Intermittent fault on a Polo
25 November 2013 
By James Dillon
The vehicle was referred to us for an inspection and assessment. The first question in my mind was: Are these symptoms linked or did they have separate causes? And, was the root cause or causes going to be financially reasonable to sort? A logical and steady approach was called for. The first thing to aim for was to get the vehicle to misbehave whilst the test equipment was attached so that I could see and feel what was going on.
The next morning I tried to start the vehicle, to no avail. It cranked well, and at a good speed, but simply failed to fire up. I noticed that the ECU had a ‘been tested’ sticker on it, which was confirmed by the customer. It had been sent away for testing and had come back no fault found. In addition to this, the crankshaft sensor had also previously been replaced. After several subsequent cranking attempts, the vehicle fired into life and fired up ten times in a row straight after. Hmm… Once it had started initially, it seemed that starting was no longer an issue!
I aborted the road test and attached the scope to the throttle once again. I revved the vehicle up and down until the symptom re-appeared. Sure enough, the throttle body (which had been recently fitted) was performing to specification and reacting directly in line with the throttle pedal position. Something else had to be breaking down and going wrong causing the cut-out, but what? The next job was to prove the presence of the vital signals.
Figure 3 shows a waveform capture of the cam, crank, injector and ignition signals, taken directly at the ECU. I managed to catch a waveform showing cranking, running and the cut-out. What this waveform proved is that the cam and crank sensor were present (and correct) when the vehicle cut out. The strange element of the capture was that the injection and ignition command signals went AWOL as the cut-out began. This pointed to an ECU malfunction. The inputs were present (cam and crank) but the outputs (the components were tested and operated correctly) had gone missing. I was guided by the test data that the ECU, despite having been previously bench tested, may indeed be faulty. There was only one way to confirm and that was to check powers, grounds and other key signals. If they all checked out – it had to be the ECU misbehaving.
During another Saturday training course, the Polo came up as a discussion point. We were discussing the input, output logic of the ECU. I recalled the test and showed the guys on the course the capture I’d taken from the Polo which provoked much ‘umming’ and ‘arring’. ‘How is that possible?’ they asked. A short while later, Andrew, one of the course attendees, recalled a similar problem which affects some Fiat Punto models. “It’s because of interference from the spark plugs”, said he! Ahhh, was this the symptom of a symptom I was looking for? I had done several of these Punto jobs (they also set an odd cluster of fault codes and usually start misbehaving shortly after a service). On the Fiat, incorrect plugs cause RFI (radio frequency interference) which can cause the ECU to malfunction internally. Could this be what’s happening with the Polo? Is there a way of proving this other than replacing the plugs (my waveforms didn’t show excessive signal ‘noise’)? It was certainly worth a try though so new plugs were ordered and fitted (figure 4 – old spark plug). Low and behold – the symptoms were solved. Well I’ll be…
The difficulty in this process was the lack of diagnostic information relating to the root cause. None of the gathered waveforms suggested there was significant RFI present. In addition to this, no fault codes were present in the ECU and there were no other overt symptoms. So what, you may ask, was the benefit in performing in-depth scope analysis? In my view, diagnostics could be summarised as the process of analysing data and proving what is working well, in the search for that which isn’t working well. The gathered data ruled out many things (cam & crank sensors, coils, injectors), but left a few suspects (ECU & sparkplugs) for which there wasn’t enough evidence to ‘convict’. The fall back in this case was experience. Problem jobs like this, where the root cause cannot be diagnosed without firm evidence, are quite rare. As was shown, collaboration, or bouncing the problem off other technicians, can often cause a spark of inspiration, which in turn leads to the solution.