First off I have never owned a Honda car so take this with a grain of salt. There has been some discussion of this problem but here is a little different path laid out that might lead to a solution.
First you have mentioned a miss. It seems like you are describing a miss at a specific cylinder. You would need to determine this by the rythm of the miss. If that is the case you should make a significant effort to determine which cylinder it is. You might determine this by plug color immediately after running with the miss or you might be able to detect an exhaust manifold temperature difference with an infrared thermometer. They are getting quite cheap. If you are talking a random miss then it could be anything and separating fuel and ignition problems can be difficult.
Second: If the miss is specific to one cylinder it will not likely be fruitful to replace components that are common to all cylinders. It might be but it is not likely. MAP, coil, ignitor, or ECU fall in this category. It would be more promising to determine the cylinder and then trade what components you can such as fuel injector, spark plug, or wire and see if the problem follows or stays.
Your description makes it sound like the missing occurs when it is going into lean burn. That is it after it warms up and you are cruising.
Third: I have never heard of a failure of the pin that locks the valves in the proper configuration but if the valves failed to change properly it could certainly cause a problem in lean burn. If the second valve is opening too much you will not get the proper swirl for the lean burn. It just crossed my mind but remember I have never worked on one of these.
Fourth: Diagnosing weak ignition. One time I had a Eureka moment. It took embarrassingly long for this concept to emerge from my unconcious as it is very simple. I noticed that when spark plug wires were old I often got shocked. When spark plug wires were new I could handle them carelessly with the engine running and not get shocked. A discharge will find the path of least resistance. What emerged from this was a dual spark gap. This only works if you have a separate coil feeding a distributor. Construct a simple spark gap arrangement on something that inuslates. Attach the coil wire to a central contact point. To one side of this arrange an adjustabe gap with a wire connected to ground. On the other side arrange an adjustable gap with the wire leading to the distributor. Please pardon the verbage I need a picture here.
Anyway adjust the gap to ground very wide and the gap on to the distributor quite small. Start the engine. Open the gap to the distributor to about 1/4 inch. It should easily jump that gap with a blue spark. If the spark is weak and reddish you may have a problem. Bear in mind that if the coil output has no path to ground it may break down the insulation in the coil and in unlucky cases it can then follow the primary circuit back to expensive electronic components and damage them. At this time bring the ground close. On the first protoype this was just done by placing a plastic handled screwdriver bit between the wires to finess the gap. If one plug wire has higher resistance than the others with just a little finess you can adjust the gap so that the engine will miss on that cylinder but not on the others. This is only useful for finding high resistance problems. If you have a low resistance problem this is pretty useless. Sometimes Honda rotor insulation breaks down in the center and the current will go to ground through the center of the rotor. This will not help with that diagnosis unless it is placed in place of one of the spark plug wires.
Trivial Anectdote. About thirty years ago the head mechanic spent a couple hours trying to get a service tractor running. I was a new hire and had grown up in a repair shop. On my lunch break I used the first prototype to find a broken rotor. Fortunately these were mature people or I would have been in a spot of trouble.
For this thing to be useful you would need to play with it to develop a sense of the typical gaps and all. At one time I thought the thing had some usefulness but soon people won't know what a distributor is.
I tried to patent it through an outfit that was later featured on "60 minutes." That did not work out too well.
Pardon the length and if I mislead you I apologize.
Usedgo, I really like your spark tester. I'm going to make one of those up.
I don't think this is a case of a bad cylinder. I know that "loping" effect, and this isn't it. It misfires on all of 'em - all four plugs were black with carbon when I pulled them, and this was after only 2000 miles of use.
It's not just on the road, so I don't think it's a leanburn problem. I can hear the misfiring in the exhaust note just blipping the throttle in neutral in the driveway. But only after the engine's been running for a couple of minutes - it runs fine when I first start it, or at least seems to.
I thought "plugged cat" for a while, but I did a vacuum test, and it passed. I suppose I should really do an exhaust pressure test, the O2 sensor port would be the perfect place, but I don't know what tool to use. Could I use my compression tester, or will the hot exhaust gas kill it?
I have a used ECU here, thanks to the post by Mrmad. However, I'm kind of nervous about trying it before I'm sure everything else is OK, because I don't want to risk damaging it.
I really think something weird is going on with the ignition timing.
If anybody has a Civic (probably any engine would work about the same way) and a handy timing light, you could do me a BIG favor. Hook it up, fire up the engine (trans in neutral of course), aim the timing light at the timing mark, and blip the throttle. Tell me which way the timing mark appears to move when you open the throttle. Does it jerk toward the grille, or toward the firewall? And about how much does it move?
Oh, I also tried pulling the wire off the coolant temp sensor to see if the ECU's default map would work better (suggesting the CTS was bad), but it ran exactly the same.
Got out the timing light on my Integra. The timing marks move towards the firewall as I made small blips in the throttle by pulling on the throttle cable. I couldn't see the tach needless to say, but I'm sure I wasn't getting it much above 2000rpm and the timing marks moved to where I couldn't see them. This is normal, remember a Honda engine turns counterclockwise when looking at the crank from the driver's side. I opened up a stock bin file from a P07 in Crome and the ignition advance on the low cam is:
16.5 degrees from 500rpm-1483rpm
18.0 degrees from 1484rpm-1795rpm
18.5 degrees from 1796rpm-1999rpm
18.75 degrees from 2000rpm-2311rpm
19.0 degrees from 2312rpm-3219rpm
The high cam has more advance, but I'm assuming you are not hitting vtec while blipping the throttle while parked.
The P72 ECU in my Integra advances alot faster (at 2500rpm, it's already at 36.5 degrees, which is why the marks jump to where I can't see them)
Thanks a million times! I really appreciate your dragging out the timing light. It helps a LOT, confirming that what I'm seeing here is NOT normal.
My poor girl's timing RETARDS as I blip the throttle, the timing marks moving toward the radiator. Once the speed is steady at the new rpm timing moves back toward 16 btc. I don't see any significant advance at higher rpm.
Is there anything else BUT the ECU that could cause this? My understanding of the way this works is that the sensors report to the ECU, and the ECU alone decides when it's time to spark. Then it sends out a logic pulse that tells the ignitor to tickle the coil. That is - am I right that timing is 100% under the control of the ECU?
If so, the next question is, what would cause the ECU to lose its mind this way?
When the VX quit out on the road back in March or so, I had it towed to a shop (not much choice under the circumstances). They pronounced both coil and igniter DOA. The mechanic said that often the coil dies and takes the igniter (ICM) with it. I don't know whether that's true, but it scored him a few hundred bucks for replacing both.
If the coil whacked the ICM as it went down, maybe the ICM in turn knocked over the ECU. If so, I'd expect the ECU output to the ignitor to just go dead. However, it acts like its internal tables have gotten reversed or something.
You seem to know quite a bit about these ECUs. Any thoughts on what kind of failure could cause an ECU to somehow reverse spark advance?
And is there a possibility that something else still lurking there could fry the replacement ECU in the same way? Should I consider swapping out the distributor too?
From what I understand, the main things controlling the timing are the TPS and MAP sensor, though nothing should ever get the timing less then the 16.5 degrees in the tables.
I see 3 possibilities.
1. Your ECU is messed up.
2. The VR (TDC) sensor in your distributer is messed up, so the ECU doesn't know where TDC is.
3. There is/are poor conections between the distributer and the ECU.
I would troubleshoot to eliminate these possibilites. I don't believe there would be anything latent in the wiring harness that could fry the new ECU (though this is just a guess). I only have a CRX OBD1 manual and an Integra manual, but you can measure the resistance on the TDC sensor to test them. I wouldn't put too much faith in the test, it is only measuring for resistance and isn't exactly showing the TDC sensor has a nice output when the engine is turning. I know you hate throwing $ and parts at it, but after testing your ECU, a different distributer would be a good idea.
Though unlikely, I suppose another possibility is poor grounds. As a precaution, I'd clean the ground terminals on the main ground to the engine block and to the head,
I did check the resistance of the TDC sensor - it was within the range in the manual, though as you say that's not definitive. Flaky grounds is an intriguing possibility I hadn't thought of.
I did find that the mech who changed out the ICM had crimped a new connector on the control wire, and had done a lousy job. It wasn't very solidly crimped. I put the ohmmeter on it and could see it intermittently go open as I wiggled the wire. But I've already fixed that, and it didn't seem to make any difference.
If the weather cooperates I'll check out some of this stuff today. Thanks!
That's a drag. It is odd you have these symptoms and either ECU is not throwing a CEL, but it does sound like you have a bad TDC sensor in your distributor. Have you checked to see if the VR sensor can be replaced without replacing the entire distributor?
I'd at least pull the TDC connector off your distributor and check to see if the contacts are clean before getting another one. On the bright side, at least distributors are easy to change.
Yeah, the lack of any CEL is what really gets me. I thought the diagnostics were supposed to complain if any of the sensors was out of range.
I tested the TDC sensor with an ohmmeter (as described in the Haynes manual) and it passed. The plug contacts looked pretty good then. I could give them a little squirt of contact cleaner, I suppose.
I think a reluctance pickup uses a magnet; if something weakened the magnet it might not work right. I wonder if there's a voltage output spec for that pickup somewhere.
Haynes says "if the [resistance] test results are not correct, replace the distributor," so I suppose the pickup isn't available as a separate part, at least not fro the dealer. But then you're supposed to replace the entire throttle body rather than just the MAP sensor, too, so maybe the pickup is available as an aftermarket part.
I'm really puzzled by this. I keep going over in my mind the admittedly rather meager knowledge I have of how these systems work, and it just doesn't make sense. Why on earth would the ignition *retard* instead of advance? And why no advance as the rpm increases?