Engine behaviour.

[akima]

I started my bike this evening and as normal it chose to put the revs at about 2300RPM. It was warming up for about 30 seconds when I decided to rev the engine a very small amount, bringing it up maybe 1 or 2 thousand revs extra: the engine then cut out. I've never had that happen before. The bike was in neutral at the time. The temperature was about 12°C.

When I started the bike again it brought itself up to ~3000RPM, which is higher than I've seen it go at idle before.

Everything returned to normal after that.

Any ideas why it cut out when I revved it?
Any ideas why the first time I turned it on the revs were quite low at 2300 and then the second time they were up at 3000? If anything I'd expect it to be the other way round: first time at 3000, the engine warms a bit, then second time at 2300.

I'm not worried about the bike, I'd just like to try and understand it a bit better. I think I have a long way to go

[jdoucet93]

I think just about all of us have this issue with our carbureted bikes. They just really don't like starting in the cold. But being that yours is fuel injected... I have no idea.

Other than possibly a bubble in the fuel line, I'm not sure.

[greg737]

This is your first winter with the bike, right?

Warm-up is a weird period of engine operation, even for a fuel injected bike and especially when it starts to get colder.

If you've got a few moments to read, I'll try to partially it with my do-it-yourself system and relate it to your OEM FI EX-250. Warning: I've found that short answers are hard to come by in the fuel injection world.

On my do-it-yourself fuel injection system I've got a wideband O2 sensor installed with a readout gauge on the instrument cluster. This gives me a pretty good picture of what's going on in the engine at any moment and I can tell you that cold start-up and warm-up is a pretty strange situation.

In a cold start-up I'll switch the bike's key to "on" and in the few moments between doing that and starting it up the wideband O2 controller will send power to a heating element in the sensor (widebands are electrically heated when their controller senses less than optimum operating temperature at the sensor). After a few moments the gauge on the instrument cluster will begin indicating a steady 21.8 which is atmospheric O2 content (and is normal to find in the exhaust pipe of a bike that hasn't been running for a while).

The ECU in my project bike is fully programmable from my laptop computer. I had to spend a pretty good amount of time getting all the settings/entries for all of the fuel delivery contingencies configured so I have a pretty good mental picture of what's going on at any given moment of engine operation. I know that when I start the bike in cold weather there's a lot of fuel being injected into the engine (the amount is: the normal fuel commanded for the given RPM and throttle position + an amount of warm-up enrichment fuel based on the temperature of the coolant temperature sensor). It's getting enough fuel to be "rich".

(We know that this is the equivalent of opening the choke lever on a carbed bike.)

So on a cold morning when I hit the starter on the bike it fires right up and settles at just above normal idle speed. You would expect the O2 sensor gauge readout to show a rich air/fuel mixture, but initially it doesn't. Instead right after the engine begins turning over it shows a lean air/fuel mixture in spite of the fact it's injecting enough fuel to be really rich. Where's all that fuel going in the first few moments of cold engine operation? It's sticking to all the cold metal in the throttlebody, the engine head and on the stems and backsides of valves themselves. The fuel injection engineering types call this effect "puddling".

For about a minute (on a moderately cold start-up) and up to 2 minutes (on a really cold start-up) the engine is getting messed up feedback from its O2 sensor (telling the ECU that things are lean when it knows it's putting in enough fuel to be rich) and the engine is operating in the delicate situation of having a lot of extra fuel clinging to the walls of its intake tract.

When you blipped the throttle you almost certainly killed it with too much fuel (too rich). Because when you rapidly moved the throttle (blip) the TPS (throttle position sensor) on the bike's throttlebody sent this new (much higher) throttle position report to the bike's ECU which used this input (combined with RPM and Manifold Absolute Pressure) to look up the proper fuel from the fueling table it has in it's memory.... and deliver even more fuel to an already rich situation. (The ECU's basic fueling table doesn't reference engine temperature).

Side note: You might have even triggered what's called "acceleration enrichment" which is based on TPSdot (Throttle Position Sensor rate-of-change). Acceleration enrichment schedules in extra fuel to prevent the bike from going momentarily lean when you make rapid throttle changes (increases). If accel-enrich wasn't there you'd feel a little "bogging" or momentary hesitation when you rolled on power in a really hard acceleration. This would have thrown in even more fuel.

So, in that instant when you blipped the throttle, when: A. the bike's ECU was already mildly confused by its O2 sensor feedback, and B. just when the delicate situation of cold start fuel "puddling" was at its worst (hanging onto the walls of the bike's intake tract, just waiting for more airflow and vibration to cause it to shake loose and enter the combustion chambers), and C. when the engine was having to work against the drag of cold, thick oil.... your blip told the ECU that you wanted even more fuel and the engine went too rich and stalled.

I'm betting it started right up again, didn't it?

Fuel injection folks like to believe that FI is pure magic on cold starts but while it truly is quite an improvement, it has it's limitations.

[akima]

@greg737: That's the most amazing reply I've had on this forum! You totally get 3 nuts for that!!

You explained that very clearly and it's given me a much better idea about what was going on in the engine. Your description certainly seemed to correlate with the actual engine-stall event.

I actually presumed that the cause of the stall was along those lines. Though I was thinking in far more basic terms. I imagined the FI computer has rules that define the behavior of the FI system and sensors which help it make decisions. I made the presumption that either the program that was running on the FI computer was faulty/incomplete and didn't have a reasonable procedure in place to deal with the specific set of variables which occurred during that start-up or that the sensors were not providing the adequate or accurate information to the FI computer it needed to make sensible decisions about how much fuel to pump into the engine.

I'd love to extract all your knowledge from your mind Understanding the FI system is way more interesting to me than understanding any other part of the engine... though I imagine it's so intrinsically linked to the rest of the bike that I'd need to understand most of the bike in order to perfectly understand the FI system

To answer your questions...

Originally Posted by greg737

This is your first winter with the bike, right?

Yep

Originally Posted by greg737

I'm betting it started right up again, didn't it?

Yep... though I had to hold in the starter button about a second longer than the previous start-up.

I'd be interested to hear your theory about why the bike then jumped straight up to 3000PM the second time I started it. I've never seen it rev that high before.

Thanks again for your awesome reply!

[greg737]

Glad to help.

[setasai]

Originally Posted by akima

@greg737: That's the most amazing reply I've had on this forum! You totally get 3 nuts for that!!

I feel like 3 nuts is 1 too many...

[Alex]

What did we do before Google image search?



Forums must have been much less fun.

[Jiggles]

Originally Posted by Alex

What did we do before Google image search?



Forums must have been much less fun.

Link to original page on YouTube.

[massacremasses]

Originally Posted by Alex

What did we do before Google image search?


Forums must have been much less fun.

QFT!

where else can you find stuff like this....

[Jiggles]

Originally Posted by massacremasses

QFT!

where else can you find stuff like this....

Or this?



Or this?



Or this?

[massacremasses]

I cant actually post the picture I wanted too.... @Jiggles. you got PM cause I need to share it... HAHAHAHA

[Jiggles]

Originally Posted by massacremasses

I cant actually post the picture I wanted too.... @Jiggles. you got PM cause I need to share it... HAHAHAHA

OMG post it!

@Alex we can post pictures of nuts right? theres no rule against that?

[massacremasses]

hahahaha dude how great was that?! It took me a second to realize how that was related to my search...

[Jiggles]

[massacremasses]

@akima will never get mad if I post pictures like this....



but that is more for @Alex... hahaha

[akima]

Originally Posted by massacremasses

@akima will never get mad if I post pictures like this....



but that is more for @Alex... hahaha

[akima]

I now think the strange behaviour I mentioned in my first post may be a result of my FI system being faulty. This morning I couldn't get my bike to start. When the engine did seem like it was taking and was about to come to life for some reason it only reved up to 1k and then conked out. I spoke to a guy at the Kawi garage and after explaining the symptons he also independently came to the conclusion that my FI system might be faulty. If it is, it'll be covered under warranty which is good.

[Jiggles]

Originally Posted by akima

I now think the strange behaviour I mentioned in my first post may be a result of my FI system being faulty. This morning I couldn't get my bike to start. When the engine did seem like it was taking and was about to come to life for some reason it only reved up to 1k and then conked out. I spoke to a guy at the Kawi garage and after explaining the symptons he also independently came to the conclusion that my FI system might be faulty. If it is, it'll be covered under warranty which is good.

Hmm, I wonder what would happen if you bump start it?

[akima]

I'd be scared of dropping the bike while running along with it. There aren't any hills near by.

[TBC]

Originally Posted by akima

I'd be scared of dropping the bike while running along with it. There aren't any hills near by.

If it's flat enough, you should definitely be able to get it up to a good speed while on the bike

[Jiggles]

Originally Posted by TBC

If it's flat enough, you should definitely be able to get it up to a good speed while on the bike

I've tried it without a hill and it's pretty difficult.

[alex.s]

it sounds like the o2 sensor has gunk on it

or maybe you are blipping it too hard

have you tried starting it with WOT?

[ForceofWill]

What do the stock EFI systems use for A/F? Do they have a MAF on the intake? If they do, and that went out, it would cause symptoms that you're describing. I would also think if they use a wideband o2, if the o2 sensor went bad.

I doubt the Ninja is sophisticated enough to have a base map to revert to in the event it can't get a reading, I just am not sure what they use to meter.

[akima]

Originally Posted by alex.s

it sounds like the o2 sensor has gunk on it

or maybe you are blipping it too hard

have you tried starting it with WOT?

I tried it once with the throttle wide open and it didn't help.

[greg737]

Doesn't the Fuel Injected EX-250 have an FI fault light? If the FI system has some sort of abnormal input or output condition it will annunciate with the light and store the error code in the ECU's memory for retrieval by a service tech guy at the dealership.

So if the FI fault light isn't on then the FI system is probably operating correctly. In other words, it's getting good signals from all its peripherals (O2 sensor, throttle position sensor, battery voltage, coolant temperature sensor, air temperature sensor, manifold absolute pressure sensor, etc.) and thinks it's delivering the correct amount of fuel at the correct time. Otherwise it would throw a code and light the fault light.

In the absence of a fault light I'd bet on something thats preventing fuel from reaching the FI system in the first place, such as accidently leaving the petcock closed or a clogged fuel filter.

That's what a stumbling, almost-start sounds like to me.

[greg737]

Originally Posted by ForceofWill

What do the stock EFI systems use for A/F? Do they have a MAF on the intake? If they do, and that went out, it would cause symptoms that you're describing. I would also think if they use a wideband o2, if the o2 sensor went bad.

I doubt the Ninja is sophisticated enough to have a base map to revert to in the event it can't get a reading, I just am not sure what they use to meter.

The EX-250 Fuel Injected model does not have a MAF (hot-wire mass airflow sensor). It also does not have a wide-band O2 sensor, it uses a narrow-band O2 sensor.

If the EX-250 Fuel Injected model's ECU lost continuity with its O2 sensor it would light its FI fault light.

Actually, a factory pre-determined basic fueling map (a volumetric efficiency table) is exactly what the EX-250 fuel injected model uses. A pre-determined table is simpler than anything else you might do with things like a MAF or wideband O2 sensors.

[ForceofWill]

Originally Posted by greg737

The EX-250 Fuel Injected model does not have a MAF (hot-wire mass airflow sensor). It also does not have a wide-band O2 sensor, it uses a narrow-band O2 sensor.

If the EX-250 Fuel Injected model's ECU lost continuity with its O2 sensor it would light its FI fault light.

Actually, a factory pre-determined basic fueling map (a volumetric efficiency table) is exactly what the EX-250 fuel injected model uses. A pre-determined table is simpler than anything else you might do with things like a MAF or wideband O2 sensors.

Good info, thanks. I know my audi has a MAF but if it goes bad, will revert to a basic map. My old supra just wouldn't run with no input from the MAF.

[alex.s]

do they have a backflow o2 sensor or do they put it in the intake?

my money is on some crap on the o2 sensor

[greg737]

Originally Posted by alex.s

do they have a backflow o2 sensor or do they put it in the intake?

my money is on some crap on the o2 sensor

I think it's all ball bearings now.

Willy: What the hell you need ball bearings for?
Fletch: Awww, come on guys, it's so simple. Maybe you need a refresher course.
[leans arm on hot engine part]
Fletch: Hey! It's all ball bearings nowadays. Now you prepare that Fetzer valve with some 3-in-1 oil and some gauze pads. And I'm gonna need 'bout ten quarts of anti-freeze, preferably Prestone. No, no make that Quaker State.

[greg737]

Akima,

Thinking more about your startup problem. Once again, if there's no FI warning light then you have to assume it's operating properly. If that's true you've got to determine if the stumbling-failure-to-start behavior is because of a too-rich or too-lean condition.

I know the FI EX-250 has an idle speed adjustment knob (because I've got the FI EX-250 throttlebody in my project bike). It's the long floppy adjuster with the black knob that dangles from the left hand side of the throttlebody. I was thinking about your startup problem and I was wondering if you had tried adjusting this knob at all?

If you haven't touched it before now I'd suggest about one complete turn to the right (right turns gradually open up the butterflys and left closes them). Of course, you could confirm the direction to turn in your owners manual just to be sure. If the problem with your cold weather start is a too-rich situation then this would introduce more air into the air/fuel ratio and that would help it get "over the hump" of startup and into a sustainable idle.

[akima]

Thanks for your input Greg.

I have seen the FI light come on briefly after the engine has started. Come to think of it, it has only started happening recently. The light goes out within seconds if it comes on at all. I assumed (probably stupidly!) that because the light went out quite quickly after starting the engine, that the FI system was fine and just had a moment of cold-weather-induced confusion.

The FI light definitely wasn't coming on earlier in the year.

I wont play with the idle speed knob right now, but thanks for making me aware of it.

[akima]

The bike actually started this morning. The FL light was on for a good 5 seconds after I started the engine and the revs stayed way down at 1k for ages. It almost stalled when I was riding and came to a stop at some lights. After riding for 5 mins the bike was fine and seemed to be behaving normally.

I've left the bike with some clued up Kawasaki mechanics now so they'll find the fault and fix it for me. I'll post the details when it's fixed.

I'm also getting the mechanic to see if I have that manufacturing mistake where they failed to grease the bearings in the hub. My bike squeaks! It could just be the brake pads though.

If I really have had 2 manufacturing issues am I unlucky? Or do a lot of you guys get problems with your new bikes?

[greg737]

Definitely temperature related then. Sounds like you've got a too-lean condition during cold starts.

When it got cold here in Spokane and I started working on tuning for cold startups I had a very similar experience (the engine "hanging" at very low RPMs on a cold start). As I slowly adjusted the priming pulse fuel, the cranking pulse fuel and the after-start enrichment I got the bike to a point where it was rich enough to get it "over the hump" and into the normal idle RPM range even when very cold.

If it comes back all "fixed" from the mechanics and still has a hard time on cold-startups, here's something you could do:
I know you don't have the capability to re-program, but you can slightly manipulate the ECU's behavior for a cold start. The way you would do this is to simply key-switch the bike slowly "on" and then "off" again a couple (maybe three) times before hitting the starter button. What the bike does every time you turn the key-switch to "on" is it brings the fuel pressure up to operating PSI and then it fires a "priming pulse" through both injectors. This pulse is meant to make sure there's fuel ready for a start (no air bubbles at the injector tips). Doing it a couple or three times will simply set up a richer air/fuel state in the cylinders for when you do hit the starter button.

Try it next time you have to do a cold start. Let us know how it goes.

[akima]

Originally Posted by greg737

Definitely temperature related then. Sounds like you've got a too-lean condition during cold starts.

When it got cold here in Spokane and I started working on tuning for cold startups I had a very similar experience (the engine "hanging" at very low RPMs on a cold start). As I slowly adjusted the priming pulse fuel, the cranking pulse fuel and the after-start enrichment I got the bike to a point where it was rich enough to get it "over the hump" and into the normal idle RPM range even when very cold.

If it comes back all "fixed" from the mechanics and still has a hard time on cold-startups, here's something you could do:
I know you don't have the capability to re-program, but you can slightly manipulate the ECU's behavior for a cold start. The way you would do this is to simply key-switch the bike slowly "on" and then "off" again a couple (maybe three) times before hitting the starter button. What the bike does every time you turn the key-switch to "on" is it brings the fuel pressure up to operating PSI and then it fires a "priming pulse" through both injectors. This pulse is meant to make sure there's fuel ready for a start (no air bubbles at the injector tips). Doing it a couple or three times will simply set up a richer air/fuel state in the cylinders for when you do hit the starter button.

Try it next time you have to do a cold start. Let us know how it goes.

You're awesome Love this little hack! I'll give it a try.

I always wondered what those mechanical noises were that I could hear after I turned the key to on but before I started the engine. I guess it's the sounds made by the fuel pressure adjustment and priming pulse that you referred too.

Interesting side thought, if I leave the bike in gear and turn the key to on, the bike doesn't make the full set of noises that it does when I flick the key to the on position while the bike is in neutral. I always park my bike in gear but make sure I I flick it to neutral before switching the key to the 'on' position... I figure that when I leave the bike in gear and it misses some noises it's missing an important step! Maybe it doesn't do the priming pulse. Who knows?

When you said "I know you don't have the capability to re-program", did you mean me personally not having the ability or did you mean that there are currently no software and hardware tools available to re-program the original Kawasaki EFI chip? Or maybe that the chip isn't even reprogrammable?

[greg737]

When you said "I know you don't have the capability to re-program", did you mean me personally not having the ability or did you mean that there are currently no software and hardware tools available to re-program the original Kawasaki EFI chip? Or maybe that the chip isn't even reprogrammable?

I was just approaching your issue of difficult cold startup from my perspective of a fully user-programmable FI system. I was asking myself, "How could she manipulate the hard-programmed OEM Kawasaki system?" I knew that I would address a too-lean startup by re-programming for a richer start but you can't re-program, so I came up with the work-around I described. It will be interesting to hear if it makes a difference in your cold startup.

Interesting side thought, if I leave the bike in gear and turn the key to on, the bike doesn't make the full set of noises that it does when I flick the key to the on position while the bike is in neutral. I always park my bike in gear but make sure I I flick it to neutral before switching the key to the 'on' position... I figure that when I leave the bike in gear and it misses some noises it's missing an important step! Maybe it doesn't do the priming pulse. Who knows?

That's odd that it doesn't make all the same sounds just because of the neutral switch, but you could be onto something. You could try it both ways to see if there's a difference in startup characteristics.

[akima]

Ok! I dropped into the bike garage on the way home and spoke to the mechanic. There is good news and bad news

Bad News: An expensive part of the FI system in the throttle body is broken. It costs ~£400 to replace.

Good News: Kawasaki is paying for it

They said that they hooked up a Kawasaki tool to the on-bike computer and it told them roughly where the problem is. Apparently there was too much electrical resistance (7ohm on one side and 8ohm on another) in this throttle body area. I think they said the resistance should be down at ~5ohm.

I don't really understand what I just typed. I just tried to put in the key facts that the mechanic told me. If they weren't closing the shop I would have asked for more information so I could understand it better. Anyone got any idea what I/they were going on about? I know what electrical resistance is: I can understand why having too much electrical resistance could be bad... not enough power will be being delivered to something that needs it.

[TBC]

I read the first sentence of your post like this:

Originally Posted by akima

Ok! I dropped ...the bike

Took me a few seconds to read it properly .

[greg737]

They said that they hooked up a Kawasaki tool to the on-bike computer and it told them roughly where the problem is. Apparently there was too much electrical resistance (7ohm on one side and 8ohm on another) in this throttle body area. I think they said the resistance should be down at ~5ohm.

I don't really understand what I just typed. I just tried to put in the key facts that the mechanic told me. If they weren't closing the shop I would have asked for more information so I could understand it better. Anyone got any idea what I/they were going on about? I know what electrical resistance is: I can understand why having too much electrical resistance could be bad... not enough power will be being delivered to something that needs it.

They're probably telling you that the TPS (Throttle Position Sensor) is messed up.

There are only two electronic components on the throttlebody: 1. the Throttle Position Sensor, and 2. the servomotor controller for the secondary butterfly assembly.

The TPS is very simple to replace. Two minutes tops, with a "star drive" type screwdriver. Adjusting it might take a few more minutes.

A TPS is just a simple component that delivers a predictable series of voltage outputs, stepping higher as it is rotated throughout its range of motion. And, of course, the thing that rotates the TPS is your wrist working the throttle.

The secondary butterfly servo is a bit more complex (read as: more expensive) as it contains the servomotor, a controller board and its own TPS for feedback to the FI ECU.


In this picture the TPS is the connector (made of mostly black plastic with a bit of gray) at the far left end, in front, closer to my thumb.

The secondary butterfly servomotor/controller/position-sensor combo is behind the TPS (the larger black plastic box+connector, closer to my index finger). The large white plastic connector at the top of the picture provides power, ground, and control inputs for the secondary butterfly assembly, and the connection at the left end (and top) of the servomotor box is for the secondary butterfly postion sensor.

With regard to your startup problem, I'd guess that the TPS is more likely to cause the problem because it's one of the main influences on the ECU's fueling decisions (along with RPM and manifold absolute pressure readings).

(EDIT) Although, now that I've considered it for a few more moments it could be related to the secondary butterfly assembly servomotor because at startup the servomotor acts as part of the bike's "automatic choke" system. For startup the ECU commands the secondary butterfly servomotor to rotate all the way around to "full-open" position which brings a small metal "cam" into contact with an arm that extends off the back of the throttle pulley. When this cam contacts the arm it causes the throttle to open slightly. In addition to this action the ECU also commands the correct amount of fuel, taking into account temperature and elevation (it figures out elevation by taking a "reference" reading on the Manifold Absolute Pressure sensor before the engine is cranked over. That's just another of those "pre-start" items that occur every time you key-switch the bike to "on")

A picture is worth a thousand words:


In this picture of the left-hand side of the throttlebody you can see the throttle pulley (the big round thing that the throttle cables connect to) with the little arm extending backwards off of it. Just below the arm is the small silver metal cam (with 0004 printed on it) that is attached to the end of the secondary butterfly shaft.


In this rear-view of the throttlebody you can see that the secondary butterfly assembly is fully closed and that the small silver cam is as far away from the throttle arm as it can get.


In this picture you can see that the secondary butterfly assembly is now fully open and that the small silver cam is in contact with the thottle arm and it is very slightly lifting the arm. This opens the main throttle butterflies just a bit to admit more airflow for startup which, when combined with the extra startup fuel that the ECU is programming, should give you a reliable start whether it's cold or warm or hot.

How's that for an answer, akima? Too much information?

[akima]

Remember to put that back in your bike before you next ride

Thanks for your explanations and thoughts.

When more voltage is created as a result of twisting the throttle, how does that additional voltage ultimately produce a higher rpm? Does it speed up an electric motor that pumps fuel into the cylinders?

[nickjpass]

Fuel and air.

When you "open" throttle grip, you're pulling a cable, that opens a butterfly valve in the carburetor (depending on the type of carburetor). As the butterfly valve is opened, more air is allowed to be pulled into the cylinder which creates a vacuum that sucks more gasoline into the engine

Carbs 101