The MGF Register Forums

I finally had another chance to investigate my misfire on my '96 VVC at the weekend. As before I drove about 20 miles with the car running perfectly, happily revving to over 6500 rpm. I got home just as it started misfiring and it got progressively worse as I left it idling and the temperatures rose.

I got a pscan a couple of weeks ago and expected to see a fault logged. There were no fault codes and most of the live readings looked entirely normal, even when the engine could barely idle!

The one reading that I am uncertain about is the o2 sensor heater. It is always on. Is this correct or should it turn off once the exhaust is hot? As far as I can tell from the circuit diagram there is no feedback from the heater to tell the ECU that it is on, so I believe the reading is the output status.

What does "Cam input history" mean. I see "Input was low".

Are the water and oil temperatures up to spec when the engine is hot (93-104 °C water, 90-120 °C oil)? When a car is running normally when it’s cold but has misfires when hot, this could hint to a failing temp sensor, I had one here which showed a water temp of just 55-60 °C after an autobahn blast. As this is theoretically a reasonable value, it was not recorded as a fault, but obviously caused some havoc, like the fan not coming on in stop-and-go traffic. Also, the older MEMS (like on your car) cannot easily adjust fuelling as easy as the modern ones can, when a sensor has failed. I even know of a case where an early MGF would not start when hot, because the ECU still thought the engine to be cold and overfuelled on start.

Another thing could be the ht leads, which might be a bit brittle and, getting more flexible with the engine warming up, could short to the block somewhere on their (quite long) way from the coils on a MEMS2 VVC. This would also not be recorded in the fault memory.

Yes, when I got home the water temperature was 86C and the oil was 90C. I left it idling and the water temperature rose until the fan came on at around 104C and then fell.

So far i've fitted new coils, HT leads, water temperature sensor, inlet air temperature sensor, throttle position sensor and spark plugs.

It had a new O2 sensor as the emissions were high at the last MOT. I'm beginning to wonder if it is failing already. The engine briefly stops misfiring if I keep moving the throttle. It seems as though the engine runs properly under conditions where it is running on a open loop map rather than closed loop. That might fit with sensor giving wrong information.

Crank or cam sensor?

Failed crank sensor = engine wouldn’t start, iirc.
Failed cam sensor on a VVC = rev limiter kicks in @5,400 rpm.

Both would produce a fault code entry in the ECU memory.

Did you get an original lambda sensor or a cheap aftermarket one with open wires to be crimped on?

MAP sensor could be an idea, as long as the values it produces are within the normal range it wouldn’t be considered as faulty by the ECU, but they might be wrong nevertheless.

I forgot to add the MAP sensor to the list of things I have already replaced.

The lambda sensor was fitted by the garage when it failed the MOT. It doesn't look like a really cheap one but I don't know if it's a genuine Rover one. I am thinking it might be best to get a genuine one and see what happens. I have spent a small fortune on new sensors already!

If the heater circuit is switched on and off it will be controlled by the ECU - it is there to get the probe up to temp quickly to enable early accurate readings. I can't see it being a problem if the heater is always on, the exhaust will run hotter anyway.