Rough idle 30 sec after cold start

[Kerry in Mpls]

Good work. Since you get a rough idle with the O2 sensor on the right side connected, that sounds like the problem side--it's the side getting leaner. I would now swap left and right injectors and repeat the O2 tests and see if the problem moves to the left side.

 

If that has no effect, I would then swap the O2 sensors side to side.

Roger - right oh, except I think you got the sides mixed (I did too when I first read it.)

 

One thing to point out - on a R1200xx when an O2 sensor fails - the mixture for both cylinders is set to the mixture determined by the working O2 sensor. With both disconnected (or failed) - as noted - the mixture defaults to a map in the ECU, and the maps tend to run rich to protect the engine.

 

Same thing as the CAT-CODE plug did on the R1150xx O2 sensor equipped engines. Pull the CAT-CODE plug and it took the O2 sensor out of circuit and the bike ran under a mapped mixture. Surge usually eliminated.

Sorry about that, the way I wrote that was confusing. Let me clarify:

 

With right sensor connected and left side sensor disconnected, cold idle stayed near normal for two minutes (well past the usual problem time).

With left sensor connected and right side sensor disconnected, cold idle became rough starting at about 45-50 seconds.

 

Since the engine apparently IS able to operate properly using only the right sensor, and it is NOT able to operate properly using only the left sensor, the left would seem to be the culprit.

[roger 04 rt]

Kerry, you're right, let me rewrite it below.

 

 

Good work. Since you get a rough idle with the O2 sensor on the left side connected, that sounds like the problem side--it's the side getting leaner. I would now swap left and right injectors and repeat the O2 tests and see if the problem moves to the right side.

 

If that has no effect, I would then swap the O2 sensors side to side.

[roger 04 rt]

Good work. Since you get a rough idle with the O2 sensor on the right side connected, that sounds like the problem side--it's the side getting leaner. I would now swap left and right injectors and repeat the O2 tests and see if the problem moves to the left side.

 

If that has no effect, I would then swap the O2 sensors side to side.

Roger - right oh, except I think you got the sides mixed (I did too when I first read it.)

 

One thing to point out - on a R1200xx when an O2 sensor fails - the mixture for both cylinders is set to the mixture determined by the working O2 sensor. With both disconnected (or failed) - as noted - the mixture defaults to a map in the ECU, and the maps tend to run rich to protect the engine.

 

Same thing as the CAT-CODE plug did on the R1150xx O2 sensor equipped engines. Pull the CAT-CODE plug and it took the O2 sensor out of circuit and the bike ran under a mapped mixture. Surge usually eliminated.

 

Don, A couple things:

 

--Kerry provided a lot of gs-911 data and the O2 sensors operate normally. Hence my diagnosis. (As you and Kerry pointed out I mixed up left and right.)

 

--You're right, eventually the R1200 uses the one working O2 sensor to fuel both sides, but not right away. See this data (chart below): http://bmwsporttouring.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=835682#Post835682. Notice that the disconnected side stayed rich for a while, also note that closed loop AFR is 13.7 because this rider was using an LC-2 setup, on Kerry's bike it will be 14.7 but the point about the disconnected side staying rich will still be true.

 

 

--On the R1150, removing the Code Plug doesn't disable Closed Loop, although that does happen on the R1100GS/R/RT. And disconnecting the O2 on the R1150 results in limp home operation, which isn't richer. Link:here: http://forums.bmwmoa.org/showthread.php?56990-2004-R1150RT-Wideband-O2-Sensors&p=775329&viewfull=1#post775329.

 

Edited October 30, 2016 by roger 04 rt

[roger 04 rt]

Kerry, How're things going?

[Kerry in Mpls]

Hi Roger and all -

 

Sorry for the lack of info lately, I took a bit of a break from investigating this.

 

Grab your beverage of choice and read through this recap of pretty much everything I have done - most of these were done on individual cold starts.

- I disconnected both idle actuators just after startup, to eliminate the possibility of one of them malfunctioning after a bit and causing the problem. No effect.

- I started the bike with only the lower coils connected. Ran rough right away (no 35 second wait).

- I started the bike with only the upper coils connected. Ran even poorer than with just the lowers, and died when I tried to increase revs.

- I removed all four of my coils and installed a friend's four known good coils from his 07 R12RT. No effect on the original symptom, and I reinstalled my original coils.

- I replaced all four spark plugs with new ones. No effect on the original symptom (but they will remain, obviously).

- I have looked at GS911 data showing idle actuators operation, and they seem to be operating very well. They operate nearly identically. I did a reset on them just to be thorough, and no surprise, they still appear to operate perfectly.

- I swapped the O2 sensors left-to-right, no effect.

- I disconnected each O2 sensor, one at a time, to check effect. With the right-only connected, there was no change. With the left only connected, there seemed to be a slight improvement. But really inconclusive.

- I swapped the fuel injectors left-to-right, no real effect.

- I purchased and installed a pair of used O2 sensors from a '13 R12RT with 21000 miles on it. GS911 indicates they are working as well as my originals. No improvement.

- I reset the adaptives a while back. The additives are still at zero, the multiplicatives are still at one. I suspect it takes some extended riding at road speeds before those begin to change. I have only ridden a few miles since I started this thread (sadly).

 

I have used the GS911 to log real time data and to check for stored faults many times. There have never been any faults stored other than when something was intentionallly disconnected. The realtime data captures indicate that the idle actuators are working properly. They also indicate that all of the O2 sensors I have used are working properly, with respect to the heating cycles and the sensor voltage cycles.

 

Next on the list:

I have purchased a pair of used fuel injectors from a '13 R12RT with unknown miles, seller indicates they work fine. They should arrive in a few days. They are used items from eBay and it's hard to know that they work perfectly. But they were cheap and swapping them in will be an easy experiment. If there is no improvment, having the injectors professionally cleaned, and getting a before/after flow report is still an option.

 

Some info that I hadn't shared yet:

I used an IR thermometer/gun to measure the temps of each exhaust pipe near the engine while it is running, and there is quite a difference. Just a minute or two after startup, the right pipe is over 500 degrees F, while the left pipe is only about 400 degrees F. I didn't check the pipe temps in the early days of this investigation. But I can say that the pipe temp difference did not change sides when I swapped the O2 sensors left-to-right or when I swapped the injectors left-to-right. The right side is always hotter. All engine runs were done in a garage, tail end near the the open big door, and no noticeable breeze which might cool off one side. I don't want to jump to conclusions on this, maybe a temp difference is normal, or maybe there is some non-obvious cooling effect when I run the engine in the garage. Not sure what to think about that.

 

 

[roger 04 rt]

Kerry,

 

You've certainly tried a lot, in a systematic way. If we stick with the theory that it's misfiring on one side when the mixture leans-off, then there would be unburned fuel in the combustion chamber, as the exhaust valve opened. That fuel would burn in the header prior to reaching the O2 sensor and cause an increase in exhaust temperature.

 

Have you visually confirmed that all four plugs fire? It could be a bad wire or (unlikely) a bad driver in the bmsk, or a bad ground wire on the primary side of one of the coils.

 

If all four plugs fire and can jump an eighth inch gap, then you may need to check compression.

 

Maybe DR could add his thoughts here.

[dirtrider]

 

Maybe DR could add his thoughts here.

 

There isn't much left is there?

 

A compression test would definitely be something I would personally look at (just to eliminate that up front)

 

How do the pipe temps compare to the cylinder head temps? If both point to the SAME cooler running side then maybe something to look at there.

 

As long as there is a temperature gun handy I would probably re-run the upper plugs only & lower plugs only looking at pipe temperatures on both iterations (might point a finger--at s-o-m-e-t-h-i-n-g)

 

Looking back there seemed to be an improvement with both o2 sensors disconnected. Maybe re-address that angle with possibly a pipe temp reading with no o2 sensors in play (not sure what that will prove but maybe give you a direction to look in)--might point to BMS-K calibration or o2 sensors coming into play too soon.

 

Do we know if the o2 heaters are ACTUALLY WORKING? I see the data showing commanded o2 heating but are they actually getting the power & heating? Maybe take your second set of o2 sensors & plug them in to the bike's o2 wires then with them hanging loose start the engine to see if you can feel the o2 heaters getting hot (just to eliminate this)

 

With no newer BMS-K to try as a substitute maybe try unplugging the Intake Air Temperature sensor (in L/H top of air box) then plugging in a 15 ohm resistor into the harness connector as a quick test (see if that shows on the GS-911 data). If that helps then (I don't know, but we can look into that if the 15 ohm resistor changes anything)

 

The problem seems to be in the fueling transitional region (between cold start open loop & fully warmed up closed loop)-- That is a difficult thing to diagnose if all concerned parts seem to check out OK.-- Probably leads back to BMS-K programming or ?????.

 

 

Something else that I would personally look at (eliminate) is to remove both valve covers then rotate the engine (or unplug injectors & crank it with starter) then make sure that all the valves open about the same travel. Real long shot but something I would verify to ease my mind.

 

It seems with all that Kerry has done it looks like no outright smoking gun so it might be one of those deals that just has to lived with until it gets worse (or bad enough) to find a failing component. (it might just be an inherent thing to a few 06 1200RT's with their current BMS-K calibration)--

[sardineone]

Definitely looks to be something wrong on one side of your motor. Maybe something mechanical like throttle body sync??? As I adjust mine, the engine management automatically smoothes my idle in a few minutes using my method here Backyard mechanic throttle body sync

[dirtrider]

Definitely looks to be something wrong on one side of your motor. Maybe something mechanical like throttle body sync??? As I adjust mine, the engine management automatically smoothes my idle in a few minutes using my method here

 

Morning sardineone

 

 

Kerry's problem is happening at idle (no throttle) & unfortunately there is no way to adjust the idle balance on the hexhead at idle as that is controlled but the computer & steppers with the throttle cams sitting on the idle stop screws.

 

 

 

 

[Kerry in Mpls]

Some new data:

 

Exhaust pipe temps, measured on top of each pipe, maybe 8" away from cylinder:

 

Running on only lower coils (uppers disconnected)

Left temp: 375 Right temp: 550

 

Running on only upper coils (lowers disconnected)

Left temp: 475 Right temp: 700

 

Running with all coils connected

Left temp: 425 Right temp: 625

 

I did these engine runs in the order shown above, and did not wait for the engine to cool down between runs. The pipes actually cool down rather quickly. The temps fluctuated a bit, and the temp gun updates the reading only every second or two, so I would allow +/- 25 degrees on all of these values. Basically, the right is consistently hotter in all of these scenarios.

 

I did not do GS911 data captures during the above engine runs.

 

--

 

I then reconnected all coils and did as D.R. suggested - I unplugged the installed O2 sensors, and plugged in the spares (which are the originals) and let them hang in the ambient air.

 

Exhaust pipe temps:

Left temp: 400 Right temp: 600

 

O2 sensor temps:

Ok, this is embarrassing, I measured it but forgot to write it down. I can't remember if they were both 200+ degrees or 400+ degrees. But they were both hot and nearly the same temp.

The bike ran more smoothly in this configuration than it typically has. However the engine wasn't completely cool at startup, and there could certainly be some effect from the unusual O2 sensor setup.

 

--

 

I then did a long (6+ minute) GS911 data capture to get a nice long curve of the cylinder head temps. By the end of the engine run, the left cylinder head temp had climbed to 214 F, and the right had climbed to 224 F (right cylinder head hotter, corresponding to right exhaust pipe hotter). The temps were still climbing and still diverging when I stopped the engine and the ended the data capture.

 

Also interesting is that the adaptive trims finally changed from their reset state during this run. Between the four and five minute mark of this run, the additive trim climbed linearly from zero to about 10, where it seemed to be topped out. The multiplicative stayed at one. Not sure if this happened because of unusual O2 sensor setup, or just because this is the first time since resetting the trims that I have run the engine long enough for it to try to adapt.

 

[roger 04 rt]

Some new data:

 

Exhaust pipe temps, measured on top of each pipe, maybe 8" away from cylinder:

 

Running on only lower coils (uppers disconnected)

Left temp: 375 Right temp: 550

 

Running on only upper coils (lowers disconnected)

Left temp: 475 Right temp: 700

 

Running with all coils connected

Left temp: 425 Right temp: 625

 

I did these engine runs in the order shown above, and did not wait for the engine to cool down between runs. The pipes actually cool down rather quickly. The temps fluctuated a bit, and the temp gun updates the reading only every second or two, so I would allow +/- 25 degrees on all of these values. Basically, the right is consistently hotter in all of these scenarios.

 

I did not do GS911 data captures during the above engine runs.

 

--

 

I then reconnected all coils and did as D.R. suggested - I unplugged the installed O2 sensors, and plugged in the spares (which are the originals) and let them hang in the ambient air.

 

Exhaust pipe temps:

Left temp: 400 Right temp: 600

 

O2 sensor temps:

Ok, this is embarrassing, I measured it but forgot to write it down. I can't remember if they were both 200+ degrees or 400+ degrees. But they were both hot and nearly the same temp.

The bike ran more smoothly in this configuration than it typically has. However the engine wasn't completely cool at startup, and there could certainly be some effect from the unusual O2 sensor setup.

 

--

 

I then did a long (6+ minute) GS911 data capture to get a nice long curve of the cylinder head temps. By the end of the engine run, the left cylinder head temp had climbed to 214 F, and the right had climbed to 224 F (right cylinder head hotter, corresponding to right exhaust pipe hotter). The temps were still climbing and still diverging when I stopped the engine and the ended the data capture.

 

Also interesting is that the adaptive trims finally changed from their reset state during this run. Between the four and five minute mark of this run, the additive trim climbed linearly from zero to about 10, where it seemed to be topped out. The multiplicative stayed at one. Not sure if this happened because of unusual O2 sensor setup, or just because this is the first time since resetting the trims that I have run the engine long enough for it to try to adapt.

 

Great data. With this information, it seems to me like all the electronics are doing their jobs and that there is a mechanical reason for the rough running, that is probably leading to misfiring, which would then lead to unburnt fuel oxidizing in the right header, and explaining the higher ride side temps.

 

Since you've run on only the uppers and only the lowers, all four coils must be firing. The data says all the sensors and closed loop are doing what they should.

 

Regarding the trims, I've found that it takes up to 30 minutes at any rpm/TPS combo for the long term trims to fully develop. The behavior of yours post reset seems reasonable.

Edited November 6, 2016 by roger 04 rt

[dirtrider]

Evening Kerry

 

I'm not sure what to make of the pipe temperature difference.

 

The cylinder head temps are not that different so it doesn't sound like a major combustion difference.

 

Are the exhaust pipes about the same color (browning or bluing) IR heat guns are not the best at reading light colored or reflective surfaces.

 

Or the pipe difference might be that the L/H side has a coating of carbon inside as using the side stand can promote more oil seepage past the rings on that side, so possibly a little L/H side oil burn on start up.

 

At least it sounds like your o2 heaters are powering up & working so we can take that off the list of possibilities.

 

With all that you have tested & recorded we are running out of possibilities on this thing.

 

I guess a compression test would be the next thing to eliminate.

 

I checked my service bulletins on the 2006 1200RT & nothing showing on a BMS-K CIP for any issue like you are seeing.

 

Just be glad that you aren't paying the dealer $100.00 an hour to diagnosis this thing.

 

 

[Kerry in Mpls]

Just be glad that you aren't paying the dealer $100.00 an hour to diagnosis this thing.

Believe me, that thought has crossed my mind many times!

 

The exhaust pipes are about the same color, dull bronze/brown. I have been holding the temp gun 3 or 4 inches from the surface of the pipes so I hope the radiant heat is being detected mostly equally on each side.

 

As I mentioned, I have replacement injectors on the way. I will try them even though I suspect there will be no difference. If so, that's something else to cross off the list.

 

I intend to check the strength of spark at each plug (check ability to jump a gap) as Roger suggested.

 

For the compression test, any advice on that? Most compression test kits assume you will push a rubber cone against a spark plug hole -- is that practical on the RT? Certainly not on the upper plugs, but maybe on the lowers? Or do I need a fitting that can thread into the plug hole?

[dirtrider]

 

For the compression test, any advice on that? Most compression test kits assume you will push a rubber cone against a spark plug hole -- is that practical on the RT? Certainly not on the upper plugs, but maybe on the lowers? Or do I need a fitting that can thread into the plug hole?

 

Evening Kerry

 

 

I use a Snap-On compression gauge. The Snap-On uses about 10" long hoses with different spark plug threads on the ends (the hose easily reaches down into the deep spark plug well on the BMW boxers).

 

If you don't have access to a kit that uses a hose you might get something that will go into the lower spark plug hole.

 

Be sure to hold the throttle wide open for the compression test. You also need to unplug the injectors, OR, just jump the solenoid at the starter (key turned off).

 

I have compression tested some unusual engines in my life by adapting old (gutted) spark plugs to a hose or pipe extension. If you want to make something you can weld (or maybe even epoxy) a long tube to a gutted spark plug that fits your cyl head threads.

 

Or possibly rent a hose type compression tester from a local tool rental.

 

Edited November 6, 2016 by dirtrider

[Kerry in Mpls]

Short and sweet:

 

Left side compression: 182

Right side compression: 190

 

Pretty even.

 

[dirtrider]

Evening Kerry

 

Yep, cross that one off the list.

[roger 04 rt]

Next up is the visual test of spark ... ?

[Kerry in Mpls]

Next up is the visual test of spark ... ?

Right, probably this coming weekend.

 

The pair of fuel injectors that I bought should arrive by then too, so I'll probably swap those out too and see if anything changes.

[Kerry in Mpls]

The injectors (used, from a 2013 R12RT) showed up and have been installed. I also took this opportunity to install new O-rings on the injectors (upper and lower on each) as well as on the idle actuators, to eliminate the small chance of an air leak on any of them.

 

And ... success! Smooth idle from cold start to at least four minutes. That is well beyond the infamous 35 second point where the rough idle would usually begin.

 

I've done a grand total of one cold startup, but I will definitely do a few more to verify that the improvement is consistent.

 

The significant temperature difference between the exhaust pipes still occurs, but maybe not related, and/or not important?

 

Anyway -- progress!

 

Edited November 12, 2016 by Kerry in Mpls

[lkraus]

The significant temperature difference between the exhaust pipes still occurs, but maybe not related, and/or not important?

 

I'm glad things are changing in the right direction.

 

I checked the temps on my headers after a ten mile, 55mph ride. At the exit from the head I had 389ºF on the right side, 400ºF on the left. It may not matter at all, but a 100º difference seems pretty big.

 

Without reviewing the whole thread, I don't remember any mention of a recent valve clearance check or throttle balance. The TB should definitely be checked since the injectors have been changed.

 

[dirtrider]

 

Without reviewing the whole thread, I don't remember any mention of a recent valve clearance check or throttle balance. The TB should definitely be checked since the injectors have been changed.

 

Evening Larry

 

Kerry's issue is at idle & on the 1200RT the computer controls the idle balance (no way to change or adjust it).

 

From his earlier trapped GS-911 data the idle cross side stepper balance looks well within reason.

[roger 04 rt]

The injectors (used, from a 2013 R12RT) showed up and have been installed. I also took this opportunity to install new O-rings on the injectors (upper and lower on each) as well as on the idle actuators, to eliminate the small chance of an air leak on any of them.

 

And ... success! Smooth idle from cold start to at least four minutes. That is well beyond the infamous 35 second point where the rough idle would usually begin.

 

I've done a grand total of one cold startup, but I will definitely do a few more to verify that the improvement is consistent.

 

The significant temperature difference between the exhaust pipes still occurs, but maybe not related, and/or not important?

 

Anyway -- progress!

 

Nice progress Kerry! Once you're convinced everything is good, you might put the old injectors back in and confirm. Label them carefully.

[lkraus]

Kerry's issue is at idle & on the 1200RT the computer controls the idle balance (no way to change or adjust it).

 

I'm thinking the off-idle balance may be different now with good injectors vs whatever balance it had with (at least) one dirty/faulty injector. It sounds like he has fixed the fundamental idle problem, I'm just thinking there may be room for additional improvement.

 

Still befuddled by the large difference in exhaust temperatures. It may not significant but it would irritate my OCD tendencies.

Edited November 13, 2016 by lkraus

[Kerry in Mpls]

OK, hang on.

On the next two cold starts, the original problem still occurred, right on schedule.

I noticed that on the first startup with the replacement injectors, it took some extra cranking before the bike fired. I suspect this was because the injectors and fuel lines had to eliminate some air before they could begin delivering fuel. I wonder if that made the computer decide to the engine richer than normal on that run, or to run rich longer than usual, which masked the idle roughness?

In any case, it seems the injector change did not address the root cause after all.

[roger 04 rt]

That's too bad. 🤔

 

I would go right to the spark gap test and be certain that all four coils can fire equally healthy sparks. I think DR usually recommends 1/8-3/16".

 

It is still possible that one of the coils isn't firing. Let's rule that out.

Edited November 13, 2016 by roger 04 rt

[dirtrider]

That's too bad. 🤔

 

I would go right to the spark gap test and be certain that all four coils can fire equally healthy sparks. I think DR usually recommends 1/8-3/16".

 

It is still possible that one of the coils isn't firing. Let's rule that out.

 

Afternoon Roger/ Kerry

 

Yes, I usually use a 3/16" electrode gap.

 

On the dual electrode plugs I usually just use a cheap single electrode auto spark plug with the electrode gap opened up to 3/16" (use a 3/16" drill bit as a gap gauge)

 

In the open air a standard gap duel electrode plug won't tell you much about coil output.

 

OR you can probably buy a spark output tester at the auto parts store for a few dollars-- Those are a SPECIAL SHORT spark plug with a base clip on them to clip to the engine case & a very large spark gap. If your coil will spark one of those then your coil output is more than likely good-to-go.

 

Edited November 13, 2016 by dirtrider

[Kerry in Mpls]

I did the quick and easy spark test using a cheap spark plug with an oversize gap. Turns out 3/16” is pretty big. The gap from the center electrode to the threaded cylinder body might not even be 1/8”. But I increased the outer electrode gap out to nearly 3/16” and gave it a try anyway.

 

All four coils were able to generate a spark, but as expected the spark was between the center electrode and the closer plug body, not to the bent outer electrode. Also, the sparks on all four was orange, not blue. Not sure if (hotter) blue is expected?

 

It just occurred to me that I'm not sure if this should have be done with the engine running, I didn't do that, just turned it over with the fuel injectors disconnected so it wouldn't start.

[dirtrider]

I did the quick and easy spark test using a cheap spark plug with an oversize gap. Turns out 3/16” is pretty big. The gap from the center electrode to the threaded cylinder body might not even be 1/8”. But I increased the outer electrode gap out to nearly 3/16” and gave it a try anyway.

 

All four coils were able to generate a spark, but as expected the spark was between the center electrode and the closer plug body, not to the bent outer electrode. Also, the sparks on all four was orange, not blue. Not sure if (hotter) blue is expected?

 

It just occurred to me that I'm not sure if this should have be done with the engine running, I didn't do that, just turned it over with the fuel injectors disconnected so it wouldn't start.

 

Afternoon Kerry

 

You do want to see bluish & snappy but doing it cranking with the original spark plugs in place probably lowers the system voltage enough to not give the greatest of sparks.

 

Maybe retry with a battery charger on the battery, or a spark plug on each side removed. (or engine running)

 

If all 4 looked the same then the chances of all 4 coils being BAD & all 4 being the same is not likely.

 

 

[Kerry in Mpls]

I bought a device specifically for measuring spark strength, to be able to accurately test a known gap, and used it on all four coils.

Short story: all look good.

 

Details: The device is a Thexton 404 adjustable spark tester.

Photo: https://goo.gl/photos/fynjm3jFS4WHGfot8

It permits finely adjusting the gap you want to check. The scale is calibrated in kV rather than distance. There are lines for SE (small engine), then 20kV, 30kV, and 40kV, which are the voltages required to jump those gaps. I initially set the gap to a measured distance of 3/16" of an inch to try to duplicate the test gap you guys recommended. I connected the device to the left upper coil (therefore that plug would not fire), started the bike, and it easily jumped the gap. So I wound the device out to the 20kV mark, which is about 3/8" inch and repeated the test on all four coils. Nice strong spark on all four.

Short video while testing lower right coil:

 

We've had an unseasonably mild fall in MN. We are two weeks beyond the previous record for the latest first frost. Reality sets in today. It is raining heavily now (you may hear it in the video), and that will turn to snow later today. So I am going to put further hands-on experiments on hold at least until the weather clears up. Snow blowing into the garage while working on the bike can be annoying. :-)

 

[dirtrider]

Morning Kerry

 

Thanks for the update on your progress.

 

This is pointing more & more towards a BMS-K calibration issue.

 

Have you tried directly powering the fuel pump with 12v yet? (ie by-passing the FPC)

 

Kind of a long short but IF the FPC is not supplying enough pump current that can decrease the injector squirt (especially in open loop)

 

Any idea on what color your FPC is (silver or black?)

 

[Kerry in Mpls]

Morning Kerry

 

Thanks for the update on your progress.

 

This is pointing more & more towards a BMS-K calibration issue.

 

Have you tried directly powering the fuel pump with 12v yet? (ie by-passing the FPC)

 

Kind of a long short but IF the FPC is not supplying enough pump current that can decrease the injector squirt (especially in open loop)

 

Any idea on what color your FPC is (silver or black?)

FPC is the original silver one. I have not tried bypassing it.

[dirtrider]

Morning Kerry

 

Thanks for the update on your progress.

 

This is pointing more & more towards a BMS-K calibration issue.

 

Have you tried directly powering the fuel pump with 12v yet? (ie by-passing the FPC)

 

Kind of a long short but IF the FPC is not supplying enough pump current that can decrease the injector squirt (especially in open loop)

 

Any idea on what color your FPC is (silver or black?)

FPC is the original silver one. I have not tried bypassing it.

 

Morning Kerry

 

I think I would try by-passing it as a test (not much left to test)

 

In any case, if you still have the original silver FPC you should have (and carry) an FPC by-pass harness as those old silver FPC's can (& do) fail while riding at about any time.

 

Or just buy a new black FPC then carry your old one as a spare.

[roger 04 rt]

Kerry, I went back and looked at the early data you posted. When you get a moment, check where the tank vent hose from the tank vent solenoid runs. On most bikes I've seen data from, only one of the LCF values changes when the tank vent opens, and the hose only runs to the left TB. On your bike, both LCFs seem to be affected. See if you have connections to just the left TB or to both the left and right. You might also try plugging both TBs at the vacuum nipple and see what effect if any that has.

 

I was thinking about why the new injectors seemed to work for a while. The other thing you might try is resetting all adaptive values and seeing what happens at the 30 second mark. I'm suggesting it because I see more movement of the long term trims than I'd expect at idle. This may not indicate what's going on but if resetting the adaptive values changed things it might give us something else to look at. If you run this test, take a full set of data like your earlier runs so we can see what the LCFs and long term trims do.

 

Happy Thanksgiving

[Kerry in Mpls]

Kerry, I went back and looked at the early data you posted. When you get a moment, check where the tank vent hose from the tank vent solenoid runs. On most bikes I've seen data from, only one of the LCF values changes when the tank vent opens, and the hose only runs to the left TB. On your bike, both LCFs seem to be affected. See if you have connections to just the left TB or to both the left and right. You might also try plugging both TBs at the vacuum nipple and see what effect if any that has.

 

I was thinking about why the new injectors seemed to work for a while. The other thing you might try is resetting all adaptive values and seeing what happens at the 30 second mark. I'm suggesting it because I see more movement of the long term trims than I'd expect at idle. This may not indicate what's going on but if resetting the adaptive values changed things it might give us something else to look at. If you run this test, take a full set of data like your earlier runs so we can see what the LCFs and long term trims do.

 

Happy Thanksgiving

Hi Roger -

 

Yes on my bike there is a hose only to the left TB, the right TB has a cap on the nipple. Back when we were tossing around the idea of a vacuum/air leak, I did pull the hose off of the left TB and put my finger on the nipple, in case the hose or something upstream was a source of a leak, but there was no change. The cap on the right side seems very secure, doubt there is a leak there. Looking through several GS911 captures, the LCFs always seem to be generally in sync. Are you expecting the left side (and only left side) to see and respond to a richer mixture when the tank vents?

 

Do you know what the possible range of values is for the additive trims? I have one capture where I can see the additive trim climbed from zero to about 10 - took about a minute - and stayed there. Then a later capture shows the additive trims starting at about 10, and decreasing to about -2, that took about 90 seconds. If the possible range is -100 to +100, then this isn't a huge change. If the range is -10 to +10, I agree this seems like a large change over a rather short time period.

 

In case there is interest, here is the most recent GS911 capture, collected on Nov 12 when the fuel injectors were replaced. (this capture shows the additive trim decreasing)

https://goo.gl/photos/azx3R6wnGYep2oGHA

 

Belated Happy Thanksgiving to you.

 

[AndresB]

Hello,

 

I have the exact same problem in my 2006 RT. Have you reached a conclusion on how to fix it?

[Kerry in Mpls]

Oops, I thought I had followed up on this, but apparently not!
In short, I had checked, cleaned, and/or replaced everything I could on my own.
The next time I took the bike to the dealer for scheduled maintenance, I asked them to also listen to the rough idle to see what they thought.
When I went to pick up the bike they said they found nothing mechanically wrong with it, but that the idle improved a lot after they added a bottle of BMW fuel system cleaner.
I was very skeptical that the fix could be that simple, but on the ride home I could tell that yes it was significantly better.
They gave me a second bottle to take with me. I used the contents of that bottle over the next few fill-ups.
By the time I had used up the second bottle, the problem was completely gone.

 

No guarantees that this will fix the issue for you too, but it's something cheap and easy to try - I wish I had tried it first!

[Newbiebmw2021]

my bike is going thru similar issues. gonna try this