1998 NSX misfire on 4,5,6

[michaelwoodcock]

Hey guys! I'm solving an odd problem and figured I'd post here for anyone who may encounter a similar problem in the future. Here's what's going on:
Symptoms: car idles rough and has check engine light illuminated
Car was being driven when problem developed
we have the following diagnostic trouble codes present:


Current Fault Log
------------------
P0300: Random/Multiple Cylinder Misfire Detected
P1304: Misfire cylinder #4
P1305: Misfire cylinder #5
P1306: Misfire cylinder #6

Pending Fault Log
------------------
P1399 (what does this mean?)


Tests performed:
Compression: over 200 PSI indicated on all cylinders except number 4 which had 184 psi.
swapped coils from front to rear bank, no effect.
inspected O2 sensor readings on front and rear bank, O2 sensors both seem to be functioning properly, and indicate we are getting proper air fuel ratios with very little short term fuel trims on both banks.

I have no direct experience diagnosing misfires on the NSX until now, does the NSX detect misfires via crankshaft speed? if so I know that the crankshaft is actually slowing down when it should be speeding up. If the crankshaft sensor works off a toothed wheel with many teeth this would be the case. Does the NSX detect misfires via a knock sensor? if so, I may have more research to do.

My hypotheses are as follows:
bad sensor on bank for 4,5,6, (Ignition timing may be off, spark plugs firing too soon, firing too late, etc.)
clogged cat for bank 4,5,6, (I think this is unlikely)
mechanical engine issue for bank 4,5,6 (a shop stated that the car's timing belt may have jumped time, but I find this unlikely, I will double check this though)

Has any body else experienced a misfire all along one bank? I searched but could not find a similar case. If not, stick around for some entertainment while I solve this issue :biggrin:

 

[Valhalla]

I have experienced this before. Here is some light reading for you.....

http://www.nsxprime.com/forum/showt...iple-misfire-codes?highlight=valhalla+misfire

 

[Old Guy]

The OBDII cars have a third terminal in each ignition coil which is monitored by the misfire detection module. However, prior to OBDII and the misfire detection module addition, the ECU still reported misfire error codes. I speculate that the pre OBDII cars used crankshaft speed to detect misfires and the OBDII cars got the added coil voltage detection module, so your car probably has both systems.

If you look at the DTC list in the service manual, you will find two sets of codes associated with misfires, P1201 - P1206 and P1300 - P1306. I am going to speculate that the 1200 codes are associated with the crankshaft speed detection algorithm and the 1300 codes are associated with the coil monitoring system. The reason I say this is that for the P1201-P1206 codes, Honda lists every possible candidates as possible causes. For the P1300-P1306 codes, Honda narrows the possible causes down to the ignition system. Full disclosure, the service manual does not state that P1200 codes are based off the crankshaft speed and P1300 are based off the misfire detection module - that is just me reading between the lines in the manual. The fact that you are not getting the 1200 codes suggests that you should first start examining the ignition system (which could include Valhalla's engine wiring harness misery). I would be inclined to look for common elements shared by the 4,5,6 ignition components.

You wrote P0300; but, I assume that is a typo and should be P1300 (random misfire) since there is no P0300. The service manual does not list a P1399 code for the NSX. Honda does use a P1399 code for some of its other vehicles and it is a random misfire error code. If you are using a generic code reader with a generic Honda code set, they can sometimes report odd results.

If you go to the NSX Wiki page and look you will eventually find a link to a .pdf version of the service manual for post 1997 cars. Get a copy and check out the procedure for trouble shooting the 1300 DTC codes. Much better than me giving you a laundry list of stuff.

 

[Valhalla]

Old Guy is a wizard with our electronics. Follow his path.

 

[jwmelvin]

+1 to that. [MENTION=26435]Old Guy[/MENTION] I like your theory about the 1201/1301-series codes.

 

[michaelwoodcock]

Thanks for all the tips guys! Im finally begining working on the car today. I may be chasing the wrong lead, but the car had non resistor spark plugs installed in the front bank. I removed and replaced with the proper NGK plugs but the misfire remains. is it possible non resistor spark plugs damaged the ignition control module/ignitor?

 

[Larry Bastanza]

OK, so the rear bank had the correct plugs? If so, I think you have good logic around the ignitor module. Any other owners close to you that you can swap out the ignitor? It would be great if you could do that, before you drop $300 on a new one.

HTH,
LarryB

 

[michaelwoodcock]

yes, the rear bank had DENSO's of the correct part number. I don't like Denso's, I changed them all to NGK's. The car went into a lazy shop that only replaced the easy to access spark plugs.

I just tested the ignitor/ICM and it turns out it's good. What I did was carefully swapped the input & output for cylinders 1&4 going into and out of the ignitor. my theory was if the ignitor was bad for 4,5,&6 doing this would move a misfire code to cylinder #1 . this misfire stayed with cylinder #4 so this hypothesis is weak.

Now I'm going to test wiring after the ignitor. A co-worker really wants to replace toe O2 sensors so we may replace those.

The onther possibilities are maybe the engine did jump time. although with very similar compresion numbers on all cylinders except #4 I find this unlikely too

I just performed another test: I made a temporary jumper wire from the BLK/YEL wire on a rear coil to a front coil and the misfire problem stayed the same. The wire to the spark pllug voltage detection module had a good ground on both banks. The only other option I can think of electrically is maybe the wiring from the ICM to the coils as high resistance somewhere. I also find this unlikely but will test shortly

 

[Larry Bastanza]

Also there is a ground lead very close to the ignitor, worth a check.......

 

[Old Guy]

Thanks for all the tips guys! Im finally begining working on the car today. I may be chasing the wrong lead, but the car had non resistor spark plugs installed in the front bank. I removed and replaced with the proper NGK plugs but the misfire remains. is it possible non resistor spark plugs damaged the ignition control module/ignitor?

Since I don't know the design details of the NSX ignition system, damage might be possible; but, I would characterize the probability as unlikely. In fact, if you check the NGK product literature they suggest the application of resistor type spark plugs on any car with electronic fuel injection because of the reduction in electromagnetic interference. That is a pretty blanket recommendation.

The other factor to consider is that the spark plug resistor is in the secondary (high voltage side) circuit of the ignition coil. It should have no effect on the coil primary current which is what is controlled by the igniter module. If anything, the use of non resistor versus resistor plugs might have influenced the operation of the mis fire detection module (no damage, just incorrect operation); but, again I think this unlikely.

The preceding does not imply that the igniter or misfire detection module is not the problem. It is just to suggest that the problem did not arise from the use of resistor plugs. I also note that if you were generating misfire codes; but, the engine was running just fine then maybe something was up with the misfire detection circuit. However, you state that the engine is running poorly so the misfires are likely legitimate and not being falsely reported by the module.


The igniter module itself is a pretty simple device. Its essentially a blob with 6 power transistors in it for switching the primary currents of the ignition coil. The power transistors tend to work or not work, not somewhere in between. However, ignition coil primary currents can be pretty high (10 amps or more). If the connections (the common ground connection or the connectors) at the igniter become compromised (bad connections with increased resistance) the increased resistance leads to reduced coil currents and less spark energy. This usually shows up first as a misfire at high engine outputs. With the fairly high currents in the ignition system contact resistance can lead to contact heating which leads to contact deterioration and increased resistance leading to ... you get the picture. Having an igniter problem show up on just one cylinder bank is odd and one of those things that I would normally characterize as low probability; but, Valhalla's experience demonstrates that it is not impossible (which now gives me cause to believe that Elvis might still be alive :smile. If you can't borrow an igniter to do a replacement test, I would at the very least remove the igniter and carefully examine the connecting plugs (look for loose or burned pins) and the ground connection.

Have you checked and confirmed that the ignition timing is correct. Simple test and good to rule out as a cause.

In the old days of distributors, it was possible to check for individual cylinder ignition issues by clamping an inductive pick up timing light on each individual plug wire and then observing the operation of the timing light. The inductive pick up depends on the spark current for triggering so anything that materially altered the spark current can result in flakey / intermittent / non operation of the timing light. The use of COPs has made that kind of test more difficult; but, not impossible. You can purchase adapters which allow you to pull the coil out of the well, use the adapter to connect the coil to the plug and then clamp the timing light pick up on the adapter cable. You could use this test to compare the front and back banks and confirm that the spark on the front bank is worse than the rear bank.

The OEM igniter is a bit pricey. RockAuto sells aftermarket units for less. A trip to the junkyard might also be useful. The module was used in some (not a lot) other Acura products. I think the most recent application was in some RXs.
Edit - Not RDX which uses logic level coils, common part with the Legend, some RLs and TLs (probably not going to find a Legend in the junkyard)

 

[ralph cerbone]

Larry B is right , its what we call common point diagnosis . It may or not be the ground on the ignitor but it will be a problem common to that bank ,take a look at the fuel injector and ignition harness . Three coils or three injectors do not fail at once ,not that i've seen in the last 40yrs.

 

[michaelwoodcock]

Thanks for all the tips guys, I am unfamiliar diagnosing NSX's so you guys are a great help!

The preceding does not imply that the igniter or misfire detection module is not the problem. It is just to suggest that the problem did not arise from the use of resistor plugs. I also note that if you were generating misfire codes; but, the engine was running just fine then maybe something was up with the misfire detection circuit. However, you state that the engine is running poorly so the misfires are likely legitimate and not being falsely reported by the module.

I have tested the ignitor, and like you said it seems very unlikely. I switched the imputs & outputs around and the codes stayed the same. Also the engine seems rough to me but I've never been in a new NSX. I assume some idle variation, and vibration in the cabin is not normal for an NSX. My G1 honda insight runs much smoother than the insight does now, with a 3 cylinder 1 liter. But let me again state that I have no experience in an NSX running good as new off the showroom floor.

Have you checked and confirmed that the ignition timing is correct. Simple test and good to rule out as a cause.

Checked earlier as per your suggestion and I need to perform the test again today. using the service loop without using the service connector to hold the idle at the specified RPM I got a reading of 35* BTDC for bank 1 and 25* for bank 2 as per a little dot I placed on the crankshaft pulley when cyl#4 was at TDC. I somewhat doubt these measurements so I am going to re-do the test today using the Service check connector as detailed in the service manual. I'll see if the timing is off.

I also have a coil output voltage checker, I'll measure coil voltage output on the rear bank & the front bank and see if they match up.

I confirmed valve timing as the last shop it was in stated the timing jumped, but they were incorrect and I suspect they just said that because they didn't want to figure out what the actual problem was. Also none of the VTEC pieces are sticking so I don't really know why we got only aprox 180 PSI on cylinder 4 compression instead of 200+ PSI on all the other cylinders.

I used my cheap oscilloscope to check the outputs from the ICM but the sample rate is too low to catch the voltage spikes but they all output nearly identical otherwise. and my test indicates that the ICM is not the problem.

 

[michaelwoodcock]

I s hould also mention that the CKP & CYP have leaked all their potting into the timing belt area. but I don't htink this is cause for a concern, really. let me know what you guys think about that

 

[Larry Bastanza]

I s hould also mention that the CKP & CYP have leaked all their potting into the timing belt area. but I don't htink this is cause for a concern, really. let me know what you guys think about that

Just like every NSX engine........

 

[Old Guy]

Your test of the igniter module is sensible and would suggest that the igniter module electronics are not the problem. If you haven't done this already, one more test related to the igniter. Pull the 8 pin plug from the module (the one with the service check loop). The wires in the connector go to the ignition coils. With the ignition switch in the on position, you should have battery voltage on all 6 wires going to the coils (the other two pins are ground connections). The wires for cylinders 4, 5 and 6 are wht/blu, wht/yel and wht/red according to the '97 service manual. Measure the voltage on those wires and poke or tug at the wiring harness to the coils to see if the voltage measurements on those wires changes or is lower than the voltage for the wires to 1, 2 and 3. If the voltages are good, that suggests that the wiring is probably OK and that your misfire is not being caused by the ignition system.

The crank and cylinder position sensors do seem to have a reputation for leaking potting compound, looking like crap and continuing to work just fine. I seem to recall that I did a little investigation and concluded from the connections that they were reluctance sensors, not hall sensors. Reluctance sensors don't have any active electronics (unlike hall sensors) and tend to be pretty durable which probably explains why the loss of the potting compound does not seem to present an 'electrical' problem.

Checking the ignition timing would be a good thing to do. However, that really only confirms the timing on cylinder #1 . The engine does have separate front and back cylinder and crank position sensors. I don't know whether that means that the timing sequence is calculated separately for the front and back or whether everything is sequenced from cylinder #1 . If the sequence for the front and back is calculated separately, then it might (a really big might) be possible for the front cylinder timing to go off even if the back cylinder timing is correct. Unfortunately, this would be difficult to test. I would also put it in the out-there category as something to check if everything else comes up as a blank.

If you are confident that your valve timing is OK you should also rule out valve clearance as a potential problem. After that, its would be appropriate to look for a lean fuel condition on the 4, 5 and 6.

 

[michaelwoodcock]

Ok, I will try and triple check the results with another timing light, but on cylinder #1 I have to dial my CENTECH (Harbor Freight) timing light to 35* of advance to get it to flash at TDC. If I set it at 0* advance, it flashes way before the timing marks are reached. Assuming this is correct, this should be the source of my problem, I presume. but in that case I should have misfire codes on all cylinders. Strange. I will check to see if I can get another timing light to confirm by tomorrow. I presume there's a keyway and there's no way the pulley was put on wrong, right?

Also,

I found this thread, so I will try another timing light for sure:
http://www.nsxprime.com/forum/showthread.php/31052-Ignition-Timing-Don-at-Hilltop-finds-a-glitch?s=

EDIT:
I made an adapter off the #1 coil, and used that as the pickup for the timing light rather than the service loop, and the timing is dead on for cylinder #1. on to checking the next thing

I also checked coil output: All coils output over 40kv @ idle. no significant difference between front & rear bank.

 

[Old Guy]

Ok, I will try and triple check the results with another timing light, but on cylinder [URL=http://www.nsxprime.com/forum/usertag.php?do=list&action=hash&hash=1]#1 [/URL] I have to dial my CENTECH (Harbor Freight) timing light to 35* of advance to get it to flash at TDC. If I set it at 0* advance, it flashes way before the timing marks are reached.

That is one of the hazards of using the service check loop. The service check loop is just an exposed loop in the circuit for the primary current to the #1 ignition coil. The inductive pick-ups on timing lights are sensitive to the direction of flow of current through the pick up. The light was probably triggering on the start of the coil primary current rather than the shut off of primary current which induces the spark. Switching the direction of the pick-up would probably have fixed the problem; but, your work around also fixed it.

It sort of looks like you have addressed all the potential electrical issues on the ignition side. That's leaving you with a fuel supply / O2 sensor issue or a valve train issue. Normally I would flag fuel pressure; but, seems unlikely since there are no misfires on the back cylinders.

Curious as to how you evaluated the ignition coil output voltage. A 40 kV firing voltage seems exceedingly high for an engine operating at idle. At idle mixture densities and pressures I would have expected the firing voltage on the plugs to be in the 12 kV range, especially with fresh platinum or iridium plugs which tend to fire at a lower voltage than conventional plugs.

 

[michaelwoodcock]

I didn't measure what the spark plugs are getting, but I used an adjustable voltage indicator, the type that has two pointed pieces of metal on the inside and you vary the distance between the two. it has a scale and if the spark jumps distance X then the voltage is voltage Y sort of stuff. Primitive, but I am satisfied with the results. I grounded the meter to the car's cylinder head

 

[michaelwoodcock]

ok I think I am finally grasping what this problem really is with this car.
I unplugged the SPVDM (Spark Plug Voltage Detection Module), Result: no misfire codes, two codes for spark plug voltage detection module, one for each bank
I reconnected the SPVDM and swapped the inputs to it from coils 1 & 4, the misfires stayed on 4,5,&6, and did not move from coil number 4 to coil number one like they would have if cylinder 4 was truly misfiring. (my hypothesis, not proven yet!)
I swapped the outputs from the SPVDM and the misfire codes switched to cylinders 1,2, &3.
I then swapped all of the inputs from 1,2,3 & 4,5,6, so that going into the module cylinders 4,5,&6, if truly misfiring, would once again trigger a misfire code for cylinders 4,5,&6. But they didn't. the codes stayed on cylinders 1,2,&3.

I am going to try and find a SPVDM in the junkyard tomorrow so I know ASAP. If this turns out to be the problem I will probably find a new one to replace that one. If that doesn't turn out to be the problem then I have absoloutly no clue.....

 

[Old Guy]

That is a useful indicator that supports the conclusion that the 'electrical' side of the ignition system is probably not the source of your misfire. I was confused by the reference to '@ idle'.

 

[michaelwoodcock]

I am fairly certain I have figured this issue out, I have a SPVDM (Spark plug voltage detection module) on order.
I detailed some preliminary findings in a previous post but when I posted it it said it was pending moderator approval.
Can a moderator review my pending posts?

 

[ralph cerbone]

I am fairly certain I have figured this issue out, I have a SPVDM (Spark plug voltage detection module) on order.
I detailed some preliminary findings in a previous post but when I posted it it said it was pending moderator approval.
Can a moderator review my pending posts?

check the ground on t he base of the module , mine broke and left a check eng. light ,there were two wires on one terminal

 

[michaelwoodcock]

ok I just replaced the SPVDM and the codes are gone! Thanks for the help guys, now just to go ride to the gas station and get some fresh marine gas & add some fuel stabilizer

Also thanks for the tip Ralph, I cleaned & put some corrosion prevention on that ground, good as new.

 

[Old Guy]

So, now I am curious. In your original post you stated the car idles rough. A failed misfire detection module can definitely cause invalid codes; but, should not affect the operation of the engine. Was the rough idling a mis diagnosis or did you do something else to resolve that problem?

 

[Larry Bastanza]

A failed misfire detection module can definitely cause invalid codes; but, should not affect the operation of the engine.

But what if the failed mis-fire detector caused an intermittent short to the primary side of the spark system?