Voltage irregularities
Q During warm up idle, my 2003 Corsa 1.0XE, with 174,000 miles, occasionally – and especially on a frosty morning – just as the idle speed reduces to normal, there can be a hiccup in the idling with the engine nearly stalling.
This only occurs in the first minutes after starting – there are no problems starting even on a very cold morning. Thereafter no issues at all. Suspecting the coolant sensor I read the voltage feed to the sensor, 5.0V, and the voltage across the sensor, with the ignition on, around 4.5 or less depending on the engine temperature, and these voltages remained steady.
However, using breakout leads and measuring the coolant sensor voltage with the engine running, momentary voltage fluctuations happened fairly regularly ie. 3.99V to 1.65V, then back to 3.98V, followed by more sudden drops with the gradual expected voltage drop as the engine warmed up being the more prevalent reading. The voltage drops are less frequent on a warm engine but do still occur. I have warmed up the sensor in water with a separate DC supply with no sudden changes in voltage. I have cleaned the pins in the engine ECU which were very clean anyway and measured the resistance of the two leads in the coolant circuit which were OK and tried a direct earth from the earthing points beside the EGR valve.
Using a basic fault code reader, the live data during warm up shows a steady rise in coolant temperature with no fluctuations. Does this point to a faulty ECU? I would be very grateful for your opinion.
Richard Pattison
The best test of the sensor is to use an ohm meter with the sensor disconnected, this will give a definitive answer to the condition of the sensor without any possibility of the reading being corrupted. The reading across the sensor should be 1600-1700 Ω with a cold engine (20°C) and around 200-350 Ω with a hot engine (80°C). With a gentle progression between the two.
The readings from the ECU pin A38 (blue wire) using a voltmeter should be 3.4 volts at 20°C and 0.9V at 80°C. The pin A9 Brown wire is the earth and should give a zero reading. If this is not the case then the earth on the ECU should be checked to ensure this is not corroded. From the details I have the earth should be via the case and double checking your readings with a separated earth lead from the case to the body may help determine if this satisfactory.
If readings from the sensor are correct and the reading from the ECU are fluctuating, then as you suspect the ECU may be at fault and it would be worth sending this for testing.