Complete spark and fuel cut for no apparent reason

[empty_inside]

Hi guys,

I was driving my S15 Spec R gently down the road today and suddenly lost all power, and now it won't start.

I've checked all the usual stuff and found nothing untoward but I can explain why it won't run - there's no spark* and no fuel injection**.

* There's 12V getting to the coils so they charge, but no actual sparks, so presumably the spark signal isn't coming down the third wire (I don't have a scope so I can't check).

** The pump's fine, you can hear it prime and if you take the pipe of the end of the fuel rail and crank it, petrol *pours* out, so the delivery's fine. However after loads of cranking the cylinders are bone dry. I connected to the ECU and when I crank the fuel injection timing moves, but the injection duty cycle remains at 0.

Given that everything else functions, the lack of injection and spark signal seems to me like it can only be the ECU - the ECU even tells me it's not putting any fuel in. Any idea why? There are no ECU fault codes.

That was my first drive after fitting a new Walbro fuel pump that afternoon, but I'm at a loss to see how that's caused it. My only thought was: is there an impact sensor (crash sensor) in or near the tank I might have knocked? I assume not because the poor thing was hit by a 2tonne Mitsu GTO and that didn't trigger it. Plus I lost power half way down the road driving slowly in a straight line.

I assume it's not the barrel / alarm / etc because I can cycle through the barrel states, crank for ages, lock and unlock the car etc.
I can't imagine there's any kind of NATS issue (JDM cars don't have it, right?) or security coding of the pump because it only takes 2 wires: + and -.

I'm really at a loss, anyone got any ideas?

 

[empty_inside]

After a whole day, I finally solved it! Probably worth a quite write-up in case anyone else sees these symptoms.

tl;dr: nothing to do with the fuel pump, massive co-incidence. It was the ignition coil earth, completely stops the car.

So after putting the OEM fuel pump back (in the name of science) and getting no change, I idly checked the ECU again. This time it was showing Code 21* - the helpfully named "ignition signal primary circuit". I checked the loom, tried it with each coil off in turn, and changed the CAS. Nothing. I still wasn't sure that Code 21 was anything to do with the problem, because I've had coils go before and Code 21's come up and the car continued to start and run, albeit on 3 cylinders.

I then followed the diagnostic procedure starting page EC-114 of the manual. It's not rocket science, just has you follow the 12V and ground round the loom, from the ECU pins to the ECU relay (it has a 2nd pole just for the coil power), and so on. I got to stage 14/18 and sure enough no ground on the coils' primary circuits. Their earth doesn't go back through the loom to the ECU, instead it comes out of the loom not far from the coils and it bolted to the back of the rocker cover. This wire had frayed and broken away from its ring connector.**

My mistakes were:
- not finding a good earth to test the stray wire.
- Missing Code 21 (maybe).
- Verifying 12V /to/ the primary coil winding and assuming that the lack of spark was a missing signal (ECU issue) - which I couldn't check without a 'scope. If only I'd checked the earth...

What gets me is that the ECU seems to know the difference between "different" Code 21s. Some types cause a fuel cut too. I suppose this makes sense so you don't coast down the road with no spark and fuel pissing into the cylinders. However, it doesn't make it easy to diagnose. More than once I thought "am I sure JDM cars don't have NATS?"

I suggest everyone checks this earth wire as if it goes you get a sudden total loss of drive. Maybe the top of a hot, vibrating engine isn't the best place to secure a vital wire, Nissan.

Hopefully we've all learnt something today!

* I *swear* that wasn't there yesterday
** I had noticed it was off as soon as I opened the bonnet yesterday, but I'd tried cranking with the held firmly on an earth point to no avail. It must need a super-good (e.g. soldered connection) not to be detected as faulty.

 

[rudd-o]

This is exactly what happened with my s15 and gone away when i pushed the loom. But The problem apeared again today after the last rainy day :/ . I ll start checking the loom to see if any wire is fried or damaged. Another issue is that my boost gauge led fried (i had it replaced 2 years ago) and roof light also blown.. dont know probably i have to check all the cars wiring