Vectra Light failure

Modern cars have the ability to monitor circuits and report errors when they are detected.

Modern light systems have the ability to warn the driver that a circuit failure has occurred with a Malfunction Indicator Lamp. The driver can then take the car to the workshop for diagnosis and repair.

The owner of one such vehicle was certain that he faced an expensive repair when the lighting circuit MIL illuminated on his Vauxhall, but when he checked the lights he could not find the one not working.

A quick code read suggested a fault with the rear left brake light circuit. However the brake lamps appeared to be working when the pedal was pressed. The light cluster was accessed and the lamps checked. The nearside brake lamp had indeed failed.

This led to a number of questions from the puzzled driver.

So how did the car know?

Why are all the lamps the same?

Why did the lamp appear to work when the pedal was pressed?

The answers are linked to how modern lighting circuits are controlled. The use of vehicle networks has reduced the number of fuses and relays by as much as half. But every circuit must have a form of protection, in the case of modern lighting it is the control unit that monitors the current drawn by circuits. It can then switch the circuit off if it detects a fault. It can then elect to use another lamp and circuit to replace the faulty one. In this case the brake light circuit was inoperative so the side light circuit was used to perform the task of the brake light. This is possible as the lights are controlled using a pulse width modulated signal. The brake lights require full intensity so have a pulse width of 100% but the side lights only require around 25% of the 21Watts available. By switching the circuit 278 times a second the 25% duty cycle is seen by the lamp as a 3Volt supply. (This is what you would read on a multi-meter) The 21 Watt lamp illuminates at around a quarter of its intensity, or 5 Watts the same as a tail light. The means the same lamp can perform the task of tail and brake lights. Using just one lamp makes economic sense as it reduces inventory.

The control unit sends a signal to the lamps and checks the circuit before it is used. Here the brake light circuit is being monitored every 10 seconds. The voltage is pulsed so fast that the filament does not even start to glow. This is how the control unit can detect a fault before the circuit is used.

Using an oscilloscope you can check the current draw during the circuit check. Here the time base has been reduced to 1ms per division. The red trace shown indicates a current of 9 Amps on a brake light circuit. This has to be a fault. Or is it?

Once the lamp is switched on (100% duty cycle) the current measured is around 1.7 Amps. This is normal for a 21 Watt lamp.

Watts law current = power/voltage

21/12 = 1.75 Amps.



Same lamp same circuit. Different current draw, this is because once the lamp heats up and starts to glow its resistance increases. Try it for yourself with a multi-meter, measure the resistance of a cold 21 Watt lamp. Then use Ohms law to check to expected current flow.

Remember Amps = Voltage/resistance.

Lost your spark

Modern ignition systems have evolved from contact breaker or points systems. If you look carefully you can still see some of the DNA of these systems in the latest systems fitted today. So is it reasonable to assume that some of the tests performed on older ignition systems can still be performed on these modern systems?

The answer is yes and no. Some of the old tests can still be performed when there is suitable access but many of the tests focused on the High Tension or secondary side of the coil. However access to the high tension side is often only possible after removing the coil pack, you can use an extension between the coil and the plug to test the HT outputs. These tests can be performed with an oscilloscope or a spark tester.

Some (older) readers will remember measuring dwell angles. Testing the LT or primary circuit, has changed little since the days of points, a test lamp can still provide quick and effective proof of circuit integrity. However an oscilloscope can provide extra details that can lead to more effective diagnosis. With an oscilloscope it is possible to analyse the current draw as well as the control signal/voltage at the same time. The reason why this is so important becomes clear when you consider how the modern ignition system is controlled.

Modern ignition systems consist of various inputs, logic and outputs. Inputs are from sensors such as crankshaft position, (engine speed & position) Manifold absolute pressure, or Air Mass (load) and knock sensors (abnormal combustion).

The logic or ECU, crunches the numbers and selects the correct ignition advance, dwell period as well as monitoring the circuit for faults and providing the circuit protection.

The outputs are the low tension circuits, fault codes and malfunction indicator lamp. Or in the case of amplified coils a signal to switch the amplifier and often a conformation of ignition signal back to the ECU.

These control signals can be an internal function of the ECU or in the case of amplified coils a square waveform that is used to switch the primary coil. The on time of the coil is controlled by this signal. The output stage allows current to flow through the primary windings when the voltage is present and stops the flow of current when the voltage drops to 0V. This ‘dwell’ can be measured much like on the older systems. The yellow trace is the control signal and the blue trace is the current flow through the primary windings. 

The cursors are measuring the on time or dwell, in this case 3.16ms. A typical value for a running engine. Notice the coil has reached around 5.5 Amps. Then current in the circuit is limited. The primary coil windings have low resistances, between 0.2 and 0.8 Ohms. This allows a rapid build-up of current, and can reach 60 Amps if left unchecked in around 40ms. The current in the circuit is dependent upon the voltage so to ensure good saturation the ECU can compensate for low voltages. The chart shows the current build up in a coil of 0.2Ω for both 12 and 8 Volts. 

Current Flow Diagram Diagnosis

The ever increasing electrical and electronic content of modern vehicles can make diagnosis and repairs a real challenge.
We have seen a number of faults that without the correct approach, could lead technicians to at best take much longer to diagnose or at worse fail to or misdiagnose.
During training we encourage technicians to analyse how effective their diagnostic routine actually is.
Some have a set routine, others fly by the seat of their pants.
What ever technique you use, could you improve your diagnostic skills?
We have discovered that simple faults can fox even experienced technicians.
Take a simple central locking fault, the vehicle will lock and unlock all doors from the passenger door lock and the remote, but if the drivers door lock is used only the unlock function works.
This suggests to me,a fault with the door lock switch.
But how can this be proved quickly without stripping unnecessary trim from the vehicle.
The wiring or current flow diagram will often hold the key.
We recommend drawing your own diagram, it should contain only the detail you require to test the circuit.
This includes wires colours, pin numbers and what you expect to see on the voltmeter.

Take a look at the diagram below, where is the likely fault, where can you test easily, and what would you expect to see.

When the whole diagram is presented it can be difficult to see the wood from the trees.
However it can be easier with your own diagram. Like the one below.

Testing at the module means the door trim does not have to be removed, if the voltmeter reads 0.1Volts at pins 12, 13 & 9 when the switches make a path to ground then the circuits and switches are OK.
At pin 12 the voltmeter reads 0.1Volt when the passenger door switch is made. But remains at 12Volts when the drivers switch is made. The fault can only be between the switch and the join in the wiring.
As this join is inside the car it is possible to trace the wire, it was found to be broken in the door hinge area, a common failure due to the constant opening and closing of the door. A quick repair and normal operation was restored. All without ordering parts or stripping door trims.

Calm under Pressure

The Peugeot 307 2.0 HDi came into the workshop with a history of failed repairs and an inventry of second hand parts fitted in an attempt to get it running correctly.

We have identified the ECU as the culprit, the internal earth path for the rail pressure sensor has failed.
With the ECU repaired, we could consider what has caused the rail pressure to increase without any command from the ECU.
The engine would start and idle but it would cut out when the rail pressure exceed set limits. (400+ bar)
We monitored the current drawn by the pressure regulator during the pressure increase and it did not alter.
This means the increase in pressure is down to a mechanical fault, not an electrical fault.
All high pressure fault diagnosis must start with the low pressure system.
In this case a higher low pressure could possibly result in higher, high pressure values.
A gauge was set up to monitor the low pressure, and an oscilloscope analysed the high pressure voltage from the rail pressure sensor. (Remember the original fault affected this sensor)
The engine was started and the rail pressure settled to a even 1.3V with a low pressure supply of 4.5 Bar.
Then the engine cut out. The Supply pressure remained constant however the rail pressure increased at the time the engine cut out.

The return lines were checked for damage and they all appeared to be fine, so what is causing the rise in rail pressure?
This is where system understanding is key to the diagnosis. The pressure is regulated by a valve, this valve allows fuel to bleed past to reduce the pressure. It has a spring inside that holds around 80-100 Bar in the rail when there is no current flowing to increase this pressure by means of an electro-magnet forcing the valve closed. A cranking check using live data showed the pressure was in excess of 120Bar in live data with the valve disconnected.

The valve must be sticking or blocked. The valve can be removed for replacement or inspection.
In this case a large amount of metal swarf was found inside the fine filter that is fitted, this must be preventing the fuel escaping. We cleaned the filter and replaced the valve. The swarf must be coming from the fuel system, and without a comprehensive repair history, we could not be certain the fault had been repaired or was the component still breaking down.

Due to the high costs involved it was decided that the best way forward was to return the vehicle to the customer. The vehicle has been driven for 3-5 thousand miles without issue.

HDI rail pressure fault codes

A 2002 Peugeot 307 2.0 HDi was presented with multiple fault codes.

It was in reduced performance mode, and driving the owner and a few local garages crazy.
The codes would clear and return, and the car would drive far better with the air mass meter disconnected.
This has led to incorrect diagnosis of the AMM, and the replacement of many components with 2nd hand parts to try and eliminate the cause all without success.

A quick scan produced the same result as the previous garages, DTC's.

The codes were as follows;

P0190 rail pressure sensor


As the Rail Pressure Sensor is a primary input and sure to place the ECU into reduced performance mode it seemed logical to start here.

A quick KOEO test showed a good 5v supply, a signal of 9.19V and an earth of 4.98V.

Bingo the first problem to solve, I unplugged the sensor hoping the readings would change however they did not. Next I tested the wiring back to the ECU pins, all OK no shorts or open circuits. The problem must be in the ECU! To prove the fault I provided the rail pressure sensor with a temporary earth direct to the battery, the codes cleared.

The ECU registered normal rail pressure during cranking and the engine fired into life.

Then suddenly the engine cut out.

A quick code check now produced;

P0380 Glow Plug/Heater Circuit "A" Malfunction.
P1112 Diesel high pressure monitoring system Malfunction. P1465 A/C Relay circuit malfunction

It was the rail pressure code that caused the engine to cut out.
I cleared the codes and started the engine again, the rail pressure rose as expected during cranking, then settled to around 360 bar then climbed to 400 bar and the engine stalled. The duty cycle of the pressure regulator was fixed at this time and the idle speed constant. So what caused the rail pressure to rise?

Oscilloscope training

Technicians know they need an oscilloscope, accurate diagnosis is often only possible with one.
But is having a scope in the garage enough?
Many technicians who have purchased a automotive scope have not had suitable training in the use of the equipment. This makes using the oscilloscope more difficult when those problem jobs present themselves.
A good example is compression testing, on a modern diesel engine the thought of performing a compression test is at best risky, at worse a nightmare. Removing injectors or glow plugs to access the cylinders can result in hours of frustration or damaged components.
Imagine if this test could be performed in seconds without any parts being removed.
It can using an oscilloscope, some even have programs built into the software to allow quick easy relative compression tests to be carried out.

Civic Type R Idle fault

On the phone this sounded like a simple problem, a Honda Civic Type R has a high idle when cold and is hunting badly when warm.
Got to be an air leak, Right?

The garage has performed a smoke test and no leaks were found in the inlet system. No fault codes are stored in the fault memory and there are no pending codes.

I performed a visual inspection, and everything appeared to be good, the vehicle had a full service history and had just had a service when this fault was mentioned to the garage. The symptoms were just as described on the phone, and once warm the vehicle starting hunting. A quick test ( I placed my hand across the throttle body-the engine stalled immediately) proved the smoke test was right the vehicle had no air leak.

The scan tool showed normal live data, and no DTC's. However the idle speed was not being controlled by the ECU correctly.
If there was a circuit fault there should be a DTC. This could be a mechanical fault with the idle speed control circuit. I removed the throttle body and ISCV this is the rotary type valve on this vehicle and a build up of carbon had resulted in the valve sticking. These valves can be stripped down and cleaned. This was down and the throttle body re-fitted. The idle speed must now be reset using the scan tool, and allowing the engine to idle once warm for 10 minutes.

The vehicle no performed as it should, proving the relying on DTC's is a big problem if the fault is mechanical.