Method over madness

James Dillon explains why diagnostics is more than just code reading

Published:  24 February, 2016

The focus of this article is a recent problematic vehicle which was booked into the Technical Topics workshop. The story behind the vehicle, a 2011 plate Audi A3 2.0 TDI, DSG, Quattro, is not an unfamiliar one. The car was clean, tidy with very nice specification. It was presented to me by a motor trader who had sourced it through one of the major auction houses. My internal alarm bell was ringing, as the vehicle looked as if it should be retailed on a forecourt. The mileage, engine, condition and specification were all ‘right’, the only thing that was ‘wrong’ was the vehicle’s behaviour. The reason that it was in an auction will now become apparent.

Persistent symptoms

The description of the fault was that every once in a while, particularly when turning, but not exclusively so, the vehicle would cut out for a second or two. It would immediately restart and be fine until the next time. The next time might be 10 seconds later, or three days later. The vehicle had previously been seen by Audi, who had retrieved fault codes for communication errors with the steering wheel control unit. This had been replaced, but the symptom prevailed. The vehicle was rebooked with them, and their final diagnosis was an onboard power supply unit. This had been replaced but the symptom was still present.

It is interesting to consider whether these, or other parts, had been replaced by the previous owner before the vehicle was put to auction. It is highly likely that the unfaulty, old, new parts had just been replaced with new, new parts. The money spent on this vehicle trying to cure this fault was running into the thousands. In spite of how much money had been spent, the vehicle’s symptom was exactly the same. Pure madness! It is highly likely that because of the fault’s intermittent nature that the diagnosis may have been based on trouble codes alone. The symptom was intermittent and when it did occur it happened really fast. I am not criticising the other repairers, using fault codes alone would explain why the other garages chose to take this course of action.

The fact that the ECUs had been replaced made my job a little easier. I knew that the computers couldn’t be at the root of the problem, but they were being affected by something. This made a good place to begin to develop my test plan.

Testing time

Firstly a visual inspection was in order to check for issues such as water ingress and obvious chaffing or damage to the wiring. After this I worked through the wiring diagrams to begin to form an idea of where and what to measure. My goal was to establish wiring which has a common link to both the replaced components and the symptoms. The vehicle manufacturer’s wiring diagrams for this vehicle can take a bit of getting used to and I used the whiteboard as a jotter to provide the bones of a test plan.

This research showed that there was a key voltage supply and ground which were related to the Steering Column Control Unit, the one which had been replaced. The image below shows the wiring diagram.

The steering column control module was supplied from a fuse SB17 from the under bonnet fuse box and was grounded in the loom. The next step was to measure the supply voltage and ground at the control unit whilst the fault occurred. I hooked up the meter and began a road test. Unfortunately, the symptom would not present whilst under test, and no symptom usually means no diagnosis. After a few miles, I headed back and decided to change my approach.

It is always worth remembering that there are only three faults that can affect a wire; Open Circuit, Short Circuit and Too Much Resistance. Looking at the system’s wiring, the supply wire was protected by fuse SB17. During the time the fault had occurred, this fuse had never blown. Therefore, the fault was likely to be either an open circuit (in the supply or the ground), or too much resistance (in the supply or the ground). Resistances and open circuits can both be intermittent (consider a bad earth/loose connection).

Elusive issue

I felt that I needed to prove that the power supply and the ground to the control unit were OK. I decided to perform what Ford used to term a ‘wiggle’ test. Whilst the meter was connected to the power and ground across the ECU (with it switched on) I measured the volt drop. I proceeded to move the wiring loom and the multi plugs connected to the steering column to see if there was an issue. The voltages were exactly as expected, no drop outs. This ruled out connection issues in this section of the loom and multi plug, but the vehicle could still have power or ground issues elsewhere.

I moved on to check the component supply, as this had a definitive start point and was easy to identify, unlike the ground which terminated somewhere in a loom splice. I performed a volt drop test focussing on the voltage supply wire. It ran from the under bonnet fuse box to the Steering Column Control Unit, around 1.5 metres of wire. I commenced the wiggle test of the engine bay wiring loom, whilst observing the meter reading. When I pulled on the harness to the front of the battery tray, the volt drop remained the same. Next, I flexed the battery tray and I saw the volt drop go from 12 volts to zero volts. On this model, the wiring loom runs underneath the battery tray. I whipped out the battery and the tray and investigated the loom as it lay underneath. It was still in its conduit but there was a very small witness mark on the loom tape where the battery tray appeared to have rubbed on the loom. I released the loom and pulled it open to reveal the image in the image below.

It can be clearly seen that the loom, and in particular the red and white wire we had previously identified, was suffering from a pinch mark. What was also obvious on closer inspection was the white powder residue which results from water ingress in a wire. When I pulled on the wire, it broke clean in two. The battery tray and loom must have been interacting during turns and whilst going over bumps, causing the wire to separate and to cut the supply to the Steering Column Control Unit. The solution was simply to fit a new section of loom and after a period of prolonged use to prove it was cured, the vehicle was classified as fixed.

When I wrote the job up and presented the conclusion, I got the anticipated response “is that all it was, a broken wire”? Yes, that is all it was, and no, there isn’t a fault code for that sort of fault. When asked why the other garage couldn’t fix it, I explained that the trader had now experienced the difference between applying a diagnostic machine to a problem and applying a diagnostic technician to a problem. After all, owning a stethoscope doesn’t make you a doctor, does it?

Related Articles


©DFA Aftermarket Media Ltd
Terms and Conditions