F800GS flat lamda sensor reading

[Gary Schulte]

A buddy has been having intermittent and difficult to reproduce problems with his 09 f800gs.  It seems he only had problems after an hour plus of riding.  Essentially the bike dies abruptly for a second or two, but will eventually pick back up as long as the motor is still turning.  It is as if the ignition is being cut temporarily and has a very abrupt on/off type of problem.

After looking at usual suspects like sidestand switches and such and the dealer not being able to reproduce the problem (it takes quite a while for the problem to develop).

Today he went for a 120 mile ride and was able to reproduce the problem with my GS-911 hooked up and we logged some data.  What most stands out is that during the 'events' the lambda sensor is reading dead flat - a zero.  Even when the bike is off, idling, or whatever - it never reads 0.  I found this gem on the gs-911 site:

Some points to note on evaluating your Lambda graph:

• The higher the value, the richer the mixture.
• Contrary, the lower the value, the leaner the mixture
• a flat line at any level (at operating temperature) depicts a faulty Lambda operation
• jumping from approx. 200 mV to approx. 650/700 mV depicts a correctly functioning Lambda control syste

So, it sounds to me like an o2 sensor problem.  Correct?  

If so, what is the likelihood that this is a downstream symptom rather than the problem?  I am guessing a fouled o2 sensor wouldn't magically unfoul itself and start reading normal values a second or two later.  He is certain that he had no problems before the dealer did the 12k service, so maybe a bent pin, pinched wire or just an odd coincidence.

Thoughts?

Thx

[Gary Schulte]

I also noticed that fuel pressure drops precipitously before the motor cuts out.  Log attached.  Lines 173-177 are characteristic of the stalls...

2009 F800GS

Any input appreciated.

Thx

[Haakon Aas]

This sounds totally stupid...even to me :-)

I am NOT familiar with the F800GS but believe it has a regular electric fuelpump like all the other injected BMW`s have.

There can only be one (actually 2, but the second is NOT plausible) reason for the fuel pressure to drop like that- the pump stops for some reason.

I would first check that the pump has electric power all the time. (hook up a LED or something in the pump supply circuit) I say LED as THAT will go dark at once...

A "normal" bulb will not ...

Such a drop in the fuel pressure will also make the Lambda give a zero reading- I think???

I need a BIGGER screen on my PC to see ALL the values... I will resize and filter off the less important values :-)))

Now sleep :-)

Haakon

 

PS, the other "low fuel pressure" causes are: intermittient faulty pump, faulty wiring inside the fuel tank or the pressure regulator.

PS-PS DOES the bike actually have a fuel pressure monitor?? Damn... :-)  I live in the black ages with my 2000 model GS  :-) 

 

 

 

--
You received this message because you are subscribed to the Google Groups "GS-911 Field Diagnostic Tool for BMW motorcycles" group.
To post to this group, send email to gs-...@googlegroups.com.
To unsubscribe from this group, send email to gs-911+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/gs-911?hl=en.

[Gary Schulte]

As I understand it, since there is no fuel feedback loop for the 800gs fuel pump, the pressure is variable based on the throttle/injection requirements.  So fuel pressure variation is integral to the functioning of the fuel injection system.

Also, looking later in the log at other stalls and at another log, fuel pressure does not strongly correlate to the stalls.

Still looking for confirmation/refutation on the expected values for lambda sensor reads.

Thx

Gary

[Peter Shearman]

Check that the fuel tank vent relief system is functioning OK.

This was a recall issue on earlier F800's.

A vacuum could be building up in the fuel tank eventually starving fuel to the pump.

Is the breather hose pinched or crimped.

Stop for a while, some air bleeds in and bike starts again?

Try running with the gas cap cracked partially open & see if fault remains.

 

Peter in OZ 

[Gary Schulte]

More info:  just prior to the problem rearing its head, he took the bike in for service.  They replaced the fuel tank due to exterior cosmetic cracks.  The canister venting 'fix' has been previously done and there is no vacuum build up in the tank nor 'whoosh' upon opening the cap.  When he brought the bike by yesterday I was able to reproduce the problem in under a mile and right after opening/closing the tank lid.  The second or third time that he experienced this problem and took it in for service, the dealer replaced the fuel pump with a new one.  So far as I know, the only thing fuel related that has not been replaced is the fuel pump controller.

Looking at his logs and at my logs, his fuel pressure does seem to drop particularly low.  Mine never drops below about 3500 even under hard acceleration.  His drops as low as 900 at times.  I don't know whether this is caused by the ECU (on the basis of bad sensor readings) or the fuel pump controller or what.

The stumbles seem to occur when the lambda sensor is reading 0.  Typically the fuel pressure is below 1500 at this point, but not always.  At some points, especially on extended compression braking, the lamba sensor reads 0 without a stall.  So the o2 sensor could be a red herring.

At this point it is conjecture until he can ride another 100 miles to try a remedy.  I suggested pulling the lamba sensor wire when it starts to be a problem to see if it changes the behavior.  If that does nothing, then bypass the fuel pump controller.  If that doesn't help, a last guess might be the 'empty' sensor in the tank that prevents the fuel pump from running when there is no fuel, but that one is much harder to try to bypass.

Any other suggestions?

TIA,

Gary

[Haakon Aas]

I guess...KNOW...I was wrong (I think) :-)

I checked the RepRom for the F800R and it does have a fuel pressure sensor.

(I believed the pump was running all the time (the engine was running) and that a pressure reliefe valve was used to keep a stable fuel pressure)

The repair manual says the fuel pressure settings are from 2,5 bar to 5,0 bar

My guess is the fuel pressure sensor is used to start and stop the pump (I could NOT find a pressure relife valve..) but its very late now- I must check more tomorrow!

"We" know the sensor is working... so whats left is ...DO the pump start every time the pressure is low.... think not? :-(

----------------

My (probable) mistake was to think of older models.... single cylinder F bikes.

There the pump run at all times the engine is crancked OR running- and a pressure relife valve keep the fuel pressure VERY stable... at 3,5 bar (+/- 0,4 bars)

Anyways- We know the pressure sensor IS working.... but the fuel pressure drops..., meaning the pump does NOT run when it should.

As the setup design, using a pressure sensor, makes this way more complicated (and vulnerable).

Less is better :-)

---------------

You say: "So far as I know, the only thing fuel related that has not been replaced is the fuel pump controller."

I guess THAT is a part of the BMS (ECU) and VERY expensive.

 

Lets go back to the time before this started- SERVICE? 

What could they have f...up?

IF the engine is starved of fuel the Lambda will show LOW or zero...

So my (unqualified) guess is still fuel supply- OR the lack of it

Off to bed :-)

 

 

 

----- Original Message -----

From: Gary Schulte

To: gs-...@googlegroups.com

Sent: Monday, September 12, 2011 5:00 PM

Subject: Re: [GS-911] Re: F800GS flat lamda sensor reading

More info:  just prior to the problem rearing its head, he took the bike in for service.  They replaced the fuel tank due to exterior cosmetic cracks.  The canister venting 'fix' has been previously done and there is no vacuum build up in the tank nor 'whoosh' upon opening the cap.  When he brought the bike by yesterday I was able to reproduce the problem in under a mile and right after opening/closing the tank lid.  The second or third time that he experienced this problem and took it in for service, the dealer replaced the fuel pump with a new one.  You sayLooking at his logs and at my logs, his fuel pressure does seem to drop particularly low.  Mine never drops below about 3500 even under hard acceleration.  His drops as low as 900 at times.  I don't know whether this is caused by the ECU (on the basis of bad sensor readings) or the fuel pump controller or what.

[Lloyd Provin Jr]

I don't know whether the F800 and the rest of the "New-Gen" motorcycles stop their pumps when correct fuel pressure is reached, but I suspect that the voltage to the pump is simply reduced to maintain pressure.  The fuel pump driver is available seperately: 16147720778.
 
Disconnecting the 02 sensor should eliminate it as the problem.  If your problem disappears, suspect the sensor or it's power supply.
 

Bud Provin
TheNickwackettGarage.com
802.483.2460
Now stocking Gerbing's Heated Clothing

 

---

From: b...@bmw-f650.com
To: gs-...@googlegroups.com

Subject: Re: [GS-911] Re: F800GS flat lamda sensor reading

Date: Tue, 13 Sep 2011 04:17:57 +0200

[Haakon Aas]

Yes.... might be... but the pump is the simplest form of centrifugal pump

Hmm- I think you are right- not a ON / OFF function.

Damn... I am way too old for this "new- gen" bikes....and will shut up :-))))

:-)))))

The Lambda should NOT, can not be the problem (I am still thinking of the OLD models!!!)

1: The Lambda is ONLY for "iddle" conditions. (not sure but think up to 2000 - 2500rpm`s and small trottle openings )

2: If the Lambda values are "unplausible" the ECU will just use a "none Lambda" map..a bit more fuel consumption... a bit more polution..but NOT stalling the bike- NOT!!!

The same with almost ALL other sensors, even the TPS (Throttle PositionSensor)- IF that fail you can still "ride back home"... but  it may easily stall if you open the throttle too fast :-)

AGAIN.... this is from experiences with the OLD F models:-)

I do pray a "simple" sensor fault will not leave you stranded...with the 2 cylinder F`s

The "old" (single cylinder) F`s  have a multitude of ECU backup maps to get you "home"...

------------

If you unhook the lambda the bike will just run a bit rich at iddle.

Thats all. BUT, do so, its fast and easy, just one plug.

Haakon

[spsg]

Hi Gary,

 

Just a couple of thoughts; as we find Lambda sensor failures are rare!  Oil or diesel in fuel (what was the fuel decanted into when the tank was changed?), more commonly blockage of the small air holes around the outside of the sensors body, if they are blocked it wont work causing misfires or stalling (and a flat line readout), common causes: - mud; protective oils or chain lube, if the sensor is clean now are the holes still blocked?  Severe contamination can be almost impossible to rectify.

 

Good luck

 

Regards

~Mike~

[Jim5941]

Read this thread: http://www.advrider.com/forums/showthread.php?t=577345

Pay particular attention to posts from Joel Wisman ;-)

[Haakon Aas]

Thanks!!! :-) dont even know why I tried to give advice :-)
But- I learned a lot now!!
Haakon

----- Original Message -----
From: "Jim5941" <jim.w...@gmail.com>
To: "GS-911 Field Diagnostic Tool for BMW motorcycles"
<gs-...@googlegroups.com>
Sent: Wednesday, September 14, 2011 1:06 AM
Subject: [GS-911] Re: F800GS flat lamda sensor reading

>

> Read this thread: http://www.advrider.com/forums/showthread.php?t=577345
>
> Pay particular attention to posts from Joel Wisman ;-)
>

[Gary Schulte]

Man that is a useful thread.  His problem does not sound quite like Ben's, but their troubleshooting tips are invaluable and I will forward this over to him.

All their tweaking/prodding/guessing really *really* makes me appreciate the gs-911.  I am never going any substantial distance without it.  

Thanks for the input all.  

-Gary

On Tue, Sep 13, 2011 at 6:06 PM, Jim5941 <jim.w...@gmail.com> wrote:

Read this thread:  http://www.advrider.com/forums/showthread.php?t=577345

Pay particular attention to posts from Joel Wisman  ;-)

[Jim5941]

Yes, Joel is "the man" x 10