Alright guys and gals, Got on the T last week and she started stuttering at around 5000 rpms. Started troubleshooting the old girl when all of a sudden no spark. I've tested all components both statically and under power, everything looks good. As far as i know the crank sensor is the only thing that'll stop the ECU from firing the coils. I see a really strong square wave on my scope so i know that's working. I've checked all connections, battery, grounds, she turns over really strong. I get no error codes while trying to start except for once late last night, it was a stepper motor error, never got it again. Any suggestions?
Joined Oct 2009
254 Posts | 1+
north manchester
check your cam position sensor readings,if that aint working it wont spark because it doesnt know witch cylinder is at tdc
1. Check to see if you have a hot wire to the coil- Key on.
2. Do not square wave that crank sensor unless you waved it the first day and know that same scale meets and greets the [same line]. If not, she is out of spec at the coil windings not that it can't hall effect and bump your scale some. That is a book spec read and to read it, you know breakout, right? If it breaks out of that number [by a millionth even it's] = JUNK! Break out is just that. It broke out of that number toss it.
3. Cam sensor is like what CT said, but we have to understand something about the cam sensor or you'll chase your tail with ct. This is theory so you are not going after page after page of guessing:
a. The cam sensor puts out that same wave or say one number on the cam sweep.
b. The ECM is setup like sequence fire. I can use the abstract by simply using a 1 and 0. Now it gets complicated.
c. Take your pick which is front/rear is OK, 1-0. So, here is the fire sequence 1-0-1-0-1-0-1 until you turn the key off.
d. The 1 is saved so 0 fires next. The 0 is saved, the 1 is fired next who cares, this is theory. Here comes more complex.
e. Take your pick who fires next is that abstract it runs in the absolute yes or no? With that said, here comes the crank.
f. That f for fire is while you are spinning the crank, that crank has a tone wheel or a number of cuts in the wheel so it can make a count per 360°.
g. That part is g for good known part if that books specs within bounds for that crank sensor. In other words, your crank sensor can start the recount to light the bike off if it is [NOT] the cam sensor. Narrow it down. One can fire and one can't you swap both around.
h. Then again, that crank sensor is all about a good book spec and now eliminate it.
i. That ECU saves the last good known [next 1-0-1-] number. Turn the key on, she fires for the last time it's the cam sensor. Meaning, you turn the key off, the cam sends back zip. You ran on the last good signal [the ECU saved]. Why did you turn the key off! LOL-Got it?! The ECU remembers -0-0-0-0-0-0-0-0- and there you go, the cam sensor.
j. The joke is on you, you keep throwing parts at it and not the two prongs off the ohm meter at both cam and crank sensors to know you need to eliminate the other.
By now, you should be up to 9-amigos that send you to the side of the road:
1. Fuel
2. Spark
3. Compression
4. Connector not connected
5. Wire out of connector.
6. Short to ground.
7. Cam sensor.
8. Crank sensor.
9. ECU = You now own a boat anchor. :lol
Like ct says, you need to get out more [in the garage] and 9-apply that puppy. :dunno
2. Do not square wave that crank sensor unless you waved it the first day and know that same scale meets and greets the [same line]. If not, she is out of spec at the coil windings not that it can't hall effect and bump your scale some. That is a book spec read and to read it, you know breakout, right? If it breaks out of that number [by a millionth even it's] = JUNK! Break out is just that. It broke out of that number toss it.
3. Cam sensor is like what CT said, but we have to understand something about the cam sensor or you'll chase your tail with ct. This is theory so you are not going after page after page of guessing:
a. The cam sensor puts out that same wave or say one number on the cam sweep.
b. The ECM is setup like sequence fire. I can use the abstract by simply using a 1 and 0. Now it gets complicated.
c. Take your pick which is front/rear is OK, 1-0. So, here is the fire sequence 1-0-1-0-1-0-1 until you turn the key off.
d. The 1 is saved so 0 fires next. The 0 is saved, the 1 is fired next who cares, this is theory. Here comes more complex.
e. Take your pick who fires next is that abstract it runs in the absolute yes or no? With that said, here comes the crank.
f. That f for fire is while you are spinning the crank, that crank has a tone wheel or a number of cuts in the wheel so it can make a count per 360°.
g. That part is g for good known part if that books specs within bounds for that crank sensor. In other words, your crank sensor can start the recount to light the bike off if it is [NOT] the cam sensor. Narrow it down. One can fire and one can't you swap both around.
h. Then again, that crank sensor is all about a good book spec and now eliminate it.
i. That ECU saves the last good known [next 1-0-1-] number. Turn the key on, she fires for the last time it's the cam sensor. Meaning, you turn the key off, the cam sends back zip. You ran on the last good signal [the ECU saved]. Why did you turn the key off! LOL-Got it?! The ECU remembers -0-0-0-0-0-0-0-0- and there you go, the cam sensor.
j. The joke is on you, you keep throwing parts at it and not the two prongs off the ohm meter at both cam and crank sensors to know you need to eliminate the other.
By now, you should be up to 9-amigos that send you to the side of the road:
1. Fuel
2. Spark
3. Compression
4. Connector not connected
5. Wire out of connector.
6. Short to ground.
7. Cam sensor.
8. Crank sensor.
9. ECU = You now own a boat anchor. :lol
Like ct says, you need to get out more [in the garage] and 9-apply that puppy. :dunno
I do have a hot wire, key on, at the coil. I have also inspected and tested all the connections, all look good. Both the cam and crank sensors test within specs according to the manual. The cam sensor is a hall effect, and you're right, all i know is that i get a good square wave when i crank, what i don't know is if the amplitude and frequency are acceptable. The crank sensor sends a sine wave (ac voltage) to the ECM and that tests within specs on amplitude. I have had many hall effect sensors test good statically but not work properly (on cars). I've checked all i can, i think it's time to change the cam sensor.
1. Connector not connected: Here, I agree you are at wits end. But this is the trouble tree that speaks.
2. Wire not in connector: Here again, is the trouble tree, you checked connections, but did you check from the cam sensor back to the black box?
3. Short to ground: This too says, short is out of book spec. You are pinning tail on donkey. I say it is not #3 if it is book spec and your empirical knowledge on cars is noted.
It is not #1 because you said you checked all the connections. Watt you did not do was follow or did you also follow #2? That says I disconnect the cam/crank sensor, ran each wire back to the ECU's main jack that piggies over the black box. Though the connector is together, a male/female connect may be air hanging and not touching at all.
That too says, there is a wire drop from cam to ECU. I would try that last wit if you did not try it.
Joined Jul 2008
1K Posts | 26+
If it's a GenII, check your a/c V output from the stator, it should be about 60v a/c across the yellows.....a failed/failing stator will deffo cause the motor to stutter....
Amurai (One of the resident AF1 Yoda types) has concluded that the arcing causes a field effect that interferes with the ECU.
Amurai (One of the resident AF1 Yoda types) has concluded that the arcing causes a field effect that interferes with the ECU.
I would think any stator is going to give off some sort of...
WATTever, if that is his conclusion, I do not want any part of it. That is his abstract... Now for mine. I think this makes more sense rather I keep it open so you can conclude on your own;
The bike has a **** load of systems. Most are closed or loop closed so the CAN can step over every loophole and know who to code. The engineering comes as there is a balance in said wattage/resistance/pressure and now go read up on ohm's law, the 4th variable of that wheel, and you are off to building a toaster so it does not cook the house down.
Same goes with the output from the battery as in a wall socket. ECU lives a few doors down in J-block. They have their own ground. They sport the 0 to 5v window. You look out to that 12v window with the two yellow lines heating up not transferring that A-C is it comes back at itself, no? Kind of cooks at the yellow connector with a bad path.
That battery is the output balance the ECU demands. YOu lower that output and the 0 to 5v cannot work in a 0 to 3v balance, meaning, if the engineering demands 4.5v for said sensor and you sent in 3v? So, battery output is paramount to having that puppy all in a static 12.6-12.8v at the battery leads.
The yellow leads cannot keep up with demand to keep charging the battery [balance]... Not some abstract that reads; arching a field effect? You mean low voltage balance out the battery hear me stutter with 3v at the ready steady?
Last edited:
Wait for the main wire harness from sensor connector wire(s) to ECU connector wire(s). One loop at a time. The OP did say, 'she turns over really strong' says it is not a stator problem. Would you agree? The phantom drop without codes are fun, no?
Tested the cam and crank sensor back to the ECU, it all looks good. Tested the stator while i was at it, got 7vac with it disconnected while cranking. I wouldn't expect to get the full voltage while just cranking but 7vac seems a little low to me. Regardless, the bike should start with the stator disconnected and a good battery.
I would agree; you ran zero stator and a well charged battery is that loop not needing that stator in the, "Total Loss" type running ignition for racing purposes = Remove stator for weight and magnetic friction off the crank.
Technically, the engine fires 8 times for a 4 cylinder needing that ping signal to close the loop as it were. Watt are the odds, you do the math, remove 2 cylinders, the ECU is waiting for 4 spins to send the ping. If it fails, code 12 appears double looped explaining;
1. Engine rpm signal missing = This means the sensor no longer sends in said [specific value] input ~ 3 amigos dropped out.
2. Engine rpm signal = This means the crank may turn, but the hall effect is sitting there with a broken shaft ~ No spin.
WATT turtle kept saying was in another thread, another brand computer bike; the OP said, "I changed wire harness, battery, this, that; and the crank sensor was, 'close enough" was his say so. But close enough is not book spec. After the second wire harness, I said; the crank sensor is not some, 'close enough' to start a computer engine. Go measure again, watt is the exact number? After who knows how many pages of parts changes, guy goes, "It was the crank sensor."
My point is; When book says this number is this number... No one [number from that manual] listens to turtle [the frog].
I'm just saying, frog legs does not measure close enough. Turtle wish is he was standing over your shoulder with a cold dark ale... Bringing that other set of eyes for ya. Get it nailedown, guy! It's going to 20/20 hindsight you it was so easy/obvious.
Bloody hell, this is like reading one of those threads in scottish talk, i recognise some of the words but it all makes little sense !!
Well the whole harness checks out, beginning to end, i've gone as far as i can possibly go without being able to directly read ECU data. I've tested all sensors both statically and in-situ with power, found everything working properly. I've pulled the flywheel and verified that the shear pin is not sheared. I'm beginning to believe that the ECU is toast. I've tried simulating errors, no go. The symptoms remain, no fire, no error codes.
Well, here how the bike turns into a boat anchor in 3 easy steps:
1. Crank sensor = No start if that sensor is out of spec.
2. Cam sensor = Has in memory (RAM) that one last good signal, then has no good know signal to restart.
3. ECU = If the two up top are good = Only other boat anchor is this amigo.
I like that through check of the sheered key. But next time, I would remove a spark plug and the dyno cap. If the piston comes up as you turn the rotor, you can run the rotor up to where it says TDC, then, if the piston is there too... Key is not sheered or has moved [if] rotor T (tdc) mark is in line with the piston at the top of the stroke and that static mark where all 3 line up.
1. Crank sensor = No start if that sensor is out of spec.
2. Cam sensor = Has in memory (RAM) that one last good signal, then has no good know signal to restart.
3. ECU = If the two up top are good = Only other boat anchor is this amigo.
I like that through check of the sheered key. But next time, I would remove a spark plug and the dyno cap. If the piston comes up as you turn the rotor, you can run the rotor up to where it says TDC, then, if the piston is there too... Key is not sheered or has moved [if] rotor T (tdc) mark is in line with the piston at the top of the stroke and that static mark where all 3 line up.
I just thought of something... IF... I don't know how generic this will turn out.... If you remove the ECU, turn the key on, have the same dash display, the dash pod may have a built in, "hey, check to see if the ECU is plugged in" kind of flashing at the dash. IF... the dash has the same look to it as if no ECU was taken out, this says, no, I was wrong with the dash having memory to recognize the ECU is not plugged in like my other bike. The second thing would be, no, it does not have that option and look how the dash responds as if the ECU was short as if not plugged in.
Short of knowing someone with the same bike, swap ECU's, you may have found it. What else could it be if you cleared everything else. The next thing is to remove the ECU and now see if it is hot? This is by checking with the shop manual, follow some prong resistances out those legs to see if making the ECU hot on the bench, can readout some ohm resistances against the book.
The last resort before the big bucks come out.
Short of knowing someone with the same bike, swap ECU's, you may have found it. What else could it be if you cleared everything else. The next thing is to remove the ECU and now see if it is hot? This is by checking with the shop manual, follow some prong resistances out those legs to see if making the ECU hot on the bench, can readout some ohm resistances against the book.
The last resort before the big bucks come out.
I actualy did that check, but I noticed in pictures that the slots on the flywheel didn't appear to be milled directly on the rotor. I pulled the cover and discovered that it's actually a collar, looks like it's a shrink fit on the rotor. It didn't appear to have moved so i pulled the flywheel to inspect while i had it tore down.
Did this test too. When you power up with the ECU disconnected you get a CAN error ( don't remember the exact text that was displayed). This makes sense because all the info on the dash (with the exception of the stand switch) goes to the ECU, then gets communicated to the dash via the CAN buss.
I have not powered up the ECU on the bench due to lack of info. I know what the individual sensor signal architecture should be and i have the equipment to simulate, all this does however is eliminate the factory wiring. I've checked the wiring for shorts and grounds thoroughly, 22 hours of testing and an *** load of notes later i'm very satisfied that the harness is ok. I haven't been able to locate a schematic of the ECU internals and am hesitant to open it (it's probably potted anyway), nor have I been able to locate any test resistance/voltage data that I could verify against.
Well, for 11 bucks and 9 dollars shipped, you just missed a cheap test ECU. It said it had a miss in it. Funny, I happen to have a japanese ECU of another brand, it does probably the same annoying stumble that was being sold. One of that grids burn out or something, meaning, once that calc happens to pul that particular map up. But that was more a good known test box for the tool box. At least it starts the bike.
If that Rotax is not sealed, has screws on the back, it probably can be looked into. I mean, if it is all sealed in epoxy, forget it. However, it could be a prom pin floating out or who knows? Motherboard cooked? Broken prong at one of the legs that both harness connectors snap over the ECU? Did you poke each pin to see if one lost a tight fit?
If that Rotax is not sealed, has screws on the back, it probably can be looked into. I mean, if it is all sealed in epoxy, forget it. However, it could be a prom pin floating out or who knows? Motherboard cooked? Broken prong at one of the legs that both harness connectors snap over the ECU? Did you poke each pin to see if one lost a tight fit?
Well that's passed half of my lunch break. I didn't even notice the foul tasting cupasoup. Thank you.:thumbup
Joined Jun 2009
3K Posts | 3+
North East Ricoland
