2010 K1300GT

[Honolulu]

2010 K13GT, 30,000 miles. One afternoon to come home from work (bike ran fine in the morning) the bike started poorly, ran rough and would barely hold idle. So:  sudden onset of problem.  Got home, found two headers cool, two hot. GS911 reports O2 sensor voltage about 0.8, thus running rich. Could see the catalytic converter glowing in the tailpipe. Thought this indicated one or more bad coils, no ignition events in one or more cylinders, fuel gases running out the pipes.  Replaced all four coils and four very sooty sparkplugs. The bike now will start and idle and is less rough than before, but something's not yet right. It sounds like it's "panting" and though that may be okay idle for a race bike, on this one it persists to at least 3000 rpm. Have not revved further in the garage and I'm a little hesitant to go on the road, only to end up needing a tow.

Bike has had very intermittent cam chain rattle on startup (nothing to compare it to, but it's a very harsh sound for less than a second).  Have been recommended to replace two orings in the chain tensioner.  That won't be the current problem unless the timing chain has jumped a tooth, in which case the bike may not turn over at all.  That the problem first displayed on cold startup suggest a possible jump, as I understand that is the moment of greatest stress on cam belts/chains.  Is there a way, short of disassembly, to discern correct cam timing?

Next a more basic question: what are the "typical" or "correct" value ranges for the parameters scanned while the bike is running?  I have seen correct O2 sensor behavior, in real time, but for other more esoteric items I have no way to interpret what is output.  Can someone post a tabulation of the correct values or ranges?

TIA

[Honolulu]

Hmmm Scottie, have you anything of value to add?

[Jughead]

Hi Honolulu

The only way to determine correct cam timing is to pull the cam cover, which, on the GT, is a real PITA.

WRT the lambda values, these should fluctuate between rich and lean.  There is no specific value.  If you use the PC app, you can actually display these values as a running graph.  As long as the graph fluctuates up and down, the lambda sensor is doing its job.  I will post a screenshot for you of what it should look like as soon as I can get back to my PC.

[Honolulu]

Jughead:  Thanks, I've seen correct lambda values on this bike, several months ago.  Prior to replacing coils and plugs the value hovered in the 0.8 range, rich enough that the cat was glowing a little.

Bad news, then, about verifying cam timing, although I've gotten as far as replacing coils and plugs, which in retrospect wasn't that difficult despite the humbug of vacuum coolant refilling.

Complaint:  the online fiche is only barely a parts diagram and shows few interconnections or context.  I'm trying to think in terms of sensor inputs to the ECU which could cause the current problem.  Apparently there's a air correction valve at the bottom of the airbox which can get sticky - if this were the case I can imagine that stickiness might interact periodically (less than a second and pretty regular) with the ECU to make the bike "pant" as it does now at idle and up to at least 3000 rpm.  That's where I'll be going next, unless I can get more advanced diagnostics to view the signal from that device.  I assume it's not only mechanical, but the fiche doesn't tell me.

[Jughead]

 Did you replace the coils with new ones or 2nd hand?

[Honolulu]

New Beru coils, four of them, $150 each (on sale!).  NGK plugs, $12 each.

[Jughead]

New Beru coils, four of them, $150 each (on sale!).  NGK plugs, $12 each.

[emoji106]  Have you check which cylinders, if any are running colder than the others?

[Honolulu]

At first, two header pipes were cold.  After replacing coils, all got hot.

Bench testing coil primaries (center pin to sparkplug) finds one coil has infinite resistance, so that coil is shot.  I really should test more carefully, since it's hard to get the ohmmeter test leads into the connector.  Also should mod my test leads to check resistance on pins 1-3.  However that's water under the bridge since the new coils and plugs are in, and have been run a few minutes.

Original question was whether any one can post "normal" readings for all the real-time monitoring performed by the GS911 on this 2010 K1300GT.  Since not even the manufacturer has bothered to reply, I can only suppose they don't know... but how could they not?

[Jughead]

Original question was whether any one can post "normal" readings for all the real-time monitoring performed by the GS911 on this 2010 K1300GT.  Since not even the manufacturer has bothered to reply, I can only suppose they don't know... but how could they not?

There is no such thing as "Normal" reading when it comes to lambda values.  The readings will change based on the fueling at the particular moment the reading was taken.  Since the fueling (and thus the air/fuel ratio) changes with virtually every stroke the pistons take, it is impossible to even predict what it is going to be on the next stroke.

What you need to do is to display the graph, and then make sure the lambda values jump up and down between rich and lean.  The graph should look like this:



In the graph above, the bike is fitted with 2 lambda sensors, thus the yellow and violet graphs.  As you can see here, even 2 sensors on the same motor do not react the same and do not display the same readings.  The GT only has one sensor, so you will only have the one graph.

[Honolulu]

Thanks for that plot, and as mentioned above I have seen proper lambda sensor real-time readout on my bike several months ago.  When the bike ran poorly, the sensor value wavered slightly around 0.8 volts - so it was running rich and the cat was glowing.

I will warm up the bike again this weekend and look at the lambda sensor voltage.  I think that having replaced coils and plugs, I'm in a different running scheme than before, but haven't checked into it.  Coming soon.

[Jughead]

Try resetting the adaptations.

[panason1c]

- so it was running rich and the cat was glowing.

Could it be that the cat was glowing red because of unburnt fuel flowing into it and then igniting/burning .....due to the faulty coils?

[Honolulu]

Yes. 

Poor combustion => excess fuel in exhaust => converter doing its job burning excess fuel => lambda sensor reading 0.8V

Exactly why I replaced the coils, and spark plugs, and coolant, while I was in there.  This meant I had to acquire the vacuum tool to refill the coolant, and the proper 5/8" or 16mm plug socket (the one I had was too big).  It's not a difficult job in retrospect, but given the complexity of the machine overall, I'm a leery of having to work on it.

Now that the "work" is done, and it still runs poorly, if a idle actuator reset doesn't get it right, I'm afraid I'll have to talk to the local dealer.  The only one in 2,500 miles.  Whose parts tech doesn't return calls.  Bad sign...

[Gunleif]

I think the O2-sensor is the problem. Put in a new one. 

[Honolulu]

Gunleif:

Not one to argue, but 02 sensors aren't cheap, so why do you believe the sensor is bad?  It only has 30,000 miles, It "ought" to last much longer.  I think with it reading 0.8 volts, one coil bad, the cat glowing red, the sensor was correctly reporting a rich exhaust condition.

Is there a way, other than by substitution, to test an 02 sensor?  I've heard of putting the sensor tip in a propane flame (need to remove from bike first obviously) and using a DVOM to check voltage, but I've never done it.