Flaming out on the interstate, 10 miles on the low fuel lite

[CycleRob]

Yeah. Riding along at the speed limit in heavy traffic at 65+ MPH (105+ kph). The air temp's about 75F (23.9C) and just before this I was going slightly faster for about 15 minutes. The motor is literally loafing along at that speed under a small throttle opening. I roll off the throttle for the traffic just a little, then roll it back on, and . . . . . there's nothing there for about 2 full seconds! Zero power like the kill switch was turned off! The sensation was definitely fuel starvation. I worked the throttle on-n-off rapidly while it decelerated quickly in 6th gear, and it came back to life. Cars behind me luckily were distant. Did the same thing again 2 minutes up the road, but only for a shorter 1 second duration. Also noted is that the normally crisp, immediate throttle response is now a little flat with minute hesitations. I'm feeling really worried now, imagining where in the bright sun a dead engine will leave me and if I will have roadside room to stop. Bummer.

I immediately remembered the posts on the other F800 site where a handfull of F800-S/ST riders described just such an event. It occurs when it's hot outside, the speed is relatively fast/sustained -and- the low fuel light has been on, but there is enough fuel left for at least another 15--20 minutes riding at speed. It logically points to either the fuel pump or it's finned electronic controller module reacting badly to heat buildup. With the low fuel light on for a while, the fuel level drops below the usually submerged fuel pump, completely canceling the liquid fuel's cooling effect. That may or may not be a cause, but it's an important factor.

Minutes after the flame-out I exited the divided highway and stopped at the first good gas station (Shell) to fill up. I removed the seat for a look-see at the fuel pump controller and wires. I knew immediately the small flannel rag neatly covering the fuel pump housing's outer circumference was a bad thing. What was I thinking? With the seat on, it cuts off major cooling airflow to the finned fuel pump controller. I have made this same type of a trip before, running perfect, with the low fuel lite on, but in much colder weather. After the gas-up and cleaning rag relocation to the left rear body panel side pocket, out on the road it ran perfect immediately. Just like it has for the previous 5,850 miles (9,415km). At the BMW open house, my intended destination, I spoke with the service dept manager about the flame-out event and got a surprised reaction from him. He said he'd "speak with the rep about it" relaying all the details (Low fuel lite, sustained speed, warm/hot weather). He said I could leave the bike for them to check it out, but it wouldn't be until Tuesday that it'd be looked at. I said no, it's running AOK after the gas up. The 50 minute fast ride home was totally flawless, assuring me the flame-out was NOT a preliminary symptom of component failure.

Lessons learned so far:
--Don't put anything in/on/over the fuel pump's rubber cover area that disrupts normal airflow.
--The F800 S/ST fuel pump controller is susceptible to malfunctioning if overheated.
--The oddly shaped rubber cover over the fuel pump appears to be a post production correction for a design error.

In my garage with the seat off again I looked at the odd protruding rubber tower where the pressurized fuel line exits the fuel pump housing. It is shaped like that to function as an air scoop, capturing the breeze blowing under the seat and forcing it out around the controller's Aluminum heat exchanger fins. My putting a rag there apparently caused an out-of-the-blue, new-2-me, potentially serious problem.

.

[owldaddy]

Rob, have you posted this on the other F800 site? As I understand it, many have had similar problems, but I don't think there has been a real fix found. I have a buddy who has a F800GS that was doing something similar that the dealer was unable to fix for a long time. BMW did something to the ECU the last time. So far so good for him this time, but many have had repeated problems that the dealers have been unable to fix. I don't talk to my friend much about his GS, he gets annoyed when he thinks about all the problems this bike has given him. The GS almost got sold he was so angry with it.

[CycleRob]

I'm hoping it's the underseat rag that caused an overheated fuel pump controller to get wiggy. Fortunately there was no permanent overheating damage to the controller and NO repeat of the flameout anomaly. I will try the same 70 MPH + low fuel light conditions again (but without the underseat rag!!) to recreate the problem. It is unusual how out-of-the-blue the problem surfaced -and- then just went away after a fill-up and the heat trapping rag was removed. I wonder if other F800 riders are/were aware of the dangers and consequences of blocking underseat cooling airflow to the controller and fuel pump cover. I occasionally lurk, but Posting "there", lately has had a lower priority.

.

[touchton]

This is interesting about the under-seat cooling. My only low fuel incident with my F800ST was I ran it out of gas with 4 miles to empty on the OBC. I just had to check the accuracy of that thing. But, I was assuming it would go to zero before the gas ran out.

I wonder if the Sargent seat I've installed will cause a problem with the under-seat cooling? I hope they took that into account when they built a seat for the F800ST.

[sweatmark]

Had a similar experience myself while riding my wife's F800S down to BMW dealer for fork seal replacement. Wish I could remember the status of fuel level, but do remember that the summer morning was cool in western Oregon, perhaps 55-60 degF.

In my case, was riding for 20 minutes at sustained 60-70 MPH along gently curving country road, stopped at an intersection, rolled on throttle through 1st gear, and then lost all power, coasting to a stop at side of road. Don't recall whether it took time to restart or not.

Prior to my dead throttle experience had read about similar occurences on the F800 board.

[CycleRob]

I have duplicated the same riding conditions (but without the rag) with no repeat of the flame-out failure. It looks like the heatstroke inducing underseat rag was the cause. I went just over 20 miles on the low fuel lite, cruising 60+MPH in 75F.

.

[owldaddy]

I was chatting with the local dealer last week, he tells me that they have not had any complaints about this problem. Seems it is only common on the web. Ya right

[CycleRob]

After the service rep showed his "first time ever" thoughtful surprise toward my flame-out problem, I got the a polite sour response from him when I mentioned it happened to several others on the F800 Net ChatBoards. His tone of voice implied `those people have no credibility' . . . . as if we're making stuff up as a sport. Here's a newsflash to dealers; most riders with a problem look for a simple solution on the Net first, KNOWING the dealer's answer is to make an appointment or leave the bike . . . with a credit card number.

It's a good thing `the problem' has not returned, because I really dread leaving my bike there for others to ride. Especially during the prime-time riding season, trying to recreate the problem I'm sure the BMS-K memory will have no record of. We'll see when my GS-911 arrives.

.

[sweatmark]

Another flameout:

Mrs. Sweatmark was riding her F800S today during return from our annual Oregon BWM rally.

Downhill stretch of highway after crossing 5000 ft. mountain pass
55-65 MPH sustained
Cool rainy weather, 45-50 deg F
Fuel tank 3/4 full
Running about 90 minutes non-stop since fuel/coffee stop (where bike rested 1 hour in 60 deg F)
Fresh, 92 octane gas (10% ethanol for us on west coast)
Full of clean oil
Electrical system at standard voltage
Nothing under seat in region of fuel pump and its controller
F800S/ST low seat in place
My wife was using her AirHawk seat pad

She shifted 5th to 6th, motor died, she coasted to stop. We were very lucky - no, blessed actually - that this occured right at a small roadside restaurant around 2500 ft elevation, with nothing but forest and narrow road shoulders for 10 miles either direction. I parked Rockster and then pushed F800S to the restaurant parking lot and semi-dry place under a big Douglas Fir.

We took the hint and had ourselves a nice bowl of home-made beef barley soup, plus coffee.

I told Mrs. Sweatmark about the F800 fueling glitch that Rob reported here, plus the anecdotes on the F800 Riders board. My guess was that today's problem was hot pump/controller... and only a restart attempt would help confirm or deny.

Following that great soup warm-up (45 minutes cool-down for the bike), the F800 started first try. No problems for the ensuing 90 minutes to home.

I'll post this over on F800 and check to see what folks (and hopefully BMW Motorrad) are doing for resolution.

[CycleRob]

My bike's flameout in gear under way has never returned. It runs perfect. It had to be the fined fuel pump controller getting overheated by the rag obstructing the cooling air, the BMS-K computer reading it's abnormal power draw, then shutting it (and the engine) down.

There can also be a sidestand switch malfunction causing Mrs. Sweatmark's F800S flameout. If/when it happens again in 5th or 6th gear, quickly pull in the clutch and gas it. If it stalls, try the starter. If it runs well with the clutch pulled in but flames out when you let out the clutch, it's the sidestand switch! Another trouble spot is the battery terminal bolts tightness. They loosen over time and should both have a SS lockwasher between the bolthead and terminal washer.

My GS-911blu showed a "gear position switch one-time malfunction" and a "left signal bulb open" trouble codes. No mention of anything that could or would have caused the flame-out.

One surprise in the GS-911's component realtime readout was the #1 and #2 detonation sensor voltage readouts. The numbers changed rapidly from under 200 to over 2,000 as the throttle was worked rapidly. Even though there isn't an actual detonation sensor, the latest electronics can monitor the nanoseconds of sparkplug voltage buildup and/or the voltage itself before the sparkplug actually fires, reactively detecting pre-ignition. Whatever! It works and I get no pinging using Chevron/Texaco Regular grade gasoline for the last 7,857 miles (12,645km) since new.

.

[sweatmark]

There can also be a sidestand switch malfunction causing Mrs. Sweatmark's F800S flameout. If/when it happens again in 5th or 6th gear, quickly pull in the clutch and gas it. If it stalls, try the starter. If it runs well with the clutch pulled in but flames out when you let out the clutch, it's the sidestand switch! Another trouble spot is the battery terminal bolts tightness. They loosen over time and should both have a SS lockwasher between the bolthead and terminal washer.

Rob, got the battery connections under control. Our F800S did have the required washers set on both battery terminals.

Based on catch-up reading over on the F800 board, I'm of the opinion that the lambda sensor temperature theory might apply, or some combination of contributors to over-lean condition (and magic detection of knock) that causes engine shutdown.

We've experienced this flame-out problem twice (1st with me riding to dealer for fork seal replacement) and conditions were similar:

[*]long stretch of easy riding at constant higher speed,
[*]moderate ambient temp,
[*]1/2-2/3 tank of 92 octane,
[*]nothing under seat to impact fuel pump

Makes me think that high compression motor tuned lean for emissions and fuel efficiency has weak spot during closed-loop running at small throttle angle and light load.

[CycleRob]

There is no excuse to justify these flame-outs. Either there's a faulty component or the computer's firmware has hidden errors. Has your car EVER stalled or flamed out on the highway, in traffic, in ANY kind of weather? Of course not!!

My bike just had another stall episode in 94 degF heat after just over 20 miles running on the low fuel light. The computer's "miles remaining" was over 57 miles when it stalled after I stopped for a red light. I had a very brief warning of the impending stall out because the engine braking was strong even after it went below 2,500 RPM, where it usually ends the deceleration fuel cut then turns on the idle mixture. I could also hear the misfiring in the exhaust. Just before the engine stopped turning in 1st gear I gave it a few quick throttle jabs to prevent the stall. It did absolutely no good. It stopped dead. Trying to restart it was like starting an old bike after it stalled out because you didn't turn on the reserve tap soon enough. It clearly was fuel starved and hitting randomly on only 1 cylinder for about 5 seconds of cranking, then it started. The traffic light had already turned green as I was whacking the throttle on/off making the tach needle dance to keep confirming it was responding OK. Tapped into 1 st and took off. It ran AOK after it restarted, even at 3 subsequent stops. I gassed up about 5 miles later and it never malfunctioned again. I was MAD as hell because it means there still is a serious heat related stalling problem, but so far onlywhen riding on the low fuel light.

There is definitely a heat buildup or cooling problem with the fuel pump, as the flame-outs only occur in hot weather with tens of miles riding on the low fuel light, when the fuel pump is no longer submerged in the fuel to cool it. Locating the plastic fuel tank down low under the seat, in the hot air blast from the radiator and engine, more easily reveals any marginal heat related failure problem.

When I first used my GS-911 diagnostic there wasn't any stored fault code on the engine. That's after the "event" that prompted this topic. I'll try to work out a component R&R test with ATL BMW this fall. BMW may want to test those faulty parts to solve this almost common.

[sweatmark]

There is no excuse to justify these flame-outs. Either there's a faulty component or the computer's firmware has hidden errors. Has your car EVER stalled or flamed out on the highway, in traffic, in ANY kind of weather? Of course not!!

Hearty "Amen"... neither have I ever experienced splined transmission input shaft failure on a manual transmission car or truck! But that's another topic.

There is definitely a heat buildup or cooling problem with the fuel pump, as the flame-outs only occur in hot weather with tens of miles riding on the low fuel light, when the fuel pump is no longer submerged in the fuel to cool it.

Based on rider reports from the F800 site, there's both a hot weather/fuel/pump stall problem, AND a steady-throttle, closed-loop, lean-running stall problem that does not necessarily occur during warm weather. Such was the case during our F800S stalls.

When I first used my GS-911 diagnostic there wasn't any stored fault code on the engine.

That's my concern when I ride the F down to our dealership for their troubleshooting. I believe I have a good rapport with the tech staff concerning problem description and investigation, but the going will be tough if/when there's no fault recorded for stall.

Does the 911 have data logging capability? I'll have to check the web site. Need to buy one of these for both our F800 and Rockster.

[boxermania]

I have read all the posts with great interest as the premptors of the event are many but the common denominator I gathered was a leisurely pace at reduced engine loads.

I don't know how many O2 sensors the bike has and if they are the narrow or wide band type, however, I would subscribe to the theory that under specific light load conditions the ECU is leaning the A/F mixture to that range where combustion is no longer supported.

It would be very interesting to fit the bike with an Innovate LM-1 wideband O2 sensor and monitor the A/F ratio when the event takes place. If Robbie wants to validate my theory he is welcome to borrow my LM-1.

I will also venture to say, based on prior experience with the brand that the dealer techs have no idea of what could be taking place and BMW's response will be "these are isolated incidents, we have yet to see one"

[CycleRob]

My engine stall out conditions were during hot weather and being 10--20 miles on the low fuel light. Because of the almost uninterrupted 40--55 MPH 6th gear cruising I've been doing since the last fuel up, the usually close to actual Average MPG dash display now says 72!!! That says a lot about the EFI programming. There is NO light load small throttle surging or lean indication when riding the bike. It runs sweet at every RPM . . . . until the 2 times I run tens of miles into the low fuel light in 94+ degF temps. It doesn't happen every time under those 2 conditions, but city traffic with the resulting heat buildup seems to make it more likely. Both of the flame-outs happened without engine heat that would have run the radiator fan. The fan noise and voltmeter reading dip tells me when it comes on. Both times it has "recovered" in a few seconds and still miles before fueling up. It's exactly as if the episodes were caused by a temperature sensitive, poor power connection to the fuel pump. To me, there's no doubt the problem is in the EFI system, most likely either the underseat finned Aluminum fuel pump controller or the pump itself.

.

[CycleRob]

I have mentally analyzed my bike's hot weather, low fuel light flame-out problem and have come to what I believe is the logical cause -and- an easy solution for it. All overly detailed and carefully put into fun filled sections

The F800 series twins have a "non recirculating" EFI system. There is only one fuel line to the engine. That means all the fuel that the fuel pump pumps goes to the injectors. There isn't any fuel return line from the pressure regulator like the majority of all EFI systems have. With the F800 system, the fuel travels extremely slow in the fuel lines during normal riding. The BMS-K computer powering the fuel pump controller controls the fuel pump to better match fuel demand. The problem that is happening on my bike I believe is the combination of:

--an engine heated underseat fuel tank almost devoid of road breeze cooling.
--high ambient Summer heat.
--a heat generating electric fuel pump exposed to air rather than cooling gasoline after tens of miles on the low fuel light.
--a low fuel flow typical of normal speed traffic or idling.

***causes***
The fuel being whirled around in the now air cooled, electric motor heated pump chamber to boil - - - AKA "vapor lock". That generats fuel vapor in the downstream lines that will quickly stall the engine -and- make it difficult to restart. Makes sense to me. The boiling fuel heat source will decrease some time after the pump stops. Soon after that the cessation of boiling will enable a shuddering, stumbling, fuel starved, heavy throttle restart.

***The solution***
Provide liquid cooling to the fuel pump!! Keep the fuel moving through it by adding a tiny metered bypass to maintain fuel pressure and provide a tiny flow amount back to the tank's fuel level. Place it after the fuel pump and before it leaves the fuel tank. A simple "T" fitting with a small carburetor main jet sized orifice integrated into it. When the bike runs, it "pees" a small stream thru an attached hose back to the fuel level in the tank. That moving fuel will provide liquid cooling to the fuel pump as it works with the bike running 50 miles deep into a low fuel light condition. BTW, if there already such a bypass bleed system on the pump, a simple external heat exchanger may be required.

I'm sure the engineers at BMW would come to the same conclusion . . . that an improperly designed, non return EFI fuel system has some annoying and serious performance limitations. Maybe it doesn't get that hot in Germany?

[Boxer]

My engine is basically the same, although as Rob noticed it is tilted a little more upright compared to his ST. Also, I probably have more air flow under the bike and around the engine than the ST, but I rode in all sorts of weather conditions in June and was on the low fuel light several times, but I never experienced what you all are describing. At one fill up I put 3.85 gals in the tank which only holds 4.2 gals. This was in central New Mexico and it was quite warm. This flame-out issue only occurred to me 2 or 3 times back when I had under 1500 miles on her, and it was under wet conditions. So I attribute that to the recall issue about the canister drain hose, which I redirected myself after it did that.

[CycleRob]

It definitely was a hot weather fuel pump overheating vapor lock issue. Twice I've run over 30 miles on the low fuel light in cooler weather, at cruising speed for one and in traffic for the other, with zero running malfunctions. Since removal of the fuel pump assembly requires a big special tool of dubious availability, I may just gas up in the really hot summer before getting 10 miles on the low fuel light.