Ignition Problem?

wjbertrand said:
So Mark are you saying there's a loss of 12V power only to the ECU somehow?
Click to expand...

That's what the behavior suggests. The dividing line between what shuts off (engine, tach) and what stays on (everything else) appears to run right through the ECM.

Just to be clear, I don't think something outside the ECM triggering a software failure that results in a watchdog reset. It's possible, but it's a long shot and not a rabbit hole we should be in just yet. Those resets happen in a handful of milliseconds and are hard to notice unless something is really hosed up and creating them by the hundred. What you're seeing goes on a lot longer. The real takeaway and the reason I brought it up is that when the ECM is first powered up or reset, the firmware will start running the engine if it sees the crank and cams turning.

If this is a supply dropout and inertia keeps the engine turning until it comes back, it'll start running again once the ECM regains its senses. If it's long enough to see on the tach, it occurs to me that a $5.00 LED test lamp connected to the 12V supply lead (or the corresponding lead on the diagnostic connector) would show it. Something else like the system voltage sagging would take a 'scope to nail down.

--Mark
 
Ah, I re-read Mark's post above at home tonight and realize that in my haste at work earlier I mis-read what he was saying. So he thinks the watchdog timer cutout would be too short to notice, but a power failure to the ECM would be long enough to cause your symptoms. The first time I read it I thought he was saying the opposite, power failure would be too fast, watchdog would show symptom.

So now that I have that straightened out, do you think its worth swapping around a few fuses to see if the symptom goes away, or elsewhere? I'm wondering if you could somehow have a marginal fuse that vibrates open at the critical RPM and then reconnects a split second later. Or a marginal connector in that path. Are there connection points anywhere near the ECM where you could monitor voltage with a multimeter and rev it in your garage and see if its stable? Also, like Mark suggested you might want something faster responding like a LED, or maybe one of those 12v testlights from the auto parts store.
 
Last edited:
dwalby said:
So now that I have that straightened out, do you think its worth swapping around a few fuses to see if the symptom goes away, or elsewhere? I'm wondering if you could somehow have a marginal fuse that vibrates open at the critical RPM and then reconnects a split second later. Or a marginal connector in that path.
Click to expand...

I keep harping on watching the ECM supply line because a drop or sag immediately rules out a lot of hard-to-diagnose things and points in the direction of things that are very easy to run down. There's no reason to guess at it when you can find out one way or the other through measurement.

--Mark
 
I have had more than a few cars with ground problems do this type of action ,ecm will not set due to loss of ground .Check all the grounds particularly the ground for the ecm and maybe the relays . It seems like a temperature related issue so that might involve a bad connection ,these a tough thing to find sometimes needed to do a voltage drop using millivolts to find it .
 
Blrfl said:
I keep harping on watching the ECM supply line because a drop or sag immediately rules out a lot of hard-to-diagnose things and points in the direction of things that are very easy to run down. There's no reason to guess at it when you can find out one way or the other through measurement.
Click to expand...

You deleted the part of my post where I asked if there is an accessible point at which the measurement can be taken. I don't have a 1300, so its not obvious to me. Since you keep saying do it, but not how to do it, does that mean to the average 1300 owner that the monitoring point for the ECM power supply is obvious and easily accessible?
 
dwalby said:
You deleted the part of my post where I asked if there is an accessible point at which the measurement can be taken. I don't have a 1300, so its not obvious to me. Since you keep saying do it, but not how to do it, does that mean to the average 1300 owner that the monitoring point for the ECM power supply is obvious and easily accessible?
Click to expand...

Honda has a breakout diagnostic cable (you need two of them) for the main ECU harness but they are really expensive. Neither of my local dealers have them so cannot even loan them to me. Almost all of the diagnostic work described in the manual and by Mark would I think require the use these break out cables. No easy way to access the power connections for the ECU as the harness is pretty well wrapped up an the ECU is buried in the tail section. I'm not even sure I could identify which wire is the supply. I mentioned in an earlier post that I examined all of the ground points that I could find. The main spot is under the tank where perhaps a dozen wires come together and are bolted to the frame cross member. Though that connection point was clean and tight, I took it all apart and abrasively cleaned all of the connection lugs and the spot on the frame anyway. The one thing I can't find is where the battery negative cable is grounded, though I don't think it's the problem because there is absolutely no change in my system voltage on the Datel meter when the misfire occurs. I would expect if it went all the way back to a bad battery ground there'd be other symptoms, eg clock resets, etc.

I changed out the 58# fuel pressure regulator back to the OEM one last night and test rode it this morning, as expected - no change at all.
 
dwalby said:
an accessible point at which the measurement can be taken.
Click to expand...
The +12vdc circuit that supplies voltage to the ECM can be found in the RED Service Connector (Black/White conductor), located beneath the right side (brake lever/pedal side), of the pillion seat.

Perhaps instead of trying to position, and read a voltage meter at speed, installing a LED lamp in the circuit, and locating the LED lamp where it can be monitored by the operator would be safer.

The advantage of the LED is that it draws a miniscule current load, which should not skew the test results, also the LED lamp does not suffer from filament shut-off lag, such that even a voltage drop of short duration will be noticeable.
 
dwalby said:
Since you keep saying do it, but not how to do it, does that mean to the average 1300 owner that the monitoring point for the ECM power supply is obvious and easily accessible?
Click to expand...

Yes, it does. The black/white wire on the diagnostic connector in the back of the bike is tapped from the same circuit that supplies the ECM. The ideal thing would be to slip a needle probe into the ECM connector where that wire lands, but the diagnostic connector is close enough to at least get us pointed in the right direction.

The circuit we're looking at is the one highlighted in this picture:
ignition.jpg
The diagnostic connector isn't shown here, but it is on the wiring diagram. Note that this circuit also supplies the low side of the ignition coils.

--Mark
 
Anna'sDad said:
The +12vdc circuit that supplies voltage to the ECM can be found in the RED Service Connector (Black/White conductor), located beneath the right side (brake lever/pedal side), of the pillion seat.

Perhaps instead of trying to position, and read a voltage meter at speed, installing a LED lamp in the circuit, and locating the LED lamp where it can be monitored by the operator would be safer.

The advantage of the LED is that it draws a miniscule current load, which should not skew the test results, also the LED lamp does not suffer from filament shut-off lag, such that even a voltage drop of short duration will be noticeable.
Click to expand...

Blrfl said:
Yes, it does. The black/white wire on the diagnostic connector in the back of the bike is tapped from the same circuit that supplies the ECM. The ideal thing would be to slip a needle probe into the ECM connector where that wire lands, but the diagnostic connector is close enough to at least get us pointed in the right direction.

The circuit we're looking at is the one highlighted in this picture:
ignition.jpg
The diagnostic connector isn't shown here, but it is on the wiring diagram. Note that this circuit also supplies the low side of the ignition coils.

--Mark
Click to expand...

Thanks guys, this is very useful. I'll try to cobble something up this weekend.
 
Anna'sDad said:
Perhaps instead of trying to position, and read a voltage meter at speed, installing a LED lamp in the circuit, and locating the LED lamp where it can be monitored by the operator would be safer.

The advantage of the LED is that it draws a miniscule current load, which should not skew the test results, also the LED lamp does not suffer from filament shut-off lag, such that even a voltage drop of short duration will be noticeable.
Click to expand...

IIRC, he is currently able to reproduce the problem in neutral in the garage so the instrumentation being used isn't a big deal (Jeff correct me if I misunderstood, the thread is getting too long to go back and search things out). I agree with everyone though that a LED will provide a quicker response than a typical voltmeter. My Fluke has a bargraph at the bottom that is supposed to respond faster than the main digits in the display, but even with that I don't know how fast it really is at responding, and not all DMMs have that feature to begin with.
 
Well, this is weird. As of lately, I can no longer get the bike to misfire when stopped in neutral and just revving the engine. I tried at a couple of stop lights and then again when I arrived home in the garage. It got hot enough to switch the radiator fans on, but no misfire. Took it back out on the road and sure enough there it is again, very consistently and reproducible. This kinda throws a wrench into my in-garage diagnostic plans then...

So I've been talking to a couple of Dyno shops, one very close to me in Oxnard called Ventura County Motorsports and another, not too much farther, in Camarillo called Jett Tuning. Jett doesn't have Saturday hours though so also being closer, I may have VCMS put it on their dyno to try and figure out what it's doing.
 
How about a flaky CKP (Crank Position Sensor) aka Ignition Pulse Generator? Intermittent no tach signal to ecm, = no spark control from ecm and also no tach sig to tachometer. The design of the crank sensor is a simple AC generator and you might be able to capture a voltage dropout with voltmeter set on AC volts (low voltage) backprobed to the sensor. If no output at sensor when hicup happens - you found it. It may not be happening long enough for the ecm to log a fault, as the efi system on these bikes is about the equivelent design of early 90's honda cars.
 
Thanks for the thought. I've already replaced the crank position sensor but it did not solve the problem.


-Jeff
 
Well, I can scratch plug wires and caps off the list. Installed brand new out of the package ones to the tune of $120 last night and rode her into work today. Absolutely no effect at all on the problem. Whatever it is I've yet to even touch it evidently. At this point I'm making the decision not to take her to WeSTOC. Even if I somehow fix it now, there's too little time to establish any confidence.
 
Maybe a broken wire somewhere which is a tough one.
 
Mellow said:
Maybe a broken wire somewhere which is a tough one.
Click to expand...

Must be. Somewhere deep down in a harness somewhere a wire has broken with no outward sign to give me a clue where to look. It's like finding a needle in a haystack. I sad say, I'm getting close to done with this thing.
 
I guess you can always part it out. This whole diagnostic nightmare must be a............. diagnostic nightmare :(
 
Last edited:
Jeff, you seem to get the 5 sigma ST copies. First the only known oil burning 1100, now this. Honda must have your picture hanging on a wall somewhere, but I don't know what you did to deserve it.
 
You must log in or register to reply here.
Back
Top Bottom