Under pressure

2 June 2014  

By James Dillon

It never ceases to amaze me how a similar symptom can have many different root causes. Perhaps this is why the specialism that is ‘vehicle diagnostics’ piques the interest of many vehicle technicians. I often hear the words satisfaction and challenge mentioned when discussing why people chose to tackle problem vehicles. Undoubtedly, satisfaction and challenge are the flip side of frustrating and time consuming. The daily diagnostic challenge is one I wouldn’t choose to be without.

The supplying garage had been very thorough in their tests and checks, going over the usual suspects of compression, ignition and fuel. The compression test they’d done showed nothing out of the ordinary. All of the cylinders were within 10% of each other, the ignition system had been checked and suspect and worn components had been replaced accordingly. However, even with the ignition system in tip-top form, the misfire prevailed and continued to set a diagnostic trouble code for a cylinder two misfire.

With the mechanical function tested and verified, and the ignition system as good as new, logic dictated that the fuel system warranted further investigation. The injector function was tested and the injector on the suspect cylinder was swapped from the misfiring cylinder to a neighbouring cylinder but still the vehicle misbehaved. Having spent their allocated diagnostic and investigation time, the supplying garage dropped it off for me to take a look.

It was a bit puzzling that the fundamental tests and checks hadn’t uncovered the root cause. The car had a fault, so there definitely had to be a cause. Normally in cases such as this, the likelihood was that the right thing wasn’t being measured at the right time or in the right way. I decided to start at the beginning and run through compression, spark and fuel to see if there were perhaps any subtle clues which may indicate what was going on. I did a relative compression test using the Pico and the current clamp. Upon current clamping the battery cables to the starter, a 10-second cranking test showed the waveform, Figure 1.

It is clear from the cranking amps waveform that there isn’t a major issue with compression. However, when the data was inspected more closely, the compression was seen to be slightly lower on each corresponding peak (but well within the 10% ‘rule of thumb’). An interesting and possible ‘clue’ but un-actionable at this stage of the diagnosis: Fundamentally, the cranking compressions were within spec. Next was to verify the spark event for each cylinder using a voltage and current analysis of the primary circuit. This is a quick and accurate test which encompasses the key elements of the ignition system. The waveforms showed no signs of abnormality (apart from the anticipated impact when the ECU shut down the injector on cylinder two). So the ignition system wasn’t the root cause of the issue.

In my opinion, the best method for verifying the quality of the constituent elements of combustion is the running ‘in cylinder’ compression test, using a pressure transducer. I am a raving fan of this method as it allows the technician to see absolute pressure values (at running speed if possible), valve timing events (relative), intake pressure and exhaust pressure; a fantastic, efficient test method. One data gathering event and multiple data points, highly efficient. I pose the question, is there another non-intrusive, rapid method of obtaining this type of data?

I started by measuring the ‘suspect’ cylinder number two. This involved screwing the pressure transducer, via an adapter pipe, into the spark plug hole of the cylinder under test. The data capture I took was a little uninspiring, as at first glance, it looked pretty much like it ought to have. All of the constituent elements were present. I decided to compare this cylinder with the other two and used the Pico reference waveform feature to perform a comparison, see Figure 2. Firstly, there is a large difference between the running compression values; 5.7 bar on cylinder two, 7 bar on cylinder three and 9.3 bar on cylinder one.

Although the time taken to complete the four-stroke cycle differs due to engine speed between the different cylinders, the critical exhaust and intake valve timing slopes show a variation in cam timing between them. How can this be possible with cast camshafts? There are also additional clues in the vacuum element of the waveforms, all was not well with the ‘breathing’ of the cylinders.

At this point, the data was pointing towards valve train trouble. The exact nature of the fault is not obvious, but there was enough data to warrant a visual inspection. After the rocker cover was removed, the root cause of the problem became obvious, see Figure 3. As I rotated the engine to observe the valve train mechanism, it became apparent that one of the cam cap bolts had snapped and the cap for cylinder three was loose. Several of the other cam caps were also loose. The cams were lifting during their rotation causing the compression and valve timing differences.

Find out more