Annoying running problem. Griff 500

[steve 500]

I would be grateful for any ideas. Over the last few months, whenever I use the Griff hard, either on the motorway for half an hour or so, our around the lanes. After five minutes of more normal (sensible) driving following the fun, the Griff starts running very poorly and will not idle. If I disconnect the ECU and reconnect after 30 seconds all is well again. I do have a fault code reader, and nothing has shown up on that when tested before disconnecting the ECU. I'm assuming that the fault code reader is ok, as if I disconnect the ECU and then test it, it does show the correct code. Any ideas will be appreciated, as its getting a little boring limping to the side of the road every few days!.

[Gatecrasher]

I've had the same sorts of issues before and it's usually the stepper motor. Try cleaning it - there are a few guides on here about how to clean it properly. It's a big black thing screwed into the back of the plenum on the driver's side - needs a deep 32mm socket to remove it, just make sure you don't overtighten it when you put it back in as they're a bit fragile and not cheap.

[steve 500]

Thanks, I will check it at the weekend but I fear that its something else as it also runs very poorly above tickover before resetting the ECU. I'm sure that I have seen smoke from the exhaust when it is playing up. Again, this clears after the ECU has been disconnected.

[blitzracing]

If this is a cat car it may not be controlling the mixture correctly after a long run, as it needs to idle for a short while to set the overall mixture, although the engine condition or a sensor input would have to alter significantly for the fuelling to alter that much. Something like an unstable AFM output when hot could do this. Resetting the ECU will clear any fuelling presets, so the ECU with simply retrim the mixture again as it idles, so compensate for any variables. It would be worth pulling out a plug after its run badly and has cooled and see if its blackened and it is mixture related.

[steve 500]

It played up again when I took the car for a spin yesterday, and I made a point of looking for smoke from the exhausts, and it's amazing how much black smoke there is when you look for it! There was also a strong smell of petrol when I started the car up, and it was so lumpy pulling away that it was dying when pulling away at low revs, so I guess it is a mixture problem. I have a new set of lambda sensors that I could put in, but what other sensors control the mixture?

[red griff roger]

I think there is always a smell of petrol at start up. If your starter motor turns the engine at (say) 1,500 rpm, and it takes (say) 3 seconds to fire, then the engine has turned 75 times. At (say) 5 litres a turn, that is 375 litres of unburnt mix. There is a pong.

I am sure some tecchie will come along and correct this.

[blitzracing]

steve 500 said:

It played up again when I took the car for a spin yesterday, and I made a point of looking for smoke from the exhausts, and it's amazing how much black smoke there is when you look for it! There was also a strong smell of petrol when I started the car up, and it was so lumpy pulling away that it was dying when pulling away at low revs, so I guess it is a mixture problem. I have a new set of lambda sensors that I could put in, but what other sensors control the mixture?

By far the biggest is AFM output, and with some input from the temp sender(s) and RPM and throttle pot. If its getting to the point of puffing out smoke, its seriously rich. The AFM can fail by the output going to high so it overfuels, but not enough to throw a fault code. The AFM gets a pretty hard time getting cooked by the exhaust, so its not unreasonable they fail. You might be lucky and get a fault code from the ECU if you plug in a reader, or else heres a Mark Adams test for the AFM:


Note that you can drive without an airflow meter in case of emergency (i.e. airflow meter disconnected), because the system will drop into a default (limp-home) mode based on throttle opening.

Most airflow meter faults will cause the engine to run excessively rich. However if the airflow meter remains connected whilst defective then the vehicle will probably not run. In most cases the output from a defective airflow meter will be in the range 2.0-2.5 Volts, which is a viable value. This represents a moderate load and will cause heavy over-fuelling without setting a fault code.

Testing is performed in the following manner. Peel back the rubber boot on the airflow meter connector and leave it plugged in to the airflow meter. Set up the digital multimeter to read voltage. Insert the negative probe into the Red/Black wire (sensor ground), and the positive into the Blue/Green wire (Airflow signal).

Turn on the ignition, but do not start the engine. The meter should immediately indicate a reading of approximately 0.3-0.34 Volts. Most defective airflow meters will overshoot to 0.8 Volts or higher, and take at least 2 seconds to come down to the correct voltage.

Now start the engine, and the reading should rise to 1.6 Volts (3.5 Litre engine) to 1.75 Volts (5.0 Litre engine).

The next test is full load, and as with the fuel pressure test it will require use of a rolling road or a steep hill in the same manner. Under full load the voltage should rise to 4.45 Volts (3.5 Litre engine) to 4.95 Volts (5.0 Litre engine).

On this injection system, the idle CO mixture adjuster is provided on the airflow meter. It is located in a boss on the top of the airflow meter, pointing towards the engine. Leaving the multimeter negative probe in the Red/Black wire, move the positive probe to the Blue/Red wire.

Now turn on the ignition but do not start the engine. Observe the voltage. The normal adjustment range is between 0.0 and 3.6 Volts, with the higher Voltages producing higher idle CO values. There are approximately 20 turns of the adjuster screw to cover the entire range.

Annoyingly, the adjustment may be clockwise or anticlockwise to increase the value, and this varies from meter to meter! For this reason it is always preferable to have the multimeter connected in this manner when adjusting idle CO, so that you see can something is actually happening.

Typical Voltages that would be found at this point are between 0.9 to 1.4 Volts for non-catalyst cars. This Voltage is always factory pre-set to 1.8 Volts for catalyst vehicles. A value near to 3.5 Volts will generally produce an idle CO value of 9-10%. These Voltages may be used as safe initial values particularly if no CO measuring equipment is available