Civic elusive misfire

2006 Civic EX manual transmission, 230,000 kms.
misfire mainly (95%) occuring on cyl 3 but 5% cyl 1 only under these conditions:
1900 rpm, 4th or 5th gear with light throttle application.
Can't reproduce at idle or 1900 rpm in the garage.
Misfire disappears under more throttle application.
spark, coil, injector have been moved to another cylinder. misfire remains on cyl 3.
fuel pump and filter changed, injectors cleaned
I have a video of the injector control wire which I'll attach. I don't have the experience to know what I'm looking at though. Any help?
also see the video of it here

That looks kind of like a peak and hold injector waveform.
I don't remember how to access the e-book, so a picture of the hard copy will have to do:



Did you repeat that test on a cylinder that is not reporting misfires?

Yours looks like it's being pulled fully to ground for the second half of the pattern while the examine does not.

No I haven't run it on another cylinder yet. I will today and I'll post it. If it was this peak and hold type wouldn't it always do that rather than show up momentarily? Thanks for the pic.

All Data doesn't specify, but if the Snap On troubleshooter is to be believed, then it doesn't look like it would be a peak and hold type. It's not always correct though. I am curious to see how a non misfiring cylinder's injector behaves under the same conditions. It would be cool to see them both at the same time if you've got more than one channel

Thanks Noah for the reference there. I can't run the other cylinder till I get the car back from my son tonight. But I'm pretty sure based on this info that I have a weird signal going to the injector. Next problem...what could possibly cause that? I believe it comes from the ECM.

Update 20 Nov 2019
video of Cyl3 and Cyl4 injector control wires.
Misfire condition appears only on Cyl3 at about 1900 rpm and light throttle application. At idle it disappears. At more open throttle it disappears. Can anyone help identify the cause of this weird waveform?

Wow, that's pretty cool. Nice capture.
I would scope the cam and crank sensors during that event.

I don't know if the PCM could "double tap" that injector on purpose if it was pulling too much amperage, but I would also current ramp that injector at idle and during that event and compare that to a known good cylinder under the same conditions.

It looks like a mis triggering of the injector driver, I had a similar condition on a fiat 1.2 petrol, the ecu was tested and no fault found,when the ecu was refitted and wiring multi plugs put into ecu-- no fault present,morale check pin fitment of connectors at ecu or pcm,fault has not returned, 3months since misfire.

update - don't need pinout for ECU

Check this out.
Camshaft Position Sensor on CH1 in YELLOW (middle)
Crankshaft Position Sensor on CH2 in green/blue (top)
Cyl3 Injector Control in Magenta (bottom)What could be the cause of this spike in voltage in time with the cyl3 firing? And there is a smaller spike at what I assume from the CKP waveform is TDC. Keep in mind I get misfires 90-95% on cyl 3 and the rest on Cyl1. Any thoughts on what could cause this?

@Noah Does he need a picture of the wiring diagram plug? That would be under the information on wiring withshopkey wouldn't it?

I could look the information up later today if so. LEt me know. @ The thread poster pm me the vin and I can see what I can find. Can't send diagrams though as that's a sketchy area I won't venture into. If I look it up and give you the color or I wonder if Paul is ok if I took a picture of the pinout with my phone and post the picture as I know as him being a content creator that's a hot button topic plus I don't know his position on it.

I've seen him use diagrams in his videos but I think that falls under a different area. If someone else knows otherwise please correct me.

Sent from my SM-N975U1 using Tapatalk

From what I can see from other Hondas online, the CMP signal shouldn't look like this at all it should be a form of square wave. Am I right thinking the sensor has an internal failure?
It's a 3 wire sensor. Signal to the ECU is white/green. The ground wire seems good I checked voltage and resistance. the other wire YEL/BLK is from the Main relay. I capture battery voltage there. Maybe it should be 5 volts I don't know.

Honda makes me think valve clearance or valve sealing issue which would fit the symptoms.

But compression is fine and similar in all cylinders. Valve clearance has also been double checked fine.

Forgive me if you posted and I didn't see it, but what color trace = which signal?

I'm thinking green is cam, pink is cyl 3 ignition and yellow is crank? It doesn't really look like a crank signal though? Sorry for the confusion on my end.
Also sorry for not getting back sooner. Life and work and what not...

There is a repeating double synch in the green trace, I don't have reference material in front of me at the moment, but from what I can recall of looking it up the other day I think the cam sync pulse was just a different sized pulse. I'll look it up again tomorrow and get back to you.

Any work done on the engine recently, or did this just come out of left field?

If you have the time and ambition it would cool to see if that double pulse in what I think is the cam signal corresponds to the double injector pulse.

I think the spike is just a kind of inductive resonance from the coil firing. Not really there, so to speak.
I also think the misfires on cylinder 1 really aren't there. It's not uncommon for me to see a companion cylinder count false misfires in much lower amounts during a single cylinder misfire.

Sorry I coded the colors on the youtube description, but didn't copy it, here it is:
Camshaft Position Sensor on CH1 in YELLOW (middle)
Crankshaft Position Sensor on CH2 in green/blue (top)
Cyl3 Injector Control in Magenta (bottom).

It's the yellow signal- the camshaft sensor - that looks out to lunch to me. I think it should be a square wave but I have no reference material to check it against.

I think the double sync in the green is a marker for tdc, but again I have no reference for that. It's not aligned with cyl3 in magenta either way.
If I had a spare cam sensor I'd swap it for that. Is there a test you can do to see if the cam sensor is functioning as designed? It's giving these weird pulses and then that giant pulse right in time with cyl 3 where the miss is usually happening. I wasn't driving during this last scope it's at idle.

Thank you for clearing that up, again, sorry for the confusion on my part.

The yellow trace certainly does not look like what the cam sensor should be producing.
It should be a square wave as you had mentioned.

As far as testing the sensor, I believe you said you tested and found power and ground at the sensor all ready.

That's good. If you're certain you were making good contact on the signal wire during that capture then there's something wrong there.
Before condemning the sensor, I would check the signal wire for a short to ground by disconnecting the sensor and the computer and then ohm the signal wire to ground. Looking for OL mohms.
I see the trace stays around 1-2v on the scope, so I don't think you'll find a short but I would check before replacing the sensor anyway.
You could also yank the end cover and check the tone generator for damage, can but given it location I don't think it would likely be damaged.


I would imagine that with what amounts to no cam sensor signal there would be a trouble code set...

Thanks for all your tips Noah. Against my initial urge I didn't replace the camshaft sensor and rechecked it. I must not have had a good connection. Here is an updated video of a run with signals from top to bottom as follows
1 - camshaft sensor
2 - crank sensor
3 - Cyl3 Injector Signal
4 - Cyl3 Ignition Coil (I'm not sure what this wire is called, but not batter voltage and not ground, it's a 3 wire.

Here's what I see. Nothing in any of the other channels. The others seems stable and that injector signal seems to come out of nowhere. So I think I'm getting good data to the ECU from CMP and CKP. Any other ideas? I haven't checked the current to the injector, but I'm not sure if there's a reason for that? ... Yeah, you told me to current ramp the injector at idle and during the event - i just re-read that so I will do that and post it.

Here is the current ramp - showing a current draw with the voltage. don't know what to say or where to look now for the cause of this.

That is wacky!
I don't see any funny business with the cam signal while the PCM is double firing the injector.
Aside from Landroverman's advice about checking pin fittment, I don't have any ideas beyond changing the PCM.