Diagnostics
This article is a true description of an AECS technical help desk problem and how it was solved. By Herbert Leijen, trainer/research for AECS Ltd
Vehicle
2002 Nissan Sylphy (Bluebird) 2.0-litre direct injected QR20 (DD) engine.
Problem presented to the help desk
This car has been hesitating on acceleration, pinks under load, and the engine check light comes on intermittently. We checked for fault codes; the fault codes logged so far were: P0120-Throttle position sensor circuit. P0605-ECM. P1320-Ignition signal-Primary. P0400-EGR system. Because the codes, except P1320 (a non-OBD code), did not really relate to the drivability problems the car had, the diagnostician decided to measure the injector signals at the ECU on all four injectors.
Three-wire injectors
The injectors used in these Nissan engines are of the three-wire type. As discussed in the EMS 1-3 training, the injectors have two coils, one with a very low resistance to open the injector very quickly, and one “normal” coil to hold the injector open.
A lot goes wrong with those direct injection injectors, any garage can vouch for that!
During the EMS 1-3 I show pictures of an ultra-sonically cleaned and flushed injector, made with a microscope, of the swirl channels inside the injector. The channels still fill with carbon, formed as a result of the high heat and pressure the fuel is exposed to inside the injector.
This first measurement the diagnostician made revealed that the number 1 injector scope pattern looked very different to the rest of the injectors.
Measuring
The dual channel measurement made with the ATS scope showed immediately that there was a problem with how the ECU was “driving” the injectors, specifically number one.
The recording of injectors 3 and 4 looked the same as the recording on injector 2.
ECU faulty?
This left us at the AECS help desk puzzled. We had never seen that before. Both coils of one injector need to be activated at the same time, not with a 180 degree offset as recorded at injector 1. This looked to us like as if the ECU was “confused”. If you combine that with the fault code P0605 (ECM), anyone would start to lean towards a “magic fault” (=ECU fault)
Think
Just a little bit of thinking tells you that the ECU IS capable of switching both coils of injector 1, would you not agree? That tells me straight away that there cannot be too much wrong with the ECU. It is still possibly faulty, do not get me wrong, but before we get another ECU we certainly need to check another few items!
What the recording did tell us is that injector 1 was injecting twice per two revolutions, instead of once.
This would certainly make the engine run incorrectly, to say the least: First, when the injection should take place during the compression stroke (stratified injection mode), not enough fuel will be injected, as only one coil activated the injector. Second, injection will also take place during the exhaust stroke. Since there is no pressure to speak of inside the combustion chamber during the exhaust stroke, the piston rings will not seal. A lot of that injected fuel will end up in the sump making the mixture richer for all cylinders (through the PCV). In addition, a lot of that fuel will go down the EGR channel, back into the intake manifold.
Later during this job, it was found that a lot of fuel was present in the sump, because of a leaking fuel pump seal.
So is the ECU faulty?
An ECU cannot command the injection or ignition properly when its input signals are incorrect.
The main input signals for injection are crankshaft/ca shaft activity. If the engine is not turning, the ECU does not know that injection has to take place, it is as simple as that.
If there is something wrong with injection or ignition phasing, it is usually a result of ECU calculation trouble, where it does not see the rising or falling slope of one of the sensor signals when expected.
On top of that, many Nissan chaindriven engines suffer from silent chain stretch. Again, any garage can vouch for that.
A logical next step is to record cam vs crank signal.
At first glance, this signal is wrong! The rising slope of the single pulse of the cam should no coincide with the last teeth before the ref. point of the crankshaft.
I have seen so many before. However, this seems to always give ignition faults, but I have never seen injection faults like this. To make sure I even looked up a pattern recorded on another Nissan, not a QR20 (DD) engine.
I said without too much further thought that he had to replace his timing chain and guides, and thought that this would be the end of it.
Sting The sting is as always in the detail! This is why we use high sample rate scopes.
The remark of the diagnostician after the cam chain replacement job was; “it runs sweet but the cam crank pattern
Advance
Of course, when I saw the pattern, we disrespectfully asked the diagnostician “are you sure the marks line up?” However, even before the question was expressed we realised that the cam had to advance with new chain and guides, something it obviously did.
Conclusion
The car is fixed, it runs perfectly now, that is for sure. Yes, the problem was the timing chain and not the ECU. However, this case has taught us here at the AECS help desk that we have to be more careful with comparing known good patterns with problem patterns. No complacency allowed!
The compare signals HAVE to be of the same engine. As I said, the sting is as always in the detail - detail which can be achieved with good quality equipment like the ATS scope.
is very different from what I expected”