Link to my diagnostic training manual.
We have had several Honda Fits in the workshop for a misfire or rough gear change. They are 4 cylinder vehicles but have 8 coils. A row in front and a row at the back. This is apparently to control “piston slap” so the ECU decides to fire either the front coil or the rear coil depending on engine conditions. Fortunately the first one that came into the workshop actually had a fault code 04-02 for “noise on crank sensor signal”. The crank sensor sits on the right rear of the engine [viewed from the driver’s seat].
Because of the fault code for noise on the crank sensor signal we scoped the crank sensor and one coil and found that the interference coincided with the coils firing. There was also a large voltage drop when the coil fired.
Sometimes a misfire on one of these engines is simply due to the fact that people don’t know they have two rows of coils so they only replace the rear spark plugs with each service. Eventually the spark gap in the front spark plugs wears so large that it causes a misfire.
To solve the problem with voltage drop first, new wiring was run to the coils via a relay. The wires were all run to the centre of the engine [under the manifold] and connected to a relay with a new fuse directly from the battery. This fixed the issue and we thought that was because the voltage drop was sorted out.
When the next vehicle came in with the same issue we decided to do the job ‘more professionally’ and instead of running wires to the centre of the engine the harness was opened up and the relay was wired directly to the original ignition coil wiring. Although this sorted out the voltage drop problem it did not fix the issue of interference in the crank sensor pattern. So the previous job was ‘accidentally fixed’ because the wiring was moved to the centre, not because the voltage drop was sorted out. On several discussions on the topic the initial fix was “replace all 8 coils”. Then someone said “just replace the 4 exhaust coils” [rear coils], finally it was found that simply replacing the no. 1 exhaust coil fixed the problem. Swapping the coils around didn’t fix the problem, all of the coils had reached the stage where they were causing interference but only the one mounted closest to the crank sensor caused an issue.
On one job we tried removing the crank sensor signal wire from the harness and running it along the other side of the engine bay to the ECU. This didn’t fix the issue but when we also took the sensor positive supply and ran it on the other side of the engine bay it fixed the issue.
So the problem is not due to any particular coil or it’s proximity to the crank sensor, it is based on the length of the positive wire that supplies the coil as it runs next to the crank sensor positive feed. The magnetic field that builds up in the wire when the coil fires, possibly the residual energy [the leftover energy that can’t jump the gap after the coil fires] that kicks back to the battery, induces noise in the crank sensor positive feed which interferes with the pattern. The longer wire of coil 1 induces more interference because of the greater length of wire next to the crank sensor supply and when the crank sensor signal gets cut in half to a low enough point that the ECU recognizes it as two pulses the ECU has a problem doing its calculations.
Update: Since writing this post I had another one of these with this fault. The spark plugs were all good so I simply opened up the wiring harness and separated the crank sensor wires for about half a meter and taped them up a few cm away from the coil wiring which fixed the problem.
Here is a photo of the modified wiring. The red section shows the new route of the wiring. The green line shows where it was originally. Tape the wiring up and leave it as far from the main harness as possible.
