• Protip: Profile posts are public! Use Conversations to message other members privately. Everyone can see the content of a profile post.

Cranks but won't start

Joined
13 February 2019
Messages
91
Location
Roseville, CA
Just did a timing belt, water pump, LMA, gaskets and clutch and after putting it back together it cranks but won't fire. It also sat for 12 years and had some rats visit. Only chewed one wire that I could find and everything looked really good when apart.

Now it cranks but doesn't fire. It's getting fuel to the rails and spark plugs are wet with fuel. I haven't done a compression check but it sounds "normal" when cranking. I'm thinking ignition system. Sensor, relay, etc.

All fuses are good.
Have not checked any relays.

Suggestions for next step?

Dan
 
If the starter turns the engine over, then the culprit is something in the PGM-FI loop. Main relay is usually the #1 culprit, but the symptom there is typically that the car starts but then immediately dies. It looks like you are getting fuel so I'd check the ignition system. The ignitor unit does fail on the NSX and it would mean no spark. Seems silly but check to make sure the ignitor unit is plugged in. Same for the crank position sensor.
 
Crank position sensor is the plug next to the alternator correct? That's connected and ohms checked out at around 650 I think. Where's the ignitor? I've checked all the plugs ten times. Oi.
 
The ignitors just die. They short out and zap its over. It probably happened when you first put voltage to it after the TB service.
 
If i understand the diagram correctly, doesn't the igniter need a signal from the ECU using the crank position sensor and MAP sensor inputs? How would I test that signal? Seems like without that signal, testing the igniter is only half of it.
 
Correct. Crank / cam position sensor provides tach and timing signals to the ECU which then triggers the ignitor unit to fire the ignition coils. I don't know the specifics of the NSX ECU; but, on most MAP based fuel injection systems the ignition timing is fixed value (may be temperature based) during cranking and the ignition does not transition to the ignition map until the engine transitions from starting to running, probably around 300 ish RPM. Short answer is that the MAP sensor probably does not enter into your non start problem.

You did not specify the year of your car which is important since fuse functions and wiring details did change during production. I am going to assume that the 1991 service manual applies.

In post #3 you said you checked the resistance values on the crank / cam position sensor and they were OK. There are four sensor coils. Did you check all four coils? Also, check for a shorts between the individual coils. The resistance between terminal pair A B and terminal pair C D (and all other pairs) should be infinite. If the coil resistances check out OK and there are no shorts between the coils then the sensor is probably OK. They are dumb-ass simple devices and very reliable despite looking like they are ready for the junk heap because of oozing potting compound. Rodent chews on the pigtail are probably a higher failure rate cause than failure of the actual sensor.

If you have injector noid lights you can do a confirmation that the CPS is functioning and that the ECU is reading the sensor and generating injector pulses. During the cranking phase the ECU will probably generate fuel injector pulses using a temperature based look up table. If you crank the engine and you see the injector noid lights flashing you are getting both the position and speed signals from the sensor. The fact that the spark plugs are wet suggests this is probably happening; but, when you power up the ECU it probably generates a fuel pulse to wet the intake manifold to speed start up. This happens without an RPM signal so if you powered the ECU on and off enough times this could also cause wet spark plugs.

If you have a test tachometer you should also be able to connect to the tach test point (page 23-92 in 1991 SM) and during cranking observe the engine RPM. This would confirm that at least the crank sensors are functioning.

In post #7 you said test shows a faulty ignitor. How did you test the ignitor since the service manual does not set out a test for the actual ignitor. On page 23-93 (1991 service manual) it sets out an ignitor input test. If it failed that test the problem is not the ignitor that has failed; but, likely a wiring problem to the ignitor. Anecdotally, the most common ignitor problem was a ground failure of the external ignitor grounding, specifically the G403 ground. Make sure the ignitor grounding is good. Have you checked fuse #13 which is the main 12v supply for the ignition coils and confirmed that the coils are actually getting 12v power?

The ignitor is actually a relatively simple device. It is six power transistors which may or may not have some additional voltage protection devices. One transistor failure should still leave you with 5 functioning transistors which should allow an engine start. As noted, the most common failure point is the external grounding which is easily fixable. If you have some electronic test equipment (a pulse generator and an oscilloscope) it is possible to check the operation of the six individual ignitor circuits in the ignitor module. Most people do not have that kind of equipment hanging around.

Just because it is easy to do, as a first step clean the ground connection for the ignitor module and then check for spark during cranking. If that doesn't fix the problem, as an observation, rodents don't usually stop with one wire if they are looking for nesting material. I would check that engine wiring harness for additional damage, particularly in the places that you can't see
 
Wow, awesome info. The 30a #13 fuse is good and it was the power test that I did. Both ground pins showed continuity and all six coil pins showed +12v. That's all I did. I just did the ohms test on the crank sensor and the numbers match what the SM said is good. Yes, I have a 1991 MY.

After spending some time on it last night I think I'm down to double checking the wiring one last time like you said. The damaged area on the harness is between the ECU and igniter on top of the engine but I had it on my bench when I repaired it and inspected it really well, I think. I'll look again, playing the odds...
 
If everything checked out on the ignitor input test described in the SM and you can confirm that the injectors are firing (noid lights) things are not looking so good for the ignitor. I would remove and clean the ignitor ground lug connection just to rule it out because multimeter tests are poor a confirming the condition of a ground connection. I would also make sure that the bolted connections for the ignitor module are clean and good. It is not clear whether the module also uses the mounting bolts as part of the ground path.

The ignitor input test does not check inputs signals from the ECU (the pin plug) because most people (including dealerships) don't have the required tools to do that. Those six pins are the six trigger outputs from the ECU for the coil ignitor transistors. Rodent damage on that part of the harness could kill the ignition; but, the rodent would have had to chew through all the wires to completely kill the ignition system. If they only chewed 2-3 wires the engine should still be able to start. Rather than just visual inspection, I would be inclined to do an end to end continuity test between the plug on the ECU body and the ignitor 6 pin plug.

If the wiring between the ECU and the ignitor checks out as good, replacement of the ignitor might be the next step. The NSX shares the same module with the 91-95 Legend, 96-04 RL and the 96-98 TL and there may be others. Short answer you may be able to scrounge a module cheap at a pick and pull. Ebay sellers list them for for around $50-70 used. Aliexpress lists them new for around $70.

The NSX ignitor is like a fly trapped in amber. The $400 - $500 OEM pricing is a left over from a previous century as its design is based upon technology available in the late '80s. These days, you can purchase dedicated ignition drivers on a chip from ONsemi for about $3 - $5 each so a homemade equivalent can probably be fabricated for less than $50 (might not look OEM). The Chinese reproductions on Alieexpress might be just fine.
 
Last edited:
The NSX ignitor is like a fly trapped in amber. The $400 - $500 OEM pricing is a left over from a previous century as its design is based upon technology available in the late '80s. These days, you can purchase dedicated ignition drivers on a chip from ONsemi for about $3 - $5 each so a homemade equivalent can probably be fabricated for less than $50 (might not look OEM). The Chinese reproductions on Alieexpress might be just fine.
Your advice is excellent- I especially like this part. "Fly trapped in amber"...so perfect. :)
 
There's a 99 RL at the Pick n Pull down the street from me but guys go through the yard and pull anything of value for ebay. Worth a lunchtime visit though.
 
Well, I picked up what I thought was the ignition module but it has a silver cover on it with no part number, probably aftermarket. Online those are misfire detection sensors. WTH.

Edit: I went back and found the ignition module. Wasn't where any of the photos suggested but I found it. I'll test tonight.
 
Last edited:
ok , I have a 97 which uses a misfire detection system , dont know if 91 uses it . detection sensor only has one connector , ignition module has two multiple wire connectors at each end
if you had rodents in car you have to check every compartment . had a problem with a friends corvette , two weeks of work to disassemble engine compartment , drivers compartment and rear compartment . a big PIA . Yes i did fix car ,found many wires that were chewed on which would result in future problems
 
ok , I have a 97 which uses a misfire detection system , dont know if 91 uses it . detection sensor only has one connector , ignition module has two multiple wire connectors at each end
91-94 do not use a misfire detection circuit. Those cars have no way to detect a coil misfire- they can only detect a problem on the ignition output signal from the ECU to the ignitor (DTC 15)
 
91-94 do not use a misfire detection circuit. Those cars have no way to detect a coil misfire- they can only detect a problem on the ignition output signal from the ECU to the ignitor (DTC 15)
Thanks Honcho that is what i thought but was not sure . the misfire detection system on NSX is different than what most manurfactures used . Most manufactures used crank sensor discrepancy NSX measures secondary ignition , thats why three wires to ignition coil .
 
Im spitballing here,
Did the engine run before the TB service?
Is the gasoline fresh?
Are you able to verify TB alignment?
Are the spark plugs new?
I’ve never tried with a coil pack, but it used to be a simple check to pull the spark plug wire off and while cranking, hold the boot close to the block to see spark. Not elegant, but effective. I would think it is possible to do the same with a coil pack, although a little more involved due to the harness and mounting.

If I remember correctly the factory service manual has a thorough flowchart for troubleshooting. Go down it step by step, even if you’ve already done it.
 
Back
Top