Hey Hey, its crank sensor day!

Coyote Chris

Coyote Chris

Site Supporter
Joined
Aug 25, 2011
Messages
2,970
Location
Spokane
Bike
10 Red NT 14 FJR, 17 XT
I will put up a write up (in the electrical section) on the wonderful time I had putting in the new crank sensor in my NT .
Seven hours of non stop fun but at least I developed some short cuts for next time. Will try and "prove" the bike this week
before heading to Canada.

I will say that when I did the test before the operation to see how the computer would react when I pulled apart the sensor connector while running I was surprised.
The engine just died...no check engine light. I tried to start the bike and it wouldnt start. Again no check engine light. Then I turned the key off and put the connector back.
The engine started right up. No check engine light. Oh well.

On aggravating thing about old age. With no memory, taking off the mid cowl is like a new experience every time.... :(
 
Thanks for that info Chris. It is not what I expected.
In this case, it seems that it needs BOTH Cam and Crank Sensor signals all of the time.
There is still one question in my mind though, Chris.
Was the engine at NOT when you tried the unplugging or was it still cool/warming up?
There may be a difference with that.

Macka
 
Macka said:
Thanks for that info Chris. It is not what I expected.
In this case, it seems that it needs BOTH Cam and Crank Sensor signals all of the time.
There is still one question in my mind though, Chris.
Was the engine at NOT when you tried the unplugging or was it still cool/warming up?
There may be a difference with that.

Macka
Click to expand...

That is a great question and at the time I just assumed that the O2 sensor wasnt warmed up so the computer might have taken that or something else like water temp into consideration.
The the basic question is still unanswered. Frosty and I go for some test rides Tuesday. I still want others to tell me what they thought about the slack in the oil pump chain and sprocket assy.
 
Coyote Chris said:
I will say that when I did the test before the operation to see how the computer would react when I pulled apart the sensor connector while running I was surprised.
The engine just died...no check engine light. I tried to start the bike and it wouldnt start. Again no check engine light. Then I turned the key off and put the connector back.
The engine started right up. No check engine light. Oh well.
:(
Click to expand...

How my brain thinks it works....and it may be wrong.

The Crank and Cam position sensors only create a voltage and pulse signal for the ECM only when the engine is turning, unlike the Throttle Position Sensor (and other sensors) that provide a signal to the ECM the moment the ignition key is turned on. It is normal for the Crank and Cam position sensors to have 0 Volts an 0 Amps before & after the ignition key is turned on, so there is nothing that the ECM can detect as abnormal until the engine is cranked. I think this is why you don't get a check engine light/MIL when the key is initially turned on during the self check. The engine will not start with the crank or cam sensor disconnected because ECM will never get a pulse signal during the cranking. If you crank the engine for >10 seconds the MIL should come on showing which crank or cam sensor is disconnected, or not sending a signal for 10 seconds with the engine turning. I think this is why it is in the troubleshooting to crank the engine for >10 seconds if the engine wont start and then check again for a MIL. See page 6-6 troubleshooting.

That is interesting that the running engine didn't give the MIL when you disconnected the connector, but again it may need a full 10 seconds to says "yes, I definitely lost the pulse signals from the CKP" to give the MIL although it immediately kills the engine. IDK.
 
Last edited:
dnktng said:
How my brain thinks it works....and it may be wrong.

The Crank and Cam position sensors only create a voltage and pulse signal for the ECM only when the engine is turning, unlike the Throttle Position Sensor (and other sensors) that provide a signal to the ECM the moment the ignition key is turned on. It is normal for the Crank and Cam position sensors to have 0 Volts an 0 Amps before & after the ignition key is turned on, so there is nothing that the ECM can detect as abnormal until the engine is cranked. I think this is why you don't get a check engine light/MIL when the key is initially turned on during the self check. The engine will not start with the crank or cam sensor disconnected because ECM will never get a pulse signal during the cranking. If you crank the engine for >10 seconds the MIL should come on showing which crank or cam sensor is disconnected, or not sending a signal for 10 seconds with the engine turning. I think this is why it is in the troubleshooting to crank the engine for >10 seconds if the engine wont start and then check again for a MIL. See page 6-6 troubleshooting.

That is interesting that the running engine didn't give the MIL when you disconnected the connector, but again it may need a full 10 seconds to says "yes, I definitely lost the pulse signals from the CKP" to give the MIL although it immediately kills the engine. IDK.
Click to expand...

You are correct! The ECM will not illuminate the MIL the instant a malfunction is detected otherwise the MIL would illuminate frequently when a mere hiccup occurs. Usually, the ECM requires a failure of several self tests before it will command the MIL to illuminate which may take up to ten seconds.

Mike
 
It is quite common to have slack in the oil pump chain. CX500's had much more slack than 1/2" and they redlined at around 10,000RPM and it was never a problem.

Seagrass
 
mikesim said:
You are correct! The ECM will not illuminate the MIL the instant a malfunction is detected otherwise the MIL would illuminate frequently when a mere hiccup occurs. Usually, the ECM requires a failure of several self tests before it will command the MIL to illuminate which may take up to ten seconds.

Mike
Click to expand...

I believe that dnktng and Mike are both MOSTLY correct.
Some systems do require several "cycles" before they will recognis(z)e a fault in some circuits.
On the other hand, I know from experience, that most very recent (read that as more advanced) systems will instantly see a fault in SOME circuits.
Those circuits are the "critical" ones usually associated with emission control.
For example, if you disconnect an O2 Sensor or Coolant Temperature Sensor on a late model car it will immediately bring on the MIL.
This is because both of those will have an immediate and huge affect on exhaust emissions (ie mixture control).
The engine will still run under both conditions but emissions will jump dramatically.
There are so many different programs in modern ECUs that it is impossible to keep up with them all.
After 47 years doing this 5-6 days a week I am still learning.:confused:

Macka
 
Macka said:
I believe that dnktng and Mike are both MOSTLY correct.
Some systems do require several "cycles" before they will recognis(z)e a fault in some circuits.
On the other hand, I know from experience, that most very recent (read that as more advanced) systems will instantly see a fault in SOME circuits.
Those circuits are the "critical" ones usually associated with emission control.
For example, if you disconnect an O2 Sensor or Coolant Temperature Sensor on a late model car it will immediately bring on the MIL.
This is because both of those will have an immediate and huge affect on exhaust emissions (ie mixture control).
The engine will still run under both conditions but emissions will jump dramatically.
There are so many different programs in modern ECUs that it is impossible to keep up with them all.
After 47 years doing this 5-6 days a week I am still learning.:confused:

Macka
Click to expand...

My reply was poorly worded. The ECM will instantly recognize a fault however it won't illuminate the MIL instantly in many cases to prevent false positives. Certain sensors are assigned a higher "value" when they malfunction, thus an immediate illumination of the MIL. This theory of operation pertains to US OBD II vehicles. I have to assume that Honda motorcycles and vehicle in OZ have a similar emissions strategy.

Mike
 
Coyote Chris said:
I will say that when I did the test before the operation to see how the computer would react when I pulled apart the sensor connector while running I was surprised.
The engine just died...no check engine light. I tried to start the bike and it wouldnt start. Again no check engine light. Then I turned the key off and put the connector back.
The engine started right up. No check engine light. Oh well. :(
Click to expand...

I tried to replicate your results & this is what I found.

-First, verified no fault codes in ECM.
-Disconnected CKP connector, ignition switch on..... normal self check on MIL, no locked in MIL, no stored ECM fault code.
-Attempted a normal engine start with CKP connector disconnected.....no MIL, no stored ECM fault code.
-Attempted 10 second crank with CKP connector disconnected...... MIL on at ~5-6 seconds. 19 fault code stored in ECM.

-Reset fault codes, connected CKP connector and started engine normally.
-Disconnected CKP connector......Engine died, did not notice MIL (it dies fast and I wouldn't have caught in a riding situation if it did light up), but a 19 fault code is stored in ECM. Cycled ignition and MIL self check is normal, normal (~3-4 second) crank won't give MIL, but a longer 10 second crank will. Regardless, there is a stored 19 fault code to indicate the cause of the engine dying was the CKP circuit. Chris, you most likely still have a 19 fault code stored in your ECM and if you ever get any other fault and read your codes you could be mislead into believing the CKP is at fault again.

Macka said:
For example, if you disconnect an O2 Sensor or Coolant Temperature Sensor on a late model car it will immediately bring on the MIL.
There are so many different programs in modern ECUs that it is impossible to keep up with them all.
After 47 years doing this 5-6 days a week I am still learning.:confused:
Macka
Click to expand...

Many/most sensors are voltage sensors, like the O2, Coolant temp, MAP, TPS (more?) on the NT and many other systems. With voltage sensors as soon as the ignition goes on a voltage is sent to the sensor and a signal voltage is returned to the computer to analyze if it is a normal/expected voltage for a non-running condition. But the CKP & CMP sensors in the NT (& other systems) are a small AC generator that sends no voltage (just like the stator mounted on the crank) until the engine rotates, it's passive. These sensors have the same 0 Volts & 0 amps signal with the ignition off or on and that is normal/expected. Not until the engine rotates does the computer expect to see any voltage. I imagine in today's more advanced ECU's that during the self check a quick pulse could be sent to a CkP (sensor that is a generator) to check for continuity, or some other method.
 
To amplify on what Dan said, depending of the ECU strategy, you may not retain the fault code after so many good self-tests. OBD2 vehicles will self-extinguish the MIL after three good self-tests. I can't say for certain that Honda employs this strategy in the NT but I would "assume" that it does. That would be a good exercise for Dan. Unplug it, set a fault code, plug it back in and see if the MIL goes out and if the ECU retains the code. Bet it doesn't.

Mike
 
The Honda ECU will most likely retain the code for few key cycles, then the code will expire as the CKP fault no longer exists. These strategies vary widely amongst different manufacturers, so another manufacturer like Suzuki or Yamaha might do things very differently. It doesn't always jive with our expected "human" logic about the way things should be, but there's engineering reasons behind the way it's done.

Brad
 
For those interested, page 6-16 of the service manual outlines a procedure to clear the DTC's without a scan tool.

Brad
 
dnktng said:
I tried to replicate your results & this is what I found.

-First, verified no fault codes in ECM.
-Disconnected CKP connector, ignition switch on..... normal self check on MIL, no locked in MIL, no stored ECM fault code.
-Attempted a normal engine start with CKP connector disconnected.....no MIL, no stored ECM fault code.
-Attempted 10 second crank with CKP connector disconnected...... MIL on at ~5-6 seconds. 19 fault code stored in ECM.

-Reset fault codes, connected CKP connector and started engine normally.
-Disconnected CKP connector......Engine died, did not notice MIL (it dies fast and I wouldn't have caught in a riding situation if it did light up), but a 19 fault code is stored in ECM. Cycled ignition and MIL self check is normal, normal (~3-4 second) crank won't give MIL, but a longer 10 second crank will. Regardless, there is a stored 19 fault code to indicate the cause of the engine dying was the CKP circuit. Chris, you most likely still have a 19 fault code stored in your ECM and if you ever get any other fault and read your codes you could be mislead into believing the CKP is at fault again.



Many/most sensors are voltage sensors, like the O2, Coolant temp, MAP, TPS (more?) on the NT and many other systems. With voltage sensors as soon as the ignition goes on a voltage is sent to the sensor and a signal voltage is returned to the computer to analyze if it is a normal/expected voltage for a non-running condition. But the CKP & CMP sensors in the NT (& other systems) are a small AC generator that sends no voltage (just like the stator mounted on the crank) until the engine rotates, it's passive. These sensors have the same 0 Volts & 0 amps signal with the ignition off or on and that is normal/expected. Not until the engine rotates does the computer expect to see any voltage. I imagine in today's more advanced ECU's that during the self check a quick pulse could be sent to a CkP (sensor that is a generator) to check for continuity, or some other method.
Click to expand...

Wow...double wow.....thanks for the info! I had no idea, Dan Mike and Brad.
I am afraid to ask what an HDS costs, which leaves the page 6-16 way of clearing the trouble codes using the SCS special tool. It sounds like you short two terminals in the dummy plug using the special connector, turn the engine switch to on, remove the plug, wait for the Mil to light, put the plug back with the light on and that will reset the memory. OK, I will do this but not before any trip!
(Note to Honda. Just make a darned plug that you insert and turn on the engine switch and boom, done. No hokey pokey business.)
I
 
Chris,

Yes, engine stop switch needs to be in "on" or "run" position in order for the clear DTC command to be issued inside the ECU, and the sequence of removing and installing the jumper at the right time is part of it too. The ECU will not clear the codes unless it sees the sequence of events it wants. This "dance" of events to get the ECU to do your bidding was not that uncommon in the car world, Ford did this for years prior to 1996.

Brad
 
rc_heli_nut said:
Chris,

Yes, engine stop switch needs to be in "on" or "run" position in order for the clear DTC command to be issued inside the ECU, and the sequence of removing and installing the jumper at the right time is part of it too. The ECU will not clear the codes unless it sees the sequence of events it wants. This "dance" of events to get the ECU to do your bidding was not that uncommon in the car world, Ford did this for years prior to 1996.

Brad
Click to expand...

Thanks Brad. If we dont go to Canada in a few days I will try it. I am pretty good at making jumpers after 30 years in the two way radio fixit business.
 
As was very well explained below, there are almost as many different ways of doing it as there are different brand names.
For those of us who have to diagnose and fix them it would be a lot easier (and cheaper for the customer) if all systems followed the same principles.
The manufactures, of course, don't want that as it would take work away from the Dealerships.
Fortunately for us here, the NT system is fairly basic and easy to fix.

Macka
 
You must log in or register to reply here.
Back
Top Bottom