help diagnose intermittent power

I had a problem develop on my 2003 non-ABS last night on the way to the gym.

History: The only recent thing I had done was R&R the forks and replace the seals. While I had the front disassembled, I replaced the spark plugs. I briefly verified that the front coils were plugged in ok. I did replace the rear plugs too, but I haven't verified the connections there. The bike is stock, with the exception of a tail-light flasher.

--
Problem Description:
I was riding on the freeway, on an interchange when i had a sudden loss of power. It did not seem like all cylinders were off, just some. I was 30% throttle, maybe 60 mph exiting a <10 degree bank. I say this because I was able to roll on to keep the speed from falling off as drastically as it would have if I chopped the throttle. Gas gauge was a little over 50%. Temperature at 180F. Tach jumpy, when the power would cut, it would drop faster that the RPMs would (since I was in gear, I know they don't drop1 that quickly). MIL off.

I hobbled along the feeder road with the surges until I got to a gas station. I was thinking maybe I got bad gas, so I would dilute whatever crap I got. Topped off the tank, mixed it up. Pull out of the parking lot, continue the trip. Problem still exists. Probably not bad gas.

I was thinking, well maybe VTEC spool valve is loping. Since that doesn't actuate until after 150F, I figured, I'd try it once the bike cools off. Problem still exists when cold. Its still possible, but if VTEC is coming on and off because of a spool valve problem, I would think the problem would manifest at idle as well. (I did this later).
--
Observations:
1. If I blip the throttle at a standstill, the bike dies. Takes a while to get going afterwards. Bike idles fine afterwards. MIL off.
2. This was at night. I took the opportunity to kill off a battery/alternator problem. With brights on, the lights do not dim when the symptoms appear.
3. It seems the threshold is RPM independent, but if I hammer the throttle, I get this.
4. Bumps in the road/railroad tracks do not seem to exacerbate the problem.
--
Commentary:
I don't know the electrical system of the VFR nearly as well as some of my other bikes, and haven't looked over the schematics - will try to do that later this week. Does the VFR have a pot and idle contact TPS, or is it pot only, or switch only? I was going to start with the TPS and connections next.

I think in this implementation, there is a cam/crank position sensor, if there was loss of sync, does anyone know which of the two the tach uses, or does the tach feed come over CAN or something else?

If the bank angle sensor malfunctions, does that trip the MIL, or does it just cut power?

The MIL does not come on (the light is not burned out, it does come on when I key on/engine on). I have not tried to pull codes yet.

I would hope if it was a engine control/sensor problem, the MIL would come on.

Questions restated:
1. Has anyone had anything similar, and what were the culprits?
2. Does the VFR have a pot and idle contact TPS, or is it pot only, or switch only?
3. Where does the tach feed from?
4. How does the bank angle sensor cut power?
5. Is there a history of some engine sensors commonly failing on 2002+ VFR?
6. How does the MIL behave?


Thanks for your time and feedback.

 

Jumping tach may suggest electrical problem, and FI doesn't like low voltage or coroded ground points. Test your charging system output across the battery terminals with the engine turning about 5000 rpm, whacha got ?? Start testing with a fully-charged battery and record charging volts across a range of rpms with headlights on both hi and low. Strange inconsistency would mean a bad r/r most likely.

Had harness recall work done ??

 

I have not had the harness recall done, but maybe the previous owner did. Do you have a copy of the service bulletin? What are the symptoms?

I am leaning towards an ignition fault, particularly because I touched the plugs recently.

I am leaning against alternator problems bc the headlight does not dim when the symptoms manifest. Will check it out tomorrow and update the thread. Let me know if you think of anything else. Thanks for your input.

 

I checked the frame VIN on the steering head. The recall is notated by a punch mark, which is offset down and left from the first VIN character. My bike has had this work done at some point.

I reviewed the schematic yesterday. These answers are for the 2002+ VFR.
2. The TPS is a pot only sensor.
3. The tach feed comes from the ECU.
4. The bank angle sensor disables the main relay (actuators and 02 sensor power), fuel pump, and ECU.

Some things I discovered while reviewing the schematic and design diagram:
There is a crank and cam sensor setup on this bike, it is SFI and no waste spark. Each coil and injector set is controlled individually, hence the need for the cam sync pulse. There is also a MAP/TPS/IAT setup, so this is a speed-density setup. I wasn't really expecting anything else, but its nice to confirm.



Just trying to get my bearings before I go taking stuff apart. Maybe someone else might find this this information useful.

Unanswered questions:
1. Has anyone had anything similar, and what were the culprits?
2.
3.
4.
5. Is there a history of some engine sensors commonly failing on 2002+ VFR?
6. How does the MIL behave?
- Does the ECU require multiple failure cycles to register a code or will it read out fault codes without any faults being stored (no MIL on)?

 

I just realized during the recall the dealer left the old harness in place. Any reason they would do that? Can I remove it?

 

I removed the old harness last night. Because it was left in place, some of the connections were strained, particularly the bank angle sensor.

I'm not convinced this was the root cause, because I haven't reproduced the problem - it was too late to bug my neighbors with revving engine...

I took a brief look at the differences in harness, it looks like the old one had connections for the ABS module; my bike is non-ABS. The new harness does not. Is that normal?

 

I pulled codes even though the MIL was not on and found 22,23, and 24. O2 sensor faults.. These may have been left in there from when I ran the bike open loop when I had the exhaust off. I went ahead and cleared them. If they're legit, they'll come back.

My discovery is that The MIL does not illuminate if there are stored faults.

I feel like I'm talking to myself here. Where's the centuries of cumulative forum expertise?

I'll check the alternator as suggested with my scope tonight. My visual inspections have come up fruitless so far.

 

I rode the bike around the block without the headlights (less elec load), problem still exists. Then I heard a noise that I had forgotten about. During the symptom, there is a rattle can noise. At first I thought again to my ignition shorting out, but then I remembered that its distinctively engine knock.

Knock is usually prevalent at lower rpms, with a load step if you have low octane fuel in a high compression SI engine, due to colliding flame fronts. It can be reduced by retarding ignition timing to start the flame earlier so the hot spots don't have a chance to auto ignite. The converse holds - if you overly advance ignition timing, you give more time for hot spots to set off the lower octane higher cetane crap gas, promoting knock.

Long story on knock aside, the VFR setup is an SFI with coil on plug. The ignition timing is controlled by the computer based on two things, a 12-t crank trigger and a 2/1 cam sync. The crank trigger lets the ecu resolve crank position in 30 degree increments, synchronizing the compression stroke with the cam trigger. The ecu interpolates between the 30 degree marks to plan the next timing event based on last rpm. It also feeds this as a 5vttl pulse to the cluster for the tach.

So with my symptoms, the only possible place there could be a fault is in the crank position sensor. I doubt this generation of system would perform advanced diagnostics and fail over sensors, hence the erroneous tach, and obviously bad timing causing knock.

Based on this, I pulled the clutch cover off and found that one of the sensor bolts walked loose and the other one too, and the sensor has been rubbing against the trigger wheel.

Mystery solved, I'm off to the junkyard! Hope my journey helps someone else.

 

Bad sensor.

 

Thanks for the final input. I am also having the same problem, but with no engine dieing or running rough issues, or any other symptoms at all other than power loss as if I just let off the throttle. When I go to add throttle, it lunges as if I just snapped the throttle. Drive-ability has yet to be compromised, but I know its on its way!! I will pull my cover off this weekend to check mine as well. Ill let yall know.

 

Dude, awesome troubleshooting and diagnostic. Evidence and reasoning lead you straight to the problem ! Thanx for the post!

 

Thanks, mastergregor.

@amc401: any luck? Intermittent electrical problems are the hardest to diagnose. Sometimes jiggling the harness can help expose. What does your tach do when power dies?

 

You are it!

pays off persistance does--vfr master you are...

Great catch on finding the sensor.

 

Not 100% sure on mine yet. After further digging, I noticed the PO did the O2 mod. Looking into that, i noticed that mod can cause these same problems( Power cutting out and then surging back on ) if you don’t have a power commander. So I unplugged my Power commander and un did the mod, cleaned all connections and applies electrical grease. 250 miles later still no hiccups. Unfortunately I also just got a nail in my rear tire so I can’t ride anymore till the new one comes in!! All in all, I am guessing my problem came from a bad Power commander. But not sure yet as its been a on and off problem.

And, nothing on the Tach but normal. Everything but power remained normal on my bike. FI light never even flickered.

 

@amc

You should start a new thread to get answers for your specific issues, just saying...

 

This is the pulse generator right? 30300-MBG-003

Its also referred to as the CKP sensor correct?