K75 Misfire/Gratuitous Flames/Terrible Partial Throttle

[Laitch]

Everybody on par with that?

Not me. According to what you've written, it meets all the benchmarks for resistance along the temperature curve, so why isn't it working if its ground connection is clean? I think you should test it again.  :yes

[dowdogg]

Thats reasonable. I'll break it down tonight and do some cleaning of the sensor and its ground again to see. I called the dealership and they told me that they don't have any in the country. Their site says its NLA. Awesome. I'll order online from euromoto probably, if this cleaning and check falls through.

[Martin]

Do not put thread tape or sealant on the threads, as doing so will insulate the sender from earth. The sender relies on the gasket ring to prevent leakage.
Regards Martin.

[riots100]

My coolant temp sensor tested ok outside the bike, but on the bike, the fan would never come on leading to the coolant boil over you mentioned.  Replaced with a new aftermarket one, fan comes on  now.  I believe there is a thread some where here about getting aftermarket coolant temp sensors - I can't remember for sure though.

[Martin]

You can get them through Aliexpress but I've seen reports of them failing. But they are cheap.
Regards Martin.

[dowdogg]

Not me. According to what you've written, it meets all the benchmarks for resistance along the temperature curve, so why isn't it working if its ground connection is clean? I think you should test it again.  :yes

Well, results are in, temp sensor tested at 2964 ohm cold in 52 degree weather. 307 ohm hot (increasing in quick .1 increments with every second it was off). Temp sensor seems to be working fine, however, I ran this test with the OEM temp relay and of course, the fan never came on during idle or riding. Both the connector and sensor cleaned and scrubbed with deoxit. Also, I ran the bike with 3 different exhaust set ups (including stock) to see if there was any change in the misfire. The problem persisted. What now guys  :dunno ? I can buy a new temp sensor, but I would rather keep removing variables if possible.

[rbm]

Sounds like the sensor might be ok.  To be specific, how cold is cold? 17°C?  How hot is hot? 97°C?  There are two NTCs in the sensor but you published only one pair of readings; what are the corresponding readings for the other sensor?

[rbm]

YESTERDAY:

More troubleshooting. Set the FI according to this guide (http://skylands.ibmwr.org/tom/tech/co-setting.html) assuming that my 4 hole injectors were ruining the AFR. Misfire still present but I managed to make the bike idle nice and low, maybe too low?

Correct RPM at idle is 950 +/- 50

Side note: throttle got stuck at WOT for around a second last night after this adjustment, happened the morning as well, so I'm staying off WOT. Only seems to happen at 100% throttle, not 99%. Might adjust the AFR again to previous setting for safety

If you replaced the OEM Oetiker clamps with band clamps around the TB bushings, then double check to make sure that the throttle actuator is not interfering with the screws in the band clamps at WOT.  There are specific orientations for these clamps in this area of the engine and very little tolerance for deviation.



Also, make sure the TPS is not binding the TB actuator throughout its rotation.  Loosen the screws holding the TPS and operate the throttle several times to ensure there is no binding. Adjust the TPS if you detect binding.  Retighten the screws once this operation is completed and test one more time.

[Laitch]

1. I'm assuming you are speaking about this test (http://www.ibmwr.org/ktech/fan-diagnosis.shtml)

Nope. I was referring to Bert Vogel's boiling water test for the coolant temperature sensor in his troubleshooting guide downloadable from this site.

Well, results are in, coolant temp sensor tested at 2964 ohm cold in 52 degree weather. 307 ohm hot . . .

I can't follow your testing method. Perhaps the giddiness of the season is affecting my reasoning. If your coolant were 52ºF then it would seem to me that your cold value is off by 1000 ohms. Is this the test using boiled water? If it is, why would you be using air temperature as a value here?

Consider the tests in this post and the ones following it in reference to your temp relay function.

Anyway, I'm confident you'll sort this out and admire your ability to gauge your speed without a speedometer.  :2thumbup: Not everybody can do that.

[DavidATL]

Did you measure the temp sender resistance through the wiring to the computer or at the back of the sender? If there is a break in the wiring, it would create an open and perhaps the computer thinks the engine is always really cold. A faulty wire or connector pin would explain the temp sender measuring correct and the overly rich mixture that results in the backfire. I suppose a break could also be inside the computer but that seems unlikely.

[Martin]

Have you checked the valve clearances? Have you checked the compression? And spark plug colour is a good indicator to what is going on. Read up on doing a plug chop and report on the spark plugs colour, before and after doing the chop.
Regards Martin.

[dowdogg]

Nope. I was referring to Bert Vogel's boiling water test for the coolant temperature sensor in his troubleshooting guide downloadable from this site.I can't follow your testing method. Perhaps the giddiness of the season is affecting my reasoning. If your coolant were 52ºF then it would seem to me that your cold value is off by 1000 ohms. Is this the test using boiled water? If it is, why would you be using air temperature as a value here?

Consider the tests in this post and the ones following it in reference to your temp relay function.

Anyway, I'm confident you'll sort this out and admire your ability to gauge your speed without a speedometer.  :2thumbup: Not everybody can do that.

New results are in!I proceeded with the Bert Vogel's out of bike water temp test (boiling water test). I was using an in-bike test before.

Baseline: 5000 ohm (cold)

55C: 2,500 ohm

67C: 1,000 ~ 1,400 ohm (keeps dropping, hard to get an accurate reading)

77C: 1000 ~ 850 ohm

84C: 730~750 ohm

99C: 580~600 ohm

I'm not sure what the margin of error is for this test, so please advise.

Have you checked the valve clearances? Have you checked the compression? And spark plug colour is a good indicator to what is going on. Read up on doing a plug chop and report on the spark plugs colour, before and after doing the chop.
Regards Martin.

I checked my valve clearances and exhaust and intake were both in spec. Plug color was good. Compression was fine.

Did you measure the temp sender resistance through the wiring to the computer or at the back of the sender? If there is a break in the wiring, it would create an open and perhaps the computer thinks the engine is always really cold. A faulty wire or connector pin would explain the temp sender measuring correct and the overly rich mixture that results in the backfire. I suppose a break could also be inside the computer but that seems unlikely.

At this point, I have checked both.

[The Mighty Gryphon]

Judging from your resistance readings, your temperature sensor is reading low temperature.  The jetronic never knows that the engine is up to operating temperature.  This could be causing a rich mixture.  However, you say the plug color is good.    :dunno2:

[rbm]

These are the values it should read at the given temps:

Baseline: 2500 ohm (if cold means room temp)
55C: 750 ohm
67C: 500 ohm
77C: 400 ohm
84C: 300 ohm
99C: 250 ohm


So, yours looks to be stuffed.

[dowdogg]

Judging from your resistance readings, your temperature sensor is reading low temperature.  The jetronic never knows that the engine is up to operating temperature.  This could be causing a rich mixture.  However, you say the plug color is good.    :dunno2:

I must state the fact that i just replaced the plugs less than 500 miles ago. So I wouldn't be surprised if they look fine regardless of their actual condition.

So, yours looks to be stuffed.

Yeah, seems to be the case, I just ordered a new one via euromoto. Will update once I get it in.

[Laitch]

I must state the fact that i just replaced the plugs less than 500 miles ago. So I wouldn't be surprised if they look fine regardless of their actual condition.

:hehehe

[Martin]

When my temperature switch connector played up it resulted in over fueling. The glowing headers and lack of excess fuel which you would have seen when you pulled the plugs suggest the opposite. Suggests faulty temperature switch rather than the connector which has now shown up the resistance temperature test.
Regards Martin.

[Inge K.]

So it seems that general consensus is that the water temp sensor needs to be replaced. Everybody on par with that?

Yes, if you have measured the resistance through each element to the housing, and not both elements in serie.

[dowdogg]

MORE UPDATES!

I have replaced the sensor, installed everything on the bike and I took it out for a day. The results were... interesting to say the least.

• Firstly, now the bike suddenly has trouble starting cold (a problem that I have never had). If I don't fuel advance at least one click from the get go, it will start up and promptly die and won't want to start again. Once warm, the bike has no problem starting up and it idles perfectly right away.
• As for the misfire, it is (partially) fixed. The bike will still misfire as seen in the video and shoot flames etc as it did before at partial throttle. HOWEVER, when cruising on the freeway, it still sounds like its misfiring a bit (hard to hear with wind, helmet, etc), but no more flames or glowing red hot header collector anymore. So during cruising, the problem went away (from what I can see visually), but when going through the gears and holding any rpm for too long at partial throttle, it misfires terribly just like before.

To be honest, the results are even more dumbfounding than I anticipated.

[Martin]

Even in Queensland unless it is reasonably warm I have to start on first click of the fast idle. Check for air leaks, check the "Z" hose, check the vacuum caps.
Regards Martin.

[dowdogg]

Even in Queensland unless it is reasonably warm I have to start on first click of the fast idle. Check for air leaks, check the "Z" hose, check the vacuum caps.
Regards Martin.

Crankcase breather and vacuum caps have both less than 5000 miles, all vacuum lines as well. I’ll check again, but not banking on it.

[DavidATL]

   [li]Firstly, now the bike suddenly has trouble starting cold (a problem that I have never had). If I don't fuel advance at least one click from the get go, it will start up and promptly die and won't want to start again. Once warm, the bike has no problem starting up and it idles perfectly right away. [/li]
[/list]



I don't think this is a problem but is instead normal operation. If the bike was running rich before (and therefore backfiring from unburnt fuel), it's possible that it would start at ambient temperature without the idle increase lever locked on.

[Martin]

If it was running rich it would start without having to advance the fast idle. Read up on how to do a plug chop. Take the bike up to where it is miss firing and shut her down as per the plug chop instructions. You might have to do it a couple of times to get the hang of it. Do this on a quiet road somewhere where you can push it safely to a place where you can work on i it. A back up vehicle is handy to act as a buffer. Hopefully reading the plugs after doing this will give some indication of what is going on. What exhaust are you currently running? Have you used DeoxIT on the ECU and FICU.  I have just bought the new Deoxit which now comes in two cans, cleaner and a gold coating can. I had run out of the old DeoxIT of which I only had the one can of cleaner. It was given to me many year ago so I'm not sure whether I was missing a can. Even the new DeoxIT failed to fix my 4 pin fuel pump connector that had only ever had one problem in twenty years.
Regards Martin.

[dowdogg]

If it was running rich it would start without having to advance the fast idle. Read up on how to do a plug chop. Take the bike up to where it is miss firing and shut her down as per the plug chop instructions. You might have to do it a couple of times to get the hang of it. Do this on a quiet road somewhere where you can push it safely to a place where you can work on i it. A back up vehicle is handy to act as a buffer. Hopefully reading the plugs after doing this will give some indication of what is going on. What exhaust are you currently running? Have you used DeoxIT on the ECU and FICU.  I have just bought the new Deoxit which now comes in two cans, cleaner and a gold coating can. I had run out of the old DeoxIT of which I only had the one can of cleaner. It was given to me many year ago so I'm not sure whether I was missing a can. Even the new DeoxIT failed to fix my 4 pin fuel pump connector that had only ever had one problem in twenty years.
Regards Martin.

Sorry for the late reply Martin,

Yes, I have used Deoxit on both the ECU and FICU. This exhaust is custom-made by Iron Cobra Fabrication in Long Beach, CA. It uses a 18" Cone Engineering muffler. If you want more information on it, here is a catalog of their pre-fabricated products, http://www.coneeng.com/pdf/motorcycle/MC-finished-Mufflers.pdf. I did a plug chop earlier, the color of the plug seemed fine. I also just install brand new NGK plug wires. As of recent after installing the new water temp sensor, the bike runs great once warm, no misfires at cruising anymore. However, it really struggles with start up, I need to turn the fuel advance up all the way and hold down the ignition button for a few seconds for it able to maintain idle.

[Martin]

It probably needs to have the mixture set.  The next test is a really cool feature on 2V bricks. If you have the bike idling at 1000rpm once warmed up, press the green starter button. If the mixture is correct the revs need to stay the same or rise slightly, if they rise significantly or drop the mixture needs adjusting. Either get access to a exhaust gas analyser or look up lean drop method on this site.
Regards Martin.