug for the O2 sensor goes to a Y where it meets with the lower ignition coil wire, and both go into a cloth-taped wiring harness bundle. On my R12R - the O2 sensor plugs directly into the connector shown about mid-photo. I'll try and get some photos..

 

Got the photos and found something "interesting.." - will continue when I get the photos edited and uploaded..

[Don_Eilenberger]

OK - ready to continue..

 

Here are the new and old sensors. The old one is the one with the white tip..

 

 

Here is the connector from the O2 sensor to the wiring harness:

 

 

Another view of the connector (highlighted in box):

 

 

As can be seen - no secondary connector. The wiring from the O2 connector goes directly into a harness (along with the lower plug wiring, and later the TPS from the throttle-body, and the knock sensor and cylinder head sensors.)

 

What I found - was sort of interesting. I removed the vent hose from the cylinder up to the air-box to get a better view of the wiring and make sure there were no connectors I was missing. What immediately caught my eye was the rather odd bend the wiring harness took, and a bright spot rubbed in the paint on a rib on the engine - caused by the harness rubbing against the rib. The wiring was actually being pushed into the rib by the vent hose. I looked it over carefully, and didn't see any damage to the wiring, but just in case, I added some plastic sheathing to the harness, and also rerouted it a bit so it wouldn't be trapped between the hose and the engine rib.

 

As usual - BMW had overtightened the tie-wraps holding the harness to the bottom of the throttle-body, so those got snipped (which made it easier to get the sheathing on) and new ones installed just snug enough so the wires woudn't move around. It seems every BMW I've owned has suffered from overtight tie-wraps on the wiring harness.

 

The last photo shows the funny bend in the harness with the sheathing over it - before reinstalling the vent hose (and you can see the bright spot on the engine rib):

 

 

I hooked up my GS-911 (I'm thinking of leaving it connected under the seat since I've been using it so much, I could connect via BlueTooth to my Netbook and do some one the road O2 sensor recordings.. ) and all looks well. Good crossover pattern coming off the sensors, and response like I'd expect (peaking high when accelerating (rich mixture) and going low when deaccelerating (lean mixture, partial injector cutoff..)

 

Now it just needs a good day's riding to see how it's really doing. One immediate difference when it's running right - the transmission rattle at idle is gone. That seems a not uncommon complaint on hexheads (and some oilheads).. wonder how many other bikes are suffering the same problem. With the sensors both working, idle is very smooth indeed.

 

Thanks for all the thoughts - I may seem to be ignoring some, but they do stick around, and sometimes nudge me a bit when I'm looking at the bike. I wouldn't have found the bright spot (which if it wasn't a problem was certainly going to be one sooner or later..) if we hadn't gotten to discussing the connectors and if there was a secondary one.

 

Best,