Code 16, Replace ECU?

[poopynsx]

Okay not sure where to ask this however I'll do it here. I recently got a CEL code 16 which states Fuel Injectors. I used the manual all the way down to the final few steps of grabbing a ECU Test harness to measure the injector terminals to the ground of the ECU. Considering I don't have one, nor would like to fork out $800+(according to a search on the forums) i'd rather just replace the ECU if it's that which is cheaper anyway. Do the cheaper stuff first normally . I asked a friend who does a lot of electrical work with cars and he said to check the resistance from the injector pin to the ground pin on the connector that goes to the ECU. I checked it and found no resistance which he said is good, then which he thinks could be the ECU. He also said if it were actually the ECU I should have a bunch of fuel just pumping out of the car, which the car does produce a big puddle behind the exhaust after its turned on. I also connected a noid light to the injector that isn't clicking and with the car on the light stays on.

Before I commit to anything is it common for the stock ECU to go out on a 1991 NSX? The modifications the car came with are Headers, Exhaust if it matters. Also if that is the case, is there a difference between the 37820-PR7-A03, 37820-PR7-A06, etc. part numbers that come up for a 1991-1994 ECU?

Thanks again!

 

[Old Guy]

As a matter of clarification, when you checked and found 'no resistance', does that mean a resistance of zero (a short to ground) or infinite resistance (open circuit)?

Pull all six injector electrical connectors, then with the ignition switch in the run position; but, the engine not running, connect your test light to each connector. If the light doesn't light up on the other five injectors but lights up on the one injector that isn't clicking, you definitely have a problem. The injectors are switched on by the ECU firing a transistor inside the ECU which connects the injector to ground allowing current to flow through the injector (see the attached schematic). If you have one light that is staying lit all the time with the engine not running, it means either the power transistor in the ECU has permanently shorted to ground or there is a short to ground on the wire that connects that one injector to the ECU.

My service manual is in the trunk of my car which is locked up in secure storage so I can't get at it. If you have the service manual for your car and you can figure out which plug on the ECU is for the injector wiring harness, pull that plug out of the ECU and then repeat the test with the test light. If the test light no longer goes on, it means that the ground is likely caused by a faulty injector driver transistor in the ECU. Replacement of the ECU is the obvious solution; but, not the only solution. If you check around, there are companies that do ECU repair which will be cheaper than a new ECU. If you know somebody who is savvy with electronics and a soldering iron, replacement of the power transistor should not be that difficult (providing it is just the power transistor).

If you pull the injector wiring harness plug out of the ECU and the test light stays lit, then you have a wiring harness fault and you need to carefully check the harness for chaffing or other damage. In the world of probabilities, unless somebody has been messing with the wiring harness or made modifications which caused rubbing on the wiring harness, I put a pretty low probability on a wiring harness fault.

 

[poopynsx]

Thanks for the help sir! I guess it fixed itself... I think it was a bad connection from the plug that goes to the harness to ECU. I started checking to make sure all the injector connectors were all off and #4 stayed on... and I noticed it didn't but then I realized I forgot to plug the connector back in again. So I connected the plug and the #4 did NOT stay lit like last time with the noid. did all of the injectors again and all off with IGN set to RUN with engine off. I started the engine with the noid on #4 and it was pulsing like normal... I then reconnected the injector, reset the CEL and runs normal again... Gremlins. I let it run for a bit because it has a bunch of white smoke, from my understanding thats ring wash right? it went away after letting it idle for a bit. Hope its fixed... I'll check for shorts again anyway. Thanks again for your help Old Guy!

 

[Old Guy]

The injector staying open all the time could have dumped a lot of gas into the cylinder (if the intake valve(s) were open) during the start-up phase. That gas would be mostly un burnt and could contribute to washing oil off the cylinder walls. It is not clear to me why washing oil off the cylinder walls would cause white exhaust smoke. Heavy white exhaust smoke is usually a sign that you have a coolant leak into the cylinder. Light white that goes away is usually just condensation in the exhaust gas that go away as the exhaust system heats up.

If that injector was sticking in the 'on' condition and dumping raw gas into the cylinder, I suggest that you do an oil change as your engine oil may be diluted (contaminated) with gasoline. If you were running the engine with the stuck open injector, that particular cylinder will have been running very rich. I suggest pulling the spark plug on that cylinder and having a look at it. It might require replacement.

Its nice that the ECU is not dead. Good luck with the wiring. Finding a here again gone again wiring problem is a real pain. Have a close look at the plug for the injector wiring harness that goes into the ECU. If a previous owner had been 'playing with things' they might have bent or damaged the pins on the connector which could have caused the ground problem. When checking for wiring problems I start with examination of the terminal plugs first and then any point where the harness passes through a bulk head or under a clamp.

 

[GraemeD]

Just to add a data point, I did have a bad female socket in the ECU connector. You could probe it and it was fine, but it would not make contact with the male pin in the ECU. The way I found it was to remove the ECU cover and probe the suspect connection from inside the ECU. So some times if just reseating the connector solves the problem, it might be temporary.

You up can also test for this condition, by using a spare pin and insertining it into each female socket and feeling the tension/grip. If there is very little resistance, the contact might be damaged. This can happen from people using too large of a probe/wire to test and it damages the contact.