Misfire when engine very hot, error 13 - Rear O2 Sensor always lean

[llama]

Hey everyone, I'm hoping the brain trust can shed some light on what might be the issue with my Uly.

I've got a 2008 XB12X with just over 100,000km on the clock. She's been 100% reliable over all these years, never left me stranded. In my ownership I've replaced the cooling fan that died around 85,000km but otherwise just serviced the beast with oil, filters and tyres. I do have a leaking rear rocker cover gasket so I've bought a new set for both cylinders and also intake seals and a new Bosch O2 sensor and I'll rotate the engine out next month when I have a free weekend and replace those bits.

Over the last two months however a fault has developed, but only when the engine is very hot (eg I've been out for a fast ride for an hour and then get stuck in traffic). The check engine light has come on four or five times but usually there is no difference in the running of the bike, but three times now when the check engine light is on it has stuttered a bit - it feels like it's running on one cylinder. If I get a clear stretch of road I can cruise around ~90km/h for a few minutes and the CEL goes out again and all is well. It doesn't seem to stutter when at part throttle and cruising, only when under load, and only when very hot.

I've read the check engine code with my Buelltooth and get error 13 - Rear O2 Sensor always lean. Does this mean the engine is running lean (eg intake leak or similar) or does it mean the engine is running rich and the ECM is telling the rear cylinder to lean off? Could it be the ignition coil? Cylinder head temp sensor? I've replaced the ignition leads about 10,000km ago so I don't think they're the problem as they've been fine for 8,000km, but I have since discovered that I'm running sparkplugs that are slightly too hot (DCPR8E where most people seem to run DCPR9E) but not sure that would cause this issue. I've cleaned all the earths I could find but again I don't think this is the issue as it only occurs when the engine is very hot and under load.

What should I be checking here?

Cheers,
Andrew

 

[Cooter]

Buell DDFI has a built in safety to kill one cylinder if the head temps get too hot (typically about 275*, under temp settings. Also referred to as 'Parade-mode'.
You can check and change the settings using the ECMDroid that you checked codes with. Also check historical codes, not just current codes.
You can also check your current head temp under Live Data, while the fault is happening to be sure thats the cause.

Code 13 is pretty common for an old/lazy O2. While I am wholly against swapping parts as a guess, see what happens when you swap in the new one?

Spark plug heat range has zero-point-zero affect on engine temperature. Either NGK plug will work fine.

 

[34nineteen]

A faulty head temp sensor will also cause the bike to run hot.

Is the fan turning on?

 

[Barrett]

Andrew: active fault code #13 actually consists of 3 seperate/distinct faults. sensor may be lean....rich...inactive. no definitive way to tell. do as mentioned and check history for ALL stored historical/active fault codes.

#13 almost always associated with failing ground...particularly the dedicated ECM ground. ID all grounds...remove...clean all surfaces...check integrity of wiring end connection....reinstall....ECM ground/grounds pin #11 as memory serves...specifically that one.

due to age....mileage...fact that rideability failure increases as temps increase....a great time to check your ignition coil specifically the wiring harness connection and OHMS values of the coil. your symptoms suggest it might have reached the end of its life-span.

 

[llama]

Thanks everyone. I didn't think to check the "live" values while it was happening, mostly because by the time I'd pulled over and popped the seat off, connected the Buelltooth, connected phone to that etc things had cooled off slightly and the error has cleared, it's only listed as an historical error (and the exact error is "13 Rear O2 Sensor always lean", not rich or inactive. Not sure how the Buelltooth knows the difference? It didn't just say "13").

I'll leave the Buelltooth connected next time I go for a ride so I can catch it in the act. I don't know if the o2 sensor has reached it's end of life or not, I am going to replace it while the engine is rotated anyway as it's waaaay easier to do it then than through the fan opening later.

Fan is working as it should (damn it's loud!), comes on about 5 minutes after starting riding then ramps up and down according to speed. Sometimes runs after turning off the ignition if it's been in stop-start traffic. It did fail ~15,000km ago and I did ride it around not working for a few months (thinking maybe it wasn't working due to being winter and cooler - might have cooked something during that time...)

I'll check the grounds again as it doesn't hurt, though again it only occurs when the engine is very hot and I thought a bad earth would show up more often. Though the CEL has come on 5~6 times in two months the misfire has only happened 2~3 times, and only for a few seconds. I only just discovered how to check the coil with a multimeter so I'll do that too.

It's very intermittent but coming into summer here in NZ would be good to sort

 

[Cooter]

You can just leave the Buelltooth connected. Mines been on for about 5 years now. If you have an issue, stop (as you should) and connect to ECMDroid. That 10-20 second process will be invaluable for diagnosis.

If your fan is coming on all the time even when the bike is cold, you probably have "speed-based fan control" checked. It will activate the fan anytime above 20MPH.
I can't imagine why a fan would need to blow on a cold engine, or why use it above a speed where it's getting WAY more CFM of air than it could ever provide, but hey, it's an option. Makes people feel better to hear it I guess?

IMO, reduce major wear on the fan, strain on the charging system, the assault on your ears, and uncheck that box.

If you are having an overheating issue with an air-cooled engine, I would look into your running conditions and maybe even the tune. Does it backfire on decel? Is it running lean?

 

[llama]

Ah, never realised I could just leave the Buelltooth connected, that makes sense. I usually have my phone on a RAM mount on the handlebars so I'll be able to connect to it very quickly to get live data.

I guess my fan is set to speed based? When cruising above about ~20km/h it screams away, when I stop at the lights etc it goes very quiet (kinda seems backwards to me - I thought it would run stronger while stationary and not run at all when cruising at 100km/h? Maybe the volume sitting at the lights is a bit much for most people!). I'll uncheck the speed based fan box and see how I get on.

To be clear, I'm not sure I'm having overheating issues or not. I've owned it for several years now and it doesn't seem to be hotter than it usually is (which is to say roasting my right leg from the headers), no popping or crackling on decel, it pulls really strongly. I have only had the CEL come on over the last two months, and the misfire 2~3 times more recently. I think I need to go for a decent ride on a hot day and see the live temp from the Buelltooth to know for sure. As far as I'm aware I've got a stock 2008 tune (stock exhaust but a K&N filter which may lean it out a bit, but that's been on there for years now and the issue only just cropped up in the last two months).

 

[llama]

Completed my engine rotation over the weekend - did rocker cover gastkets, intake seals and fitted a Bosch o2 sensor. As a bonus the bike even still runs after I put it all back together! Took about 11 hours over two days for me to complete.

I measured the resistance across the various pins on the ignition coil and it's a bit out of spec, all resistances are a bit high - primary measure 1.1 ohms (should be 0.5~0.7) and secondary measure 12k ohms (should be 5.5 to 7.5k ohms), so maybe that's the issue. I've asked H-D here for pricing and availability of a new coil, but in the meantime I'll just ride it with the Buelltooth fitted so at least I will know if it's getting too hot, or something else is causing the misfire (assuming I can even get it to misbehave - it's very intermittent).

It's not all beer and lamingtons however - when I get the front engine mount off I noticed that it's starting to split, it's probably about 1/4 gone now. Looks fine when on the bike, still has a good gap but I know it's only going to get worse.



Still, better to find out now than up some deserted gravel road in the middle of nowhere

 

[34nineteen]

Check ebay. There are some new ones on there.

11 hours isnt bad, especially for the first time. I prefer to just pull the frame off the engine rather than monkey around with a rotate. Now that you've done it, the next time you should be able to do it in a fraction of the time. :up:

 

[llama]

Just got back from an hour long ride, it's a nice hot day here in Auckland so a good chance to test the misfire. Of course it didn't misbehave at all, engine ran great, though I did get the check engine light coming on within a few minutes of leaving home and it stayed on the whole time this time.

I kept the Buelltooth dongle connected to my phone (in a RAM mount on the handlebars) as I rode and the hottest I saw was 241°c when on the motorway riding at around 110km/h in 4th (so up near 4k RPM). In traffic it sat around 200ish with the fan whirring away the whole time.

So based on the error and the slightly out of spec ignition coil I'm guessing (correct me if I'm wrong here! wouldn't be the first time...) that the coil isn't producing a good spark, so there is some unburnt fuel being detected by the O2 sensor which is telling the ECM to inject a bit less, hence the error "Rear O2 always lean". I'll order a new coil and see how it goes.



Cheers,
Andrew

 

[Cooter]

:angel:

 

[34nineteen]

I kept the Buelltooth dongle connected to my phone (in a RAM mount on the handlebars) as I rode and the hottest I saw was 241°c when on the motorway riding at around 110km/h in 4th (so up near 4k RPM). In traffic it sat around 200ish with the fan whirring away the whole time.


hold on there HotRod! you're seeing CYLINDER HEAD TEMPERATURE of 466 degrees fahrenheit? is that correct?
that equates to an approx. parallel oil temp of close to 400 degrees fahrenheit.
not only impossible...but catastrophic.
why? because during developmental testing of H-D air-cooled V-twins, 330/290 fahrenheit is considered the upper limits of acceptable head/oil temps before the onset of imminent damage.
you're almost 45% above those developmental standards.

Well, how hot does it get in hell?

 

[llama]

hold on there HotRod! you're seeing CYLINDER HEAD TEMPERATURE of 466 degrees fahrenheit? is that correct?

Uh, yeah, that should be 241°F (damn freedom units ;-P )

 

[llama]

I received a care package from St Paul H-D two weeks ago that included a new ignition coil and finally got around to installing it on Saturday. As Monday was a public holiday where I live and the day was nice and hot (around 29°C) I took the opportunity to go for ride - around 440km of slow traffic, high speed boring motorway, nice twisty tarmac and ~50km of dusty gravel. The bike ran like a champ, no check engine light and no misfire.

It's early days, but I'm sure on a multi hour run in such warm weather it would have stuttered and misfired in the past, and I had got used to seeing the check engine light on my commute in to work during the week. No check engine light and it pulled strong all day - I'm tentatively calling this fixed

 

[upaero]

Check your AFV on the ECM parameters/AFV settings page of ecmdroid.

 

[llama]

Hey upaero, what AFV am I looking for? I haven't actually looked at ECMDroid since replacing the ignition module, the bike is running great, no check engine errors and no misfire.

 

[upaero]

Hey upaero, what AFV am I looking for? I haven't actually looked at ECMDroid since replacing the ignition module, the bike is running great, no check engine errors and no misfire.

It should be close to 100 at sea level. rev-mo.com/afv.

 

[upaero]

I kept the Buelltooth dongle connected to my phone (in a RAM mount on the handlebars) as I rode and the hottest I saw was 241°c when on the motorway riding at around 110km/h in 4th (so up near 4k RPM). In traffic it sat around 200ish with the fan whirring away the whole time.


hold on there HotRod! you're seeing CYLINDER HEAD TEMPERATURE of 466 degrees fahrenheit? is that correct?
that equates to an approx. parallel oil temp of close to 400 degrees fahrenheit.
not only impossible...but catastrophic.
why? because during developmental testing of H-D air-cooled V-twins, 330/290 fahrenheit is considered the upper limits of acceptable head/oil temps before the onset of imminent damage.
you're almost 45% above those developmental standards.

The engine temp he is seeing is in Celsius. The warm up enrichment (WUE) tables enrich the fuel until about 200C. The fan doesn't come on until 220C. This is measured right at the top of the head, very close to the combustion chamber. Of course the cylinder walls will be much cooler. An engine temp of 241C is a bit excessive though, which is a result of running lean.

 

[34nineteen]

I thought that sensor was just a "switch" to turn the fan on and off by basically telling the ECM when it reached a certain point. I learned something today.

 

[llama]

It should be close to 100 at sea level. rev-mo.com/afv.

Sorry for the delay in getting this - I finally had time to do a 45 minute highway loop with the phone on so I could take screenshots. It looks like my AFV is normally 90% - I checked it over a few days multiple times - here it is about 20 seconds after starting the engine at idle:



Here it is after a 5th gear run up a long shallow incline at about 110~120km/h (~3500rpm), still at 90%:



I have *once* seen it at 85.5% on a different day - again at the top of the same incline but this time I went up in 4th (so ~4000rpm+ for a few minutes - this is also the highest temp I've seen over the last few runs, 237°):



That last one is a bit low according to the link you posted "...a number lower than 100 means it is removing fuel due to a sensed rich condition. Generally speaking, your AFV should be within 10% of 100" I checked the AFV setting as outlined in the Revision Moto post and it's set to 90% (ignore the front number, I only have the rear O2 sensor):



So I'm thinking I should set this to 100% (or maybe just over - from https://www.rev-mo.com/openloop: "If it was a stock bike with a stock exhaust we may want to start at 104. This starting point also depends on what your current AFV value, but it should be close to 100"). I am running a K&N air filter but factory exhaust. The bike did have a sportier exhaust on it when I bought it (which I found too loud, hence the stock one now fitted), so maybe the previous owner was messing with settings...

Also on that same page "...Buells have a nasty tendency to develop a lean condition due to inaccurate O2 sensing which shows up as an AFV of less than 100 at or near sea level. In this case, the ECM "thinks" that the motorcycle is going up in altitude, so it is reducing the fuel delivery across the board." - Do you reckon this is what I'm seeing (the engine is sensed to be too rich)? Or simply that the AFV is set too low (90% instead of 100+)?

P.S Having said all the above, since replacing the ignition coil the bike has been running fine, no hiccoughs or pops, pulls nicely through the rev range.