My NSX cranks, but won’t start

[Bregva]

Hey guys, hopefully you can help me on this. My ‘99 Comptech supercharged NSX quit starting few days ago, it cranks, but won’t fire up
Here is what I checked:
1. Verified that spark is present
2. Fuel pressure is fine, I have AEM gauge on the fuel rail and it shows fuel pressure when I crank the engine
3. Removed and checked main relay, found few pins starting to develop cracked solder around them, resoldered everything
4. Checked grounds on the igniter and pretty much everwhere
5. Inspected connectors on the ECU and FIC, they are tight, no broken wires
6. Cheked all fuses, that’s first thing I did
So here I am and don’t know what else to look for
Oh and I’m getting code P0107 when I try to start the car, bad MAF sensor?
Any suggestions other then towing the car to Ramon, ?

 

[Old Guy]

Does the MIL / check engine light come on when you turn the ignition key to the run position? If it comes on briefly then goes out then that indicates that the ECU is powering up which confirms that the main relay is probably OK. If the fuel pump goes through the 2-3 second prime cycle then the main relay and ECU are probably good.

For P0107 I am assuming you meant MAP sensor rather than MAF sensor since the NSX does not use a MAF sensor. A dead MAP sensor would definitely stop the engine from running. However, depending on how Honda configures the firmware, it might fire up while cranking; but, then immediately die once the engine catches and tries to transition from cranking to run where it absolutely needs the signal from the MAP sensor. However, Honda may have put something in the firmware that if the MAP sensor signal is hopelessly out of range, there is no point in even attempting a start. Chasing down the MAP sensor would be useful before hauling the car away. As I recall, the service manual does have the basic procedure for trouble shooting the MAP sensor. The sensor is mounted in that large square 'box full of stuff' on the left side of the engine compartment.

 

[Bregva]

Yes, I meant MAP sensor, sorry. MIL light comes on for a second or two and goes away and I can hear fuel pump priming, so that excludes ECU and main relay. I have the service manual at my work, I'll check procedure when I get there tomorrow. Thanks for a tip Old Guy, I'll keep you guys posted

 

[91 X]

I suspect it's probably the relay....my car was doing the same thing and it was the relay

 

[Bregva]

So I checked vcc it shows 5v on MAP sensor as it should be. I inserted a clip into control pin, put ignition on (I) position, got voltage reading of about 2.8v. Put ignition into (II) start position, cranked the engine and voltage dropped to about 2.6v. On a few youtube videos about how to test a map sensor, voltage was dropping to about 1.6-1.7v, but their engines always would start and idle, in my case since engine wouldn’t start and there is no actually air flow, I’m not sure if voltage reading indicates bad MAP sensor or it is normal

 

[Old Guy]

Since you pulled the DTC, you must have a code reader? Depending on your code reader, you may be able to read the actual manifold pressure on the code reader. With the ignition switched to on; but, the engine not running the manifold pressure should be right around 100 kPa for your sea level location.

When I refer to pressures I always use absolute pressure values rather than gauge values - non of that silly vacuum stuff for me. I think the service manual uses gauge values.

In your test, by control pin if you meant the sensor output pin, then I think your MAP sensor is probably OK. According to the service manual, the sensor output voltage at 100 kPa would be around 2.8 volts. Once you start cranking the engine the manifold pressure (and sensor voltage) will drop slightly because the manifold pressure is slightly reduced during cranking (unless you are holding the throttle wide open). Your supercharger may also affect the pressure drop during cranking. Once the engine starts the MAP drops quickly, probably to around 40 kPa which would give you a sensor voltage somewhere around 1.5 - 1.7 volts. If your engine doesn't start, then the sensor voltage is never going to move much. Short answer is that I think your MAP sensor is probably OK.

If you have a code reader that can read the MAP value and you are getting 100 kPa, then that suggests that the wiring between the MAP sensor and the ECU is probably OK.

If you have a CTSC, you must have one of those FIC thingies? I have no knowledge of the FIC. Does it intercept the MAP signal and modify it (since the OEM ECU fuel map is not scaled for boost)? If so, is it possible that the FIC is screwing up the MAP reading? If your code reader can report the MAP value and its not 100 kPa, that might indicate a screwed up FIC.

You have confirmed that you have spark. Did you confirm that it is timed correctly by using a timing light to check timing position? This would be a really unusual failure.

If you have access to injector test lights, check to see if the injectors are firing during cranking. If they are not firing, check the injector outputs right at the ECU before the signals are modified by the FIC. If you are getting signals out of the ECU; but, nothing out of the FIC that might confirm that you have a FIC problem.

Finally, the MIL light coming on and the pump priming confirms that the ECU is powering up and booting up. However, that does not guarantee that the ECU is completely OK. The P0107 code can be triggered by an ECU internal connection problem so don't rule out an ECU problem. However, in the world of failed stuff, my experience is that OEM ECUs are pretty low on the cause list unless somebody has fiddled with them.

 

[Bregva]

I use a very simple OBD II reader, I don't believe it is capable of reading manifold pressure, I also do not know if spark is timed correctly since I don't have a timing light and neither do have injector test lights. I can only guess at this point that injectors are not opening at all, or are opening, but not in correct sequence which points to possibly faulty FIC. Service manual mentions pulses, but won't specify of what type and amplitude. Regardless, I will try to see if I can detect any kind of pulses on the output of ECM and output of FIC, as you suggest

 

[jwmelvin]

You should be logging from the FIC during a crank cycle. Do that and post it please.

 

[Old Guy]

If the FIC supports logging, do as jwmelvin suggests before you go poking around the wiring harness to measure things. Personally, disturbing the wiring harness to measure things is my last resort because of the risk of damaging connector pins. Also, after 19 years the plastic on the retainer clips is starting to lose its flex which means stuff may snap rather than un-clip.

 

[Bregva]

You should be logging from the FIC during a crank cycle. Do that and post it please.

Jason, I e-mailed you couple logs, let us know

 

[Gary W]

Do you see any RPM while trying to start the car?

Gary

 

[jwmelvin]

Jason, I e-mailed you couple logs, let us know

I didn't receive anything yet.

 

[jwmelvin]

It occurs to me that this could be an issue with vacuum line #4 , which runs to the MAP sensor. (During the install, one must switch lines #3 and #4 on the module side of the hard lines, causing the MAP sensor to be fed from the post-blower area.) That line gets clogged in CTSC setups because in the default configuration, it comes from the bottom of the intake plenum under the blower. I had an issue with this that took me a while to find. My car would start but sputter and die. It's worth checking I think. You can pull the line off and blow carb cleaner into it, which for me temporarily resolved the issue. I eventually took that vacuum tap from the side of the intake plenum, where the FPR tap comes from. Given that you are seeing an issue with the MAP sensor, I would check all those lines, and also double check the MAP sensor connection. You need to look at OBD2/FIC data to see what the ECU is reporting for MAP sensor voltage with the engine stopped and with it turning over.

 

[Bregva]

Do you see any RPM while trying to start the car?

Gary

Yes, RPMs are about 160 as logged on my Zeitronix

 

[Bregva]

I didn't receive anything yet.

I'll resend it tonight

 

[Bregva]

If the FIC supports logging, do as jwmelvin suggests before you go poking around the wiring harness to measure things. Personally, disturbing the wiring harness to measure things is my last resort because of the risk of damaging connector pins. Also, after 19 years the plastic on the retainer clips is starting to lose its flex which means stuff may snap rather than un-clip.

I still poked around FIC connectors a bit and saw 12v power going to it. Haven't checked further

 

[Bregva]

It occurs to me that this could be an issue with vacuum line #4 , which runs to the MAP sensor. (During the install, one must switch lines #3 and #4 on the module side of the hard lines, causing the MAP sensor to be fed from the post-blower area.) That line gets clogged in CTSC setups because in the default configuration, it comes from the bottom of the intake plenum under the blower. I had an issue with this that took me a while to find. My car would start but sputter and die. It's worth checking I think. You can pull the line off and blow carb cleaner into it, which for me temporarily resolved the issue. I eventually took that vacuum tap from the side of the intake plenum, where the FPR tap comes from. Given that you are seeing an issue with the MAP sensor, I would check all those lines, and also double check the MAP sensor connection. You need to look at OBD2/FIC data to see what the ECU is reporting for MAP sensor voltage with the engine stopped and with it turning over.

I removed MAP sensor and sprayed into the vacuum hose. That did not change anything

 

[jwmelvin]

I removed MAP sensor and sprayed into the vacuum hose. That did not change anything

Bummer. Thanks for trying.

 

[jwmelvin]

The log of you cranking can be seen here. I'm not sure how helpful it is but it seems to show that there is no response of the Zeitronix internal MAP sensor while cranking (this is a log from the Zeitronix system). It is possible to log from the FIC itself, which gives you information about what the factory sensor is seeing. That is how I eventually diagnosed my problem, which seems not to be the same as yours.

Have you checked all the vacuum connections? If you have a hand-driven vacuum pump (like this or this), you can pull a vacuum on the MAP sensor (without the engine running, while monitoring via OBD2 or the FIC) to make sure it is responding as you expect. The flaw with this line of inquiry is that, in my experience, the engine will run but sputter if the MAP sensor is reading atmospheric (WOT no boost). But we really need to see the voltage coming out of your sensor to know more, easily available from OBD2/FIC.

 

[Old Guy]

The factory service manual does not list the P0107 code as a cause that would prevent the engine from starting. The rather limited number of EFI systems that I am familiar with do not use the MAP sensor / fuel map on start up (primarily because the MAP signal can be bouncing around a lot during cranking). They typically use a coolant temperature based fixed fuel pulse width during cranking to get the engine to fire and then when the engine fires and the RPM transitions above some limit (probably 400 - 700 RPM) they switch to the fuel equation using the fuel map, RPM and MAP inputs. On those systems, a faulty MAP sensor won't stop the engine from starting; but, as soon as the engine transitions above the crank / run threshold it promptly dies or runs like absolute crap depending on which way the MAP sensor fails. The OP is not reporting any cylinder firing at all which makes me think that it is something more than just the MAP sensor, hence my suggestion about using an injector test light to see if the ECU / FIC is even firing the injectors. The earlier voltage measurements that the OP reported also suggest that the actual factory MAP sensor is likely OK.

Injector test lights are pretty cheap and you really only need one, not six for this test; however, if the OP can data log on the FIC during cranking and can record the MAP, RPM and fuel PW values, that would confirm that the FIC is calculating the fuel pulse width during cranking. However, it is possible that even though the FIC is alive, the injector output stage of the FIC has failed which makes the injector test light the ultimate end device test.

As a precaution, if it turns out that the injectors have been firing during all this testing and that the absence of engine starting has been caused by something else, depending on how much testing / cranking the OP has done, they may want to consider changing the oil and perhaps pulling the plugs and applying a little oil into the cylinders. Depending on how much cranking has taken place, the cylinder walls could be pretty devoid of oil if they have been well washed with gasoline. Perhaps pull the rubber intake boot off of the throttle body and give a sniff test (this won't be as effective with the CTSC) - if it reeks of raw gasoline you know the injectors are probably firing and you should probably be doing something about dealing with that residual gas. He could also pull a spark plug to check to see if it is wet with fuel. Not to be alarmist, but, if there is a lot of fuel vapour hanging around in the intake plenum, if you had some weird ignition timing condition you could have a spectacularly bad back-fire event. True back fires on modern cars are super rare and on an NA car they just scare the crap out of everybody with flame coming out the intake. On an FA car with a fixed displacement blower and no BOV device, there is no place to vent the explosion potentially creating an intake pressure cooker bomb with disastrous results. My current bet is that this is alarmist because my opinion is that your injectors have not been firing and you don't have an engine full of raw gasoline; however, do the safety check.

 

[Bregva]

I went ahead and bought a noid injector test lights. Checked one of the injectors and sure enough there was no pulse. So at this pount I’m pretty confident that my FIC is dead

 

[jwmelvin]

Thanks for the report. That’s a bit frustrating about the FIC.

 

[Old Guy]

I went ahead and bought a noid injector test lights. Checked one of the injectors and sure enough there was no pulse. So at this pount I’m pretty confident that my FIC is dead

Did you ever attempt to connect a lap top up to the FIC to see if you can generate a log of a start attempt? If you attempted to initiate a log and the lap top could not establish communications with the FIC, that would reinforce the conclusion that the FIC is dead.

Someone with direct experience with a FIC on a CTSC car can perhaps advise on this; but, it occurs to me that since the FIC is a piggyback system, you should be able to disconnect the FIC and restore all the original connections to the factory ECU and sensors. This should allow you to try a start on the engine because the engine should start and idle fine on the factory ECU settings. I would not drive it or operate the engine in an RPM range where the SC is delivering material boost. If the engine starts with the FIC disconnected, that is pretty definitive that it is dead. However, if you don't have the original fuel and ignition maps stored some place and the tuning software to load the maps, then you can't replaced the FIC yourself so perhaps its time to haul the car off to your favorite repair guy.

 

[Bregva]

I connected laptop to FIC and it did not recognize it, then I removed the FIC and reconnected ECU harnesses and the car started right up! So that tells us the unit is bad for sure.
I want to thank you guys for guiding me through all this and for invaluable info I gained in the process. Hopefully this could help others too if they encounter similar problems. What a great group NSX Prime is! And generally NSX owners, and I met quite a few
Now I’ll have to either look for another FIC, luckily I have my tune saved and could reinstall it into another unit. Or go standalone. Please let me know if anyone is having anything for sale

David
1999 CTSC NSX

 

[tdiggity]

If you’re in Los Angeles, legally your only option is another FIC (smog).


Sent from my iPhone using Tapatalk