One cylinder not firing

[Pierce]

Dropping my daughter off after a casual drive across Dublin I very suddenly found only five cylinders were firing. Got back home (20 miles) and found that no. 5 exhaust was cold. Compression test-pass. Leak down test - pass. Swap coil pack, swap plug, swap injector - no change. Removed inlet manifold and cleaned EGR system (anyway). Checked and compared signals on all front injectors and coil packs using an oscilloscope. Measured coil resistance again. Measured injector resistor values again. Continuity from ECU to injectors checked. So the question is: what is it that is right in front of me that I am ignoring/overlooking. I assume that the cam and crank sensors must be alright if all other cylinders are firing, it start every time except for a very distressing burble where 5 should be.
This is now a matter of pride. Several of my best friends are mechanics. I am not.
My 92 (early 93 UK registered) 111K mile sad bastard car (as my wife refers to it) needed a clutch when I bought it four years ago (that took me a while) and apart from a few small jobs has lived up to the 'unburstable' Honda label I have expected from a number of their bikes before I 'retired' to four wheels after a couple of spinal discs took offense to the abuse they had suffered.
It's off the road for the winter (Irish??) so sad bastard car has plenty of time to keep frustrating me before I bow to the now seemingly inevitable.
I had a Lancia Delta many moons ago and left it with a recommended mechanic to blow out the jets and got a bill for a valet (Pay the bill or I call the cops). That was the last time I paid for anything other than a puncture.
Any point in asking for HELP?
And before anyone asks, yes I have tried multiple search options to get a lead on this one, to no avail.

 

[JD Cross]

Not sure this is relevant but some years back I had the five cylinder problem.
I tried the same things you have done and finally gave in and took it to my dealer's NSX tech
He found the igniter was at fault and replaced it.
No problems since.

 

[Pierce]

Hello again JD. Been there done that. Not only did I try my igniter on a friends car successfully but I tracked down one from ebay for $30 which also ran the five good cylinders. Apologies for not mentioning that. I'm sure he will let me install my ECU in his car if I ask but that to me is last resort, touching on defeat as I doubt it will solve the problem, at least I hope not, given the cost of replacement. For the record the ECU is a 37820-PR7-G02. I mention that in case the G02 detail is relevant. Manual 5 speed RHD.

 

[Old Guy]

JD picked my next choice for the likely candidate after you eliminated the coil through the coil swap! Now that you have eliminated the igniter as the problem, you are pretty much down to the ECU or the wiring between the igniter and the ECU or the igniter and the coil. See the attached diagram (obviously pre OBDII) . The white / yellow wire in the cable between the ECU and the igniter is the control for the power transistor in the igniter that switches #5 coil. The white / yellow wire in the cable between the #5 coil and the igniter is the wire that carries the coil current that is switched by the power transistor in the igniter. Check for terminator to terminator continuity on both wires and check that neither wire is shorted to ground. C heck to make sure that the termination plugs are correctly seated at the ECU and the igniter.

If the wires are good and the termination plugs are good, then it is pretty much down to an ECU problem. To confirm this, you would need access to an oscilloscope so that you could back probe the six connections between the ECU and the igniter while the engine is running. This would allow you to see whether the driver signal for the #5 power transistor is missing, indicating that the #5 driver circuit in the ECU has failed. If there is some good news to this, it is that because your engine still runs, it is likely that 99.999% of the ECU is OK, and that the only thing that has failed is about 35 cents worth of transistors and resistors which form the #5 driver circuit. If you have any friends who are electronics wizards, they should be able to confirm that the driver circuit has failed and effect a repair.

The easy test is to do an ECU swap; but, before doing that make sure that the white / yellow wire between the ECU and the igniter is not shorted to ground. If that were the case, then you risk damaging the #5 driver circuit on the swapped ECU!

If you don't have access to any electronics wizards, there are a couple of companies around that should be able to repair for a lot less than the cost of a replacement

 

[Pierce]

Come on lads, someone must have seen this before. Am I to believe that the fault I have is unique? Sorry to point out Old Guy, but you didn't read the challenge too closely so why not give it another try. I had already mentioned that I had used an oscilloscope. The next step which I have now taken was to attach a signal generator to each of the leads and see if anything appeared on any of the other wires. I had concentrated on all connections between the ECU and the heads - all six igniter leads before and after the igniter itself and I also checked all leads to the injectors. I don't class myself as a wizard but as I have earned my bread and butter for thirty years in hands on electronics this is not strange territory to me.
Update: I took the ECU out and installed it in a friends car and guess what -- perfect. I thought someone might suggest that one of the injection resistors was faulty but no joy there either.
I was hoping someone would tell me that the jiggling pin or the whatjamawingding was what had failed but I can't find any reference to either in the service manual and still feel it is something I am overlooking. I pulled out the timing advance this evening and without opening it it seems it is just a 5k pot so after re-checking the timing, the leak down and the compression I am going to be back to square zero, not even square one!
A bad earth somewhere I haven't spotted, low fuel pressure only affecting one cylinder???
Running the engine, I disconnected each injector (one at a time of course) and got pretty much the same response to each disconnection, including when I disconnected cylinder five.
I am trying to think of a suitable reward for whoever enlightens me although that would be only after the fact and as I said previously, it is a matter of pride so if the problem is so basic that I have to kick myself I doubt I will be so honest as to admit what the cure was. There again, you never know so come on try a few wild suggestions, I've run out of them.
T minus three months before I expect, sorry NEED, to have this problem sorted. Plenty of time.

 

[JD Cross]

This may sound simplistic but to have ignition you need fuel, spark, and compression.
If you've eliminated spark problems then it has to move to fuel or compression.
Is it possible the fuel rail is plugged over the bad cylinder resulting in no fuel to that injector?
Is it possible you have some kind of valve/ring/piston failure resulting in low compression in the bad cylinder?

 

[eddie_82]

Sounds like an electrical issue if all the mechanical components check out.

 

[Briank]

When you disconnect injector 5 and you get a change them its not a dead cylinder. Are you sure you have a miss? I am thinking you may have a jumped timing belt.

 

[Old Guy]

Going back and re reading your original post, you have swapped the #5 injector, the #5 coil and #5 spark plug and confirmed that they are operational - correct? You have also confirmed that the #5 injector is getting a signal applied to it and #5 coil is getting current switched on its primary winding. So presumably you have fuel and spark. It would be easy to double - double check the spark part by cranking the engine with the coil removed and the sparkplug inserted in the coil and grounded checking for a spark; however, if the primary current in #5 has the same waveform as the other coils, I would expect that it should be OK.

The injection and ignition timing signals originate from the crank / cylinder position sensor. If the sensor was faulty, it should generate a 4, 9, 54 or 59 trouble code on your pre OBDII car, which you have not reported. Also, as you allude to in post#1, its not possible for a faulty CKP/CYL sensor to just screw up #5 cylinder and leave the other cylinder’s timing unaffected, so I think it is fair to say that the sensor is OK. You know that the ECU works fine, the CKP/CYL sensor is OK so the overall ignition timing and fuel pulse durations should be fine and it sounds like you have done your due diligence in terms of assuring all of the appropriate electrical connectivities. You should have spark at the appropriate time and fuel being delivered.

Just to confirm, after all your testing and fiddling, the #5 exhaust is still cold? When you pulled the sparkplug out after running the engine, is it wet with fuel or dry? If its wet, I am at a loss. If there is fuel entering the cylinder there has to be air entering the cylinder and if ignition is occurring at the right time, then you should have combustion. We know you don’t have a big hole somewhere because your compression and leak down tests were good. However, if the sparkplug is dry, then there is no fuel entering the cylinder. If the injector is firing and the plug is dry, I think that pretty much means that the fuel is not entering the cylinder which suggests a valve problem of some kind (I am leaving it as ‘some kind’ because I am having trouble conceiving of a failure that would result in non operation of both intakes or both exhaust valves). With a lightly damped vacuum gauge, you might be able to pick up on non operating valves by tapping into an intake manifold port and watching for a pressure fluctuation; but, it will be pretty difficult. You have a scope, so it might be possible to see this by connecting the scope to the signal line of the MAP sensor. I have a 4 cylinder engine with Megasquirt and have done this on the MAP sensor and you can see the pressure drops associated with the individual cylinder openings (I am mostly retired and like a 10 year old, my mind sometimes wanders off into the I wonder what area). Whether or not this would be successful will depend on how much dampening the NSX intake manifold provides and whether the NSX MAP sensor has a dampening orifice on it. If the pressure drops associated with individual cylinder openings are visible on the scope, you would be able to pick out a ‘missing’ cylinder. The only other way to test this hypothesis is to pop off the valve cover and look; however, I personally think this hypothesis is so flakey that I would not suggest doing that at this point. I also think non operating valves would have messed up your compression test.

To JD Cross’s comment on the plugged fuel rail, if the plug is coming out dry, maybe that is a possibility. You know that the #5 injector is good and that the main part of the rail is clear since there is fuel delivery to the outside cylinders so it would have to be a blockage right where the injector pushes into the rail. Seems low probability, but???
<o
</oYou seem to have checked the obvious and not so obvious. I suggest the next step involves a Shaman or perhaps reading some chicken entrails.

 

[A.S. Motorsport]

After reading the posts briefly, (lunch break :tongue

Let me get this straight,

Swapped and confirmed working properly
- Plug
- Coil
- Injector

Now can you confirm,
#5 Injector clean and nozzle firing correctly?
ICM Module on throttle body functioning properly? (swap it for a known good one)
Can you check fuel pressure give me the figures? (Should be 46-53 PSI)
- If incorrect first check FPR located on the rear bank, than check pump.
- Fuel filter still good?

When checking resistors,
- Fuel pump resistor (unlikely)
- Injector resistor
- Ignition resistor


If everything "simple" checks out,
Rotate crank to TDC on cyl #1 , remove valve covers.
Inspect belt and confirm cam gears alignment.

Make sure the adjustment bolt has tension, also always rotate the crank clockwise.

To make sure we're all on the same page Honda numbers are per bank, not crossed like on certain American engines

 

[shawn110975]

I have a customers NSX ( AUTO ) in my shop right now and 4,5,6 are running lean and you can feel Cyl 6 misfireing I have replaced everything the last thing is I have sent the injectors off to be rebuilt/flow tested etc.

Idle is rough at stand still, but WOT it feels perfect only at idle you can feel the wheel shake, and it lunges forward slowly at a stop sign, I have replaced every sensor, its practicly a new engine LOL compression was good for all 6 cyl,

cyl 1,2,3 brand new spark plugs are clean and perfect after 4 miles and 4,5,6 are black as night. I am gonna check the fuel rails today see if they are clogged on that side, now one thing to note I have found plasti dip inside all the vacuum hoses
as he colored all his hoses etc. I think plasti dip might have clogged that fuel rail not sure yet its driving me bonkers

 

[Old Guy]

I have a customers NSX ( AUTO ) in my shop right now and 4,5,6 are running lean and you can feel Cyl 6 misfireing I have replaced everything the last thing is I have sent the injectors off to be rebuilt/flow tested etc.


cyl 1,2,3 brand new spark plugs are clean and perfect after 4 miles and 4,5,6 are black as night.

Did you get that mixed up? If 4, 5 & 6 are running lean, the sparkplugs should not be black as night.

Did you replace or check the O2 sensors? The NSX ECU uses the O2 sensors in front of the catalytic converters to trim the fuel mixture. The ECU may be able to apply different trims to the front versus back cylinders which might cause all the front cylinders to run rich / lean relative to the back cylinders if one of the O2 sensors is sending a bad signal. Normally the ECU would attempt to trim to 14.7:1 and normally a bad sensor would throw an out of range error code (if it is OBDII, don't know about pre OBDII). Its a last ditch suggestion!

If you have dealt with the O2 sensors already, then I default to the same advice I provided to Pierce, time to consult the chicken entrails!

 

[Pierce]

Lads, I really appreciate all the attention. Apologies that other duties are not going to allow me to solve this very soon as I probably won't get a chance to do much in the next week.
However a few more details. O2 sensors -not an issue as I have a Comptech straight through exhaust system although the sensors are still there - no cats thanks to Irish nonexistent smog/emmision laws until after 93.
Back to the original event, it happened suddenly rather than gradually, so in that respect everything agrees with Eddie's electrical diagnosis except the facts (wires).
I think you are referring to the Igniter, Adnan (A.S.) when you mention the ICM module on the throttle body. I covered that one. I will check TDC as per the cam gear markings to confirm these are all aligned. Also I fully understand why you mention clockwise.
I have not seen a single trouble code since this started. The engine starts on the button like usual which I don't think would be the case if the TB had jumped a tooth or the fuel pressure was low.
I must now wipe my memory banks using (industrial) alcohol and start afresh. I will record each episode and eliminate the impossible (?) even if it means paying good money on a service manual on chicken entrails - compliments of KFC.
I have already discussed with a full time mechanic friend. I may ask him just to have a listen. You never know, sometimes it is best to leave it to the professionals but that would be.............

 

[Old Guy]

.
However a few more details. O2 sensors -not an issue as I have a Comptech straight through exhaust system although the sensors are still there - no cats thanks to Irish nonexistent smog/emmision laws until after 93.

My comments about O2 sensors were with respect to Shawn's half rich / half lean engine. A cylinder missing in action would not be related to an O2 sensor problem.

If you are going to consult chicken entrails, they have to be uncooked. No cheating with KFC!

 

[ftuhy]

OK, a WAG & chicken entrails idea. But first, my thoughts on potential problem areas, many already expressed elsewhere:
1. Valves not opening & closing - do a simple compression test. (Can't imagine how this would happen, but ......)
2. Valve timing off/TB slip- as noted elsewhere, not likely if 5 other cylinders are working
3. Coil pack not firing - but waveforms look OK & pack works in other cylinders
4. Fuel injector not injecting - somewhat ugly to check.
5. Fuel injector injecting or coil firing at the wrong time. Not sure how with 5 good cylinders - Bad ECU/ECM?
6. Bad connector somewhere. +12V is connected to the injectors or coil packs via common circuits, so its not a bad 12V lug. Ground is not directly involved, at these devices, so its not a bad ground lug issue.

For item 3, when there was such a thing as spark plug wires, I would use a small neon bulb (the physical size of those 2.5V incandescent Christmas tree light bulbs) with one side grounded and the other side going to a wire wrapped about 4 to 6 turns around the spark plug wire. By induction, the neon bulb would light when the spark plug fired. Its seems to me that you could do this by wrapping the wire around the tube of the coil pack in the area of the top of the spark plug. The fine wire (26 gage/.4 or .5 mm) should fit past the coil pack insulator. Crazy idea, but this is chicken entrails time. If you don't have access to the neon bulb (I probably have a couple in my electronics parts bins), I could try this myself, then mail you a bulb.

Or maybe, just get out a bottle of your favorite wine/liquor/whiskey/etc and contemplate other solutions.

 

[docjohn]

now you have a vigor engine..

 

[A.S. Motorsport]

I have a customers NSX ( AUTO ) in my shop right now and 4,5,6 are running lean and you can feel Cyl 6 misfireing I have replaced everything the last thing is I have sent the injectors off to be rebuilt/flow tested etc.

Idle is rough at stand still, but WOT it feels perfect only at idle you can feel the wheel shake, and it lunges forward slowly at a stop sign, I have replaced every sensor, its practicly a new engine LOL compression was good for all 6 cyl,

cyl 1,2,3 brand new spark plugs are clean and perfect after 4 miles and 4,5,6 are black as night. I am gonna check the fuel rails today see if they are clogged on that side, now one thing to note I have found plasti dip inside all the vacuum hoses
as he colored all his hoses etc. I think plasti dip might have clogged that fuel rail not sure yet its driving me bonkers

Bit off topic, I have had nearly a exact same issue with a Ferrari a couple of weeks ago.

 

[shawn110975]

A guy I know that owns Chily Performance here in tampa they only work on high end cars told me that he has seen that in Ferrari's and its normally the cat is clogged he said cats are more trouble than they are worth

 

[A.S. Motorsport]

Yes especially on the 90's cars they are highly restrictive and cause issues.That issue wasn't related to the catalytic's on the car though. Anyway back on topic. How did you test the ICM module? (swap or just measuring?)

 

[Pierce]

Sorry Adnan, no joy there. I installed the ICM (ECU?) as with the igniter, in another 92 car and it worked beautifully.
Back to square zero. Valve/belt timing - the 5mm punch slipped easily (about 35mm) into the alignment holes in all but the rear exhaust where it only went in 25mm. Looking at the TDC mark it didn't quite line up with the timing cover key but this has a bit of damage so I assumed the cover was slightly out of line showing about two degrees off TDC. The next step was to get the strobe light and see whether that would reveal anything. I don't know why I didn't think of it before but I have to admit I have been misleading you a little bit. When I initially drove the car home I had to immediately go back out so it was the following day that I plucked up the courage to start the engine and that is when I felt exhaust pipe no. 5 as cold. That may have been partly because I didn't run the engine for long. I discovered after I was strobing yesterday that no. 5 is firing but it is still not right as I was able to touch that pipe but the others (4&6) were far too hot.
Back to the strobe (which I notice that no-one suggested which is strange - except that I have to include myself) and even when fairly warm I am looking at the engine running at about 30 before TDC. It occurred to me that maybe (just maybe) the timing belt had jumped let's say just one tooth on the crankshaft. This would still allow all four cams to line up correctly but would, sorry might, explain the strobe issue. I spoke to a helpful Hondawise mechanic (Ken) today who insisted that even with only one tooth off, the car could be made to run but would definitely not idle. I'd like to hear of anyone who has had a jumped belt in case they know otherwise.
Back to the manual again it seems. I've seen some dispute about how much variation the advance/retard ignition adjuster can have with claims which Ken agrees with - +/- 2 degrees. Has anyone seen a case where this was disconnected as in an open circuit. I'm about to find out. If I simply disconnect and see no change in the strobe position (-30) I think I might be onto something. I've already had a look at it and measured the resistance at 5K with the sweep centered, what I mean is 2.5K either side of the centre contact. Nothing to lose anyway.
Turkey entrails are popular this time of year, I wonder.......

 

[Honcho]

Pierce-

Got your PM. I agree with Adnan and suggest you visually confirm cam timing using the marks on the cam wheels. The NSX will run ok and idle even if off by a tooth, though as mentioned before this problem usually affects the entire bank and will show up in a compression test. What do you mean by "pass"? The service manual range is not helpful. A healthy NSX with 100k miles should have a dry compression of 190-215 psi at sea level (assuming it has not seen boost). If your bank is less than this, that would tell me something, especially if the unaffected bank is within the healthy range.

My other thought is to remove the fuel rail and check for any blockage at the injector terminal. But, I would only do that after confirming a dry plug. It sounds like you have spark and there is no problem with the ECU or ignition circuit. I would look into the timing adjuster, but if there was a problem, you would have a CEL #18 . Since you don't, that is way I'm thinking cam timing is off. The key for me is your 30 deg BTDC timing- that is beyond the adjustment range of the adjuster (13 - 17) and suggests a timing belt issue. I really do think you're off by a tooth on the front bank.

 

[Pierce]

Problem solved. I have only myself to blame. Interested? Read on. Swapped out the entire control box. No joy, so I decided to start on page 1 of chapter 11 of the bible according to NSX. It took very little time to reach O2/lambda sensor test. Lo and behold the heater resistance for the front sensor (should be 10-20R) was only 6. When I put a scope on it there was no signal, just a small DC bias, unlike the rear which had a nice (intermittent) sine wave. New sensor installed. Initially no change but as it grew a bit warmer the engine started to hunt noticeably, up to 13/1400 and down to 1200 quite slowly. Don't ask me why but I put my hand down toward the sensor and felt a noticeable breeze.
Whipped (?) off the Comptech header and found a neat crack where the weld had already been redone, badly. Bingo.
My first instinct had been to pull off the manifold but I couldn't understand how one pipe could break but the other two would remain intact. Now I know. I also know that I have a new timing belt - which had been noticeably loose between the crank and the front exhaust bank. A new pulley and crank sensor; and I've fixed the tiny leak that has been spraying steam onto the top of the oil cooler directly above the filter for many years and a myriad of other small issues like the two bolts missing from the timing cover and the inlet manifold has been drained, and the EGR has been flushed and.....I forget.
I hope this may be of use to somewhere even if only to remind them of Gary Kentosh's remark in his well thumbed write up on changing the timing belt, about understanding that if it takes half an hour to remove a bolt, then so be it. Amen.

 

[ftuhy]

Congratulations on your endurance. I suppose you have found or are about to find a good use for some nice Irish Whiskey to mark completion of the project.

 

[Pierce]

I suppose I could chuck a couple of measures of Paddy into the tank instead of STP but my recollection is that the only gasket it blows is my wife's. Let's not go there. Endurance is an interesting word. Subtle and inherently an understatement, just like....... an NSX.