I thought I would share this case study will you all. Many have said using Pico Scope is a learning experience and this was my first real diagnosis that couldn’t be found using “conventional testing methods”. Armed with my Pico 4423 I set about diagnosing this strange fault.
2004 Volvo V40 1.9 TD, D4192T3 engine code using Bosch EDC15C engine management
The customer had the vehicle recovered to the workshop with suspected timing belt failure. I removed the upper belt cover but the belt was still on and not slack, cranking the engine the car ran and drove the non-start car into the workshop with a huge sigh of relief that I wouldn’t be taking the cylinder head off.
After talking with the customer myself I managed to find out the car had cut out whist driving. The customer had some years ago suffered a timing belt failure and he had assumed the worst.
As the car was running fine I road tested the vehicle to see what was happening. At around 2500 rpm the vehicle had a bad hesitation and combustion knock. It sounded like fuel starvation. This could easily be driven through as it only effected a narrow rev range. After a while of driving the car eventually cut out. On restarting the engine the engine management light was on.
Using VIDA (Volvo’s dealer tool) the fault codes were read ECM1809 Fuel pressure too low. VIDA also gives guided fault tracing (always good for a laugh!) Interestingly the guided fault tracing first asks if the engine will run – “If the car can be started the fault should not be traced as the fault is not active”
Now it’s time to get out the Pico scope! My first trace was RPS (Rail Pressure Sensor) on Channel A and MAF (Mass Air Flow Sensor) on Channel B. The trace below shows from KOEO (Key On Engine Off) through WOT (Wide Open Throttle) and back to idle.
Now the more experienced of you will already see the fault. I must admit I did not.
In the mean time I asked the receptionist to enquire on service history and if in any doubt it would be a good idea to replace the fuel filter and retest the car. The vehicle had little service history and the customer had only owned the car for 6 months. He agreed to replace the fuel filter as recommended. The element was black. I temporarily fitted a clear hose to check for air in the system, this did not show any air in the supply system to the high pressure pump. It is also worth noting at this point the supply does not have an in tank fuel pump. I also checked the injector leak off which were all even across the rev range.
After the fuel filter was replaced I road tested the car and used the graphing function in VIDA on the MAF and RPS the fault was still present and the RPS could be seen to jump around when held at 2500rpm. My concern was is this cause of effect of the trouble. My gut feeling went with the sensor faulty but why? And how was I going to prove it?
After some research in which the use of high pressure gauge kits were decided that it’s the way to go but unfortunately not available to me at the time. I confirmed power and ground whilst wriggling the wiring harness. I carried out volts drop tests on the signal wire. All tests so far had proved good results. It was recommended to tap the sensor whilst monitoring the signal. The trace below shows power, ground and the signal at idle whilst tapping the sensor. The scope trace below shows this.
This is where it dawned on me the spikes in the first trace were not interference it was actually the signal dropping out. The sensor was shorting out internally causing the voltage to drop to near zero.
The PCM would try to correct this by increasing the rail pressure which would cause the combustion knock and the cutting out was the safe guard in the PCM programming to prevent damage by over pressure probably as the signal came back on line. A replacement rail pressure sensor was fitted and a road test completed with no problems and one happy customer.