Need some VERY EXPERIENCED HELP (misfire) (Looking for NSX GOD type HELP)

Joined
1 September 2005
Messages
540
I'll cut straight to the case, and try to make it the least confusing as possible.

I've been a member/owner for over a DECADE now. But I need some no sh!t PRO help.

Issue is a misfire on cylinder #5 . I know this is not uncommon, but please read the post to view the lengths I've gone to for resolution.
It was discovered during what should have been an easy dyno tuning session. (The car produces safely 405rwhp to the ground).


Here are the details:
The misfire was diagnosed through using the AEM EMS Series 1 that is installed in my car, and a laptop with the AEM sfwr installed which was viewed/identified by a seasoned NSX Tuner (Also a respected Prime Member abroad). Unfortunately, I no longer have access to this person and am trying to resolve this issue on my own.

It was determined that cyl #5 was incurring, and also being displayed via the AEM EMS. There are NO OTHER cylinders having misfire issues. NONE.

Trouble shooting steps performed:
Bought new, and replaced all 6 spark plugs with appropriate NGK plugs for my car. (Comptech supercharged, with HB Pulley, and mild Methanol injection)
Bought new, and replaced all 6 ignition coil packs (From Acura)
Swapped out the ignitor with a known "good" ignitor from a donor car.
Swapped out the AEM EMS from a car with an extremely similar tune.
Removed All 6 RC Engineering Injectors, sent them to RC Eng to be flow tested, and retuned as needed.
Physically viewed the fuel rail outlets that provide for cylinder #5 (but found nothing)
Compression Tested EACH of the 6 cylinders. (The individual cylinder's psi ranged as follows: 6 - 240lbs, 5 - 240lbs, 4 - 240lbs, 3 - 240lbs, 2 - 240lbs, 1 - 235-237lbs respectively).

Each cylinder held pressure well while performing the compression test, and all cylinder readings were damn near identical. I performed the test twice. Each time, the same outcome. Thankfully "positive" readings, and no readings that varied nearly at all. ***Caveat: Cylinder #1 's readings were a little harder to get a good read on, due to obstacles preventing me from fitting my arm in to tighten the very rigid rubber hose of the Compression gauge. However, it read at 237psi on my 2nd attempt to deal with the pain in the ass (While cramming/contorting my arm down and dealing with the fun sharp metal obstructions as they effortlessly opened my hide till I was leaking red. (Heh... Was better than giving up, and half assing it.)

I have also pulled the valve cover off of the front bank head for cylinders 6,5,4,(facing front of car) so that I could attempt to view the valve springs (Intake & Exhaust). Was pretty amazed at how unbelievably neat, and clean it was inside the head. But to no avail could I find any traces of metal chunks, flakes, or any shimmering bits of metal that would possibly pointed to a damaged valve spring(s). It was pretty damn immaculate honestly. I am contemplating getting the black light liquid to use on the valve springs, just in case there's anything I'm unable to see with natural day light & naked eye.


With all of the above named items that I have verified, I am left now only with my limited knowledge and experience to think that it may be the electrical that leads to one of these two:
1. The electrical lead on injector on #5 cyl
2. The electrical lead on ignition coil pack leading to #5 cyl

Or God forbid, an issue somewhere in the electrical harness that goes from the car's interior into the engine bay. I have zero experience working with or diagnosing auto electrical, but I do understand the fundamentals to a degree. Unfortunately I have not yet tried to use a Volt meter, or test light of any kind to try and get readings on the above named points #1 &2.


I'd like to keep this thread CLEAN, and to the point if at all possible for those who are kind enough reply and help a long time fellow NSX'er keep living the dream. Because at this point the dream is bordering on a living nightmare. And I don't want to be the guy who keeps putting it off until there are numerous new problems, ultimately putting the car on the eternal "I'll do it later shelf."

I know the car is otherwise sound as a pound, and with only 60K miles...... I'm not throwing in the towel on what is still debatably one of the better street & track capable NSXs (counting it's 24yr old age factor) in my tri-state area that hasn't had the engine cracked open.

I'm just a guy who is at the depths end of his experience level, and at the end of his rope (on patience). I'd really like to figure it out with some experienced & knowledgable people's input/advice. Versus taking it to a local speed shop to basically re-diagnose everything I've already replaced or named as verified. Only to find out the problem is a known issue that ONE or TWO or several of you has previously encountered, and potentially save me THOUSANDS in unnecessary labor fees.

If I've not described, or missed anything at all that could be the culprit, by all means.... Post it up. Thank you to anyone who can help.
 
I may have missed it, but under what conditions does it begin to misfire? Is it only under load? Does it happen at idle?

Also when you say your AEM EMS indicated it was cylinder #5 what exactly do you mean?
 
Is this a mis-fire, which most people interpret as a disruption in the combustion process which make occur infrequently or frequently in the case of severe misfire, or is this a case of cylinder #5 being completely dead? If you are looking for chunks of metal under the valve cover that suggests that the cylinder is completely dead.

As to testing the electrical connection to the coil. Pull the #5 coil and sparkplug. Pull the fuel pump fuse. If convenient, release any residual fuel pressure in the fuel line. You don't want the engine starting during this test. Stick the sparkplug in the coil, connect the electrical connection to the coil and press the sparkplug ground electrode against the engine block. Have someone crank the engine. If you see spark, then the electrical connections to the coil are OK. For the injector, there are test lights that you can buy that fit between the injector and the injector plug for confirming the electrical connection if don't want to get into using a voltmeter.

I have a similar question as E-Nough Logic. What feature on the AEM was used to diagnose the misfire? Does the AEM have a crankshaft speed versus cylinder firing display?

Update - rather than pull the fuel pump fuse, disconnect the injector resistors (assuming your RC injectors are low Z injectors). This eliminates the need to release fuel pressure before doing the coil test.
 
Last edited:
The misfire at the beginning was only noticeable at engine idle. When the throttle was depressed/opened the misfire was not noticeable. The person performing the the dyno knows the AEM EMS Series 1 that is in my car intimately, and has had well over 30 tuning hours invested into it. This time was invested when the exact same AEM EMS was his, in his NSX. I then bought my entire Comptech supercharger setup, and ALL of it's supporting parts (AEM included with tune). All of the following parts were purchased from, and installed by this same person. I have since had "roughly" 5,000 miles of driving with this equipment on my car.

So while I do not know "exactly" if the AEM EMS software that is on his laptop & my device had the software capable of indicating an EXACT cylinder. I am 100% positive that I was told which bank was having this intermittent misfire. Together we then started the engine, let it warm up and drop to my particular car's proper idle rpm. We then systematically began to pull the coil packs off of each cylinder individually 1 by 1 until we pulled the one that caused the engine to indicate the prominent cylinder of being suspect. It was cylinder #5 .


After time the misfire (still intermittent) began to worsen quickly. Primarily under load the misfire seemed to disappear or be less pronounced. However at idle, the car now has misfire that is very noticeable, and on frequent occasion will die when rpm's are at idle or near idle. I have tried to set the cruise control at speeds ranging from 40mph - 60mph, but do not recall the misfire being to the degree as to cause the car to shutter or symptoms that usually accompany a severe misfire while an engine is under a continuous load. Again, this is from my limited knowledge and experience.

In fact, when under load per say while traveling at 60mph. When the intermittent misfire happens it does not become greater, and is still an item that an average driver may possibly not even notice if they were not expecting or aware.


As for gas, oil, and the like.... I make it a point to always use the same gas location & octane grade. My only fuel additive is stabil for winter storage. I'm always positive to ensure the additive enters the engine before storage, and I'm also cautious to monitor the volume of fuel stabilizer that is added. Then again ensuring new fuel is added upon the next driving session as to consume all of the previously contained fuel & additive from the previously induced storage.


Fuel concerns:
I have had 4 other cars which I also refueled with the exact grades of fuel, during the exact same times as my NSX. All of which seemed to run without issue.

Electrical concerns:
I should be able to do the spark plug ignition test without issue. (I will definitely try this.)
As for the injector, I'm comfortable enough with a volt meter that if I know what voltage the injector should be receiving, I can confirm it. (Again, I will attempt this once I get some items put back together....I'm just super hesitant to put things back now.


Let's say for conversation, and any unnamed items of question... Let's say the above electrical items are performing to spec and doing so 100% of the time.
What are some other items I can look into?


Thank you guys, and I fully plan to keep this thread as up to date and detailed as possible until the issue is resolved and the car is 100% fully operational again as it should be.
 
If the misfire is intermittent, I would not bother with the electrical tests just jet. The probability of catching an intermittent electrical problem when testing is pretty low. You really need to have some kind of fault logger which would trap the failure. With the engine running, I would try wiggling the exposed portions of the FI and ignition wiring harness to see if you can initiate a stumble. I would pay particular attention to the interface between the native wiring in the NSX and the AEM ECU. Honda wiring and connectors are pretty robust. I don't have a clue about AEM stuff.

The algorithm that was used to identify cylinder #5 would be really useful. Particularly if it turns out to be a false lead. Since you seem to have addressed the usual suspects for a true misfire, when was the last time you had your valve clearances checked? The NSX seems to be quite sensitive to valve clearance.
 
If your car is like mine, every electrical connector I unclip then clip back together, never makes a solid connection... Shad also experienced this when installing my ctsc. It has been a thorn in my side...
 
I've seen similar misfires with a lot of plug and play injector clips. What injector are you running and what type of connection/connectors?

Also make sure you have all the required grounds. I know the CTSC sometimes the ground connections have had some complications.

You can, to a certain extent, switch the ignition and/or fuel injector outputs to fire another cylinder(s). Ignition would be very simple on an AEM series 2 but with the Series 1 it is wasted spark which complicates it. You would reassign the ignition outputs in the AEM software and would need to either use 4x spark testers or could get by with 2 if you felt comfortable repinning the igniter unit or ecu. This would basically test the wiring harness if you have eliminated the coil packs and injectors. I can get into which wires need swapped or how to set up the software if it's something you want to dive into.
 
Last edited:
Sounds like a wiring issue since you have replaced the common items.
Maybe disconnect the ECU connectors & other #5 Connectors & make sure the pins haven't been pushed back. Have seen this happen a few times.

As Old Guy says check valve clearance on #5 . and maybe check you can move the Vtec rocker.
 
Sounds like a wiring issue since you have replaced the common items.
Maybe disconnect the ECU connectors & other #5 Connectors & make sure the pins haven't been pushed back. Have seen this happen a few times.

As Old Guy says check valve clearance on #5 . and maybe check you can move the Vtec rocker.
Yep, I just recently pulled my hvac blower out for service, cleaning, and after reinstallation the blower wouldn't work. Then it would only work at a very slow speed. The culprit, those gosh darn Oem connectors from 1990....
 
Can anyone recommend a good point of contact, website, or company that could help me locate The Main Electrical Harness for the 1st Gen NSX that would be in reliable condition? Doing some research, and most of my searches are varying widely. Pretty much to the point that I am not certain what/which parts I should be looking for to begin with.

They all look like 2 things. A nightmare, and a mess.....
https://www.google.com/search?q=199...c6XCJsQ_AUIBygC&dpr=1.3#imgrc=psL1hBZ-sAr0MM:
 
I don't think there is anything called a main electrical harness. Got to Delray Acura (or one of the other on-line Acura parts vendors) and call up their on-line parts manual.

http://www.acuraoemparts.com/

Look under the heading Wire Harness and / or Engine Wire Harness. You should be able to find the part number for the part you want.

Get your wallet out.
 
Last edited:
Since the misfire is very noticeable at idle and it occasionally dies, why not try to plug the OEM ECU back in temporarily? This would eliminate the AEM unit as the culprit if you experience the same symptoms.

Get the engine coolant up to normal temperature and then quickly unplug the AEM ECU and plug in your old OEM one. Then idle and observe. You don't say what capacity aftermarket injectors you have installed, so it will probably run super rich. If you only do it for a few minutes when the engine is already warmed up it shouldn't wash the cylinders too much. What do others think?

Just idle though. I wouldn't drive it and certainly don't build boost.

Like others have said, it sounds like a ground or connector is working loose.

Good luck.
 
Last edited:
An update that I wish were the resolution. First, I want to say thank you to everyone whom has contributed input towards the eventual resolution.

Here's what I have replaced, and or swapped with donor (running cars) NSX parts.

>New plugs
>New plug wires
>New ignition coils
>Swapped the Igniter
**Swapped the GEN1 AEM EMS from donor car with similar tune. (we'll circle back to this item, as I did not perform this temporary swap. )
>Checked the Intake & Exhaust valve springs (Inspected for cracks, debris, or any type of metal fatigue).
>Installed a new valve cover gasket on the bank nearest the cabin (where a suspected misfire is occurring on Cyl. #5 , after viewing the valve springs)
>Checked the entire AEM Methanol Kit (also disco'ed the kit entirely, and sealed in inlet ports from the meth injectors.)
>Compression tested All cylinders (all have great numbers, see my 1st post)
>Pulled the RC Engineering 550cc Injectors, and sent all back to RC Eng. to be flow tested, and they also recalibrated all 6 injectors. Also provided new rubber O-rings.
>Performed a vacuum leak check EVERYWHERE around the engine, engine compartment, etc. (I used Carb cleaner, and then waited for the engine to cool, and checked again with a propane bottle and rubber tube long enough to reach EVERYWHERE. No vacuum leaks were discovered.
>All Engine & Trans. grounds were removed, and cleaned to a nice bright new copper state.

During the vacuum leak test, I sprayed a touch of the carb/throttle body cleaner into a port on my intake pipe that is located before the throttle body. The RPMs went up as expected, AND there was no indication of a misfire or surging idle. The engine bumped the RPMs up slightly, the misfire disappeared and the engine ran smoothly until the carb cleaner burnt off.

**I then pulled the AEM EMS after reading the post by Mac Attack. I still have my OEM ECU, and plan to swap out the AEM EMS to see if this allows the misfire to disappear. I noticed upon removal of the GEN1 AEM EMS that one of the connectors was damaged (the squeeze clip portion was 3/4 broken off). But I did not see or notice any pins inside the connectors, or any wires having issues, etc.. The only thing that caught my eye was that 2 of the 4 AEM EMS Manufacturer Warranty stickers had been removed, the third Warranty sticker looked to have been cleanly cut, and the 4th Warranty sticker was the only one intact. Which would lead me to speculate to a degree that at some point prior to my buying it from the original owner, that the AEM EMS case had been opened. This caused me to question the AEM EMS, and I will be swapping it with my OEM ECU as aforementioned.

If the OEM ECU allows the car to run smoothly, without the noticeable misfire... I'll at least know I need to buy a new one. If the misfire remains, and there are no signs of change in regard of this issue... Well, then I'm no worse off than when I began.

My only concern with the OEM ECU is that I have the Science Of Speed performance chip in it. But, I'm fairly certain that it will at least allow the car to idle. If anyone has any input on the work performed so far, feel free to contribute to this thread and I'll try my best to provide detailed responses.

Once I swap the AEM EMS, and try the OEM ECU, I will be at the end of my known ability as well as tools, and resources. This late update is due to a physical relocation on my part, and the car was driven from my now sold house, to a rental house and garaged. Then driven again, directly to my new house at present time. During each drive, I experienced no issues while driving and the misfire/rough idle only occurred during moments of the car being at idle RPM while at stop lights/stop signs.

Which made me question if our cars have an IAC, or MAF that could possibly be bad???
 
It seems from the beginning, you are SURE that there is a problem specific to cylinder #5 . Is the below described cylinder balance test your only troubleshooting to confirm this?

We then systematically began to pull the coil packs off of each cylinder individually 1 by 1 until we pulled the one that caused the engine to indicate the prominent cylinder of being suspect. It was cylinder [URL=http://www.nsxprime.com/forum/usertag.php?do=list&action=hash&hash=5]#5 [/URL] .

The reason I ask, is because this is a quick and primitive way to check for a simple misfire specific to a cylinder. This test can also provide false positives when you are chasing a problem that is either intermittent, or worsens, etc... Trust me, this has happened to me. I am not saying that you're incorrect in troubleshooting #5 as the culprit, I just want to make sure you aren't narrowing our troubleshooting options as a result of 1 test that may not have been reliable.

When I first got my NSX, I had an intermittent misfire under load, and it turned out to be a failing fuel pump, that only exhibited problems under 4K RPM. I'm not saying that is your problem, but it took me opening my eyes outside of my initial "cylinder level" thoughts to find the culprit.

Hope this helps in some way or another...
 
A misfire that only occurs at idle; but, not under load does not sound like an ignition system problem. What are your idle AFR values? Do they bounce around a lot at idle? You say that you installed RC 550 cc injectors. Larger injectors automatically add the potential for poor idle performance because with larger injectors you have to shorten the fuel pulse width to deliver the same amount of fuel as the smaller OEM injector. Injector flow rates can become very non linear at very small pulse widths. Some injector vendor will give you some guidance on the minimum recommended pulse width. RC does not publish the linear range of their injectors so you may have to email them to see if they will give you that data. You should be able to connect up to your AEM and display the fuel PW at idle. If those idle PW values are at or below the injector's minimum PW values you will likely never have a stable idle with that injector. RC does publish the injector offsets (they call it latency) for their injectors. Make sure that the offset values are correctly entered into the AEM and that the voltage correction for the offset is also correctly entered. Incorrect injector offset values will also severely screw up your idle performance. Do not switch back to your OEM ECU unless you also switch back to the OEM injectors. The OEM ECU will never, never, never operate correctly with larger injectors.
 
Last edited:
If you've swapped out the AEM ECU twice now with other boxes with identical tunes and had the same intermittent idle issues, then the AEM is probably not the issue.

My comment on swapping it out for the OEM ECU would be just for idle checking. I now see you've posted your injectors are 550's, or more than double the capacity of your OEM injectors. Yes, low non-linear flow rates, dead time, and voltage correction offsets will all be different, but, it should idle (somewhat I guess) and run very rich since the Honda engineers most likely did not account for O2 feedback that far out of whack. Hence my previous comment on doing this after the engine is warm to further reduce washing down the cylinders.

The important thing to note when switching injectors/ECUs is if they are set up as high or low impedance. Your '92 came from the factory with low impedance injectors. There is an injector resistance box to lower the current draw and protect the injectors and ECU. If your 550 RC injectors are low impedance also (you can measure between the pins with an ohmeter to verify they are around 2.5 ohms), then the OEM ECU will acceptably power them. The car probably just won't idle well.

That's a last-ditch attempt to eliminate the possibility of it being the ECU though. It sounds like you've already done that.

Can you manually wiggle the wire looms going into the AEM when the car is running to see if it improves or makes the idle worse?
Did you change the TW coolant sensor?
Clean the TB really well?
 
A misfire that only occurs at idle; but, not under load does not sound like an ignition system problem.

Under post #5 , the OP said under load the misfire seemed to disappear or be less pronounced.

Can you log the data from your AEM when it is "missing" under idle and upload the file here?
 
My comment on swapping it out for the OEM ECU would be just for idle checking. I now see you've posted your injectors are 550's, or more than double the capacity of your OEM injectors. Yes, low non-linear flow rates, dead time, and voltage correction offsets will all be different, but, it should idle (somewhat I guess) and run very rich since the Honda engineers most likely did not account for O2 feedback that far out of whack. Hence my previous comment on doing this after the engine is warm to further reduce washing down the cylinders.

My comment about the OEM ECU never operating correctly with the RC 550 injectors was related to the huge difference in fuel flow rates. Since my car is a 2000, I didn't even think of the high impedance versus the low impedance injectors on the early cars. My comment about never operating correctly was related to the resultant fuel mix problem associated with the swap. As you suggested, if he managed to effectively hot swap the OEM ECU for the AEM ECU , he might be able to get the ECU to go into closed loop immediately after start up; however, my recollection is that the OEM ECU fuel trim is limited to something like 20% or less before it limits out and triggers an error code. With the 550 cc injectors, 20% trim is not nearly enough and the engine will, as you correctly noted, still be running very rich. I think that it would be rich enough to introduce some new / different idle issues that would make diagnosis difficult. Plus, there may be so much fuel delivery during cranking when the ECU is typically running off of a fixed temperature based cranking pulse width with no EGO correction that starting may not happen.

I think your suggestion to log the engine performance during idle is the most effective suggestion. At minimum, a log of RPM and AFR on the two sensors would yield a lot of information about the engine operation. If there are true misfires occurring the AFR will spike high during the misfires. Having both AFR signals might allow you to at least narrow down which bank the misfire is occurring on (if there is a misfire) and the RPM should drop following the misfire providing the signal is not smoothed too much. If the AEM 1 is full sequential and you can log the ignition outputs at the same time the log might allow you to tie RPM / AFR changes to ignition events which could be used to point to a specific cylinder.
 
Last edited:
In quick response to Mac & Old Guy, (First off, Thank you guys)

I won't be swapping out my AEM with the OEM ECU. After speaking with a member of SoS, I was told that I should definitely not swap "IN" my OEM ECU, as it would probably foul my plugs and cause more issues than it would help. In short I was told, "You don't wanna do that with your setup." Which was also followed by a suggestion to upgrade my EMS to a better/current EMS, since my current AEM Gen1 is old beyond AEM support, etc..

So… now I am looking for my AEM CDR, so I can load to a laptop and begin educating myself further on this dinosaur piece of electronic tech (the AEM, not my old ass car).
Once I've recorded some info, I will absolutely post everything possible. Too bad the GEN1 isn't able to tell me if a specific cyl is misfiring. Only that there's been a misfire. This info again was told to me when this problem began, while trying to figure out the problem. I'm grateful for the info, and wish I had paid to just have the issue troubleshot by them.


**Fuel Pump.... My car has a Walbro Fuel pump, and I recall clearly that during the installation of the CT Supercharger, and all of the supporting parts (AEM EMS, RC Injectors, Fuel Rails, FPR, etc.) I was informed that a new Walbro Fuel pump was needed, and when the installer received/installed it, they told me it was bad and had to send it back for a new Walbro Fuel pump.

So, while all of this sucks a fat one... It's nice to have further support as I learn, and hopefully resolve this issue.
 
Back
Top