Injection problem - help needed

[Scania]

My Griff did not pass the yearly "kontrollbesiktning" (=Swedish for MOT) due to too high CO and HC values. Also from time to time it does not run as it should below 3' rpm. When starting up it runs OK but after a while it either stalls or almost stalls combined with a very poor running, not all cylinders. In order to identify the problem I got an Ecumate and ran the normal tests. This indicated problems with the Lambda sensors so I replaced them. Unfortunately the problem was still there.

I have noted the following test results from using the Ecumate
- No fault codes after replacing the lambda sensors (before replacing it I had fault codes 44 and 45 = lambda sensors out of range)
- Water and fuel temperatures OK, TPOT value OK and changes when the throttle is moved
- Throttle voltage and % normal values. Does not enter into Limp home mode
- Fuel pump values OK
- Lambda display is OK for a while when starting the engine. Then it gets stuck in either Lean or Rich and stays there until I kick the throttle pedal or restart the engine. When revving above 3' rpm it also returns to normal values for a while.
- AFM voltages are OK
- Stepper motor seems to work OK. When extending and retracting with the Ecumate it changes the rpm as it should
- ROAD and IDL indicators are OK

I also have done the following tests
- checking that there is OK connection from the lambda sensors signal cables to pins 23 and 24 on the ECU connector
- checking the power supply to the lambda sensors = 13,8 volts with engine running seems OK
- checking for injector problems by testing resistance on pins 2-13 and 2-11. Result 4,5 ohms should be OK
- disconnecting the lambda sensors and running the engine makes no obvious difference. The Ecumate even shows lambda values in the same way as when the sensors are connected. This I do not understand, rather I would expect values to be fixed or blank.

One idea would obviously be to have a TVR specialist to look at the car. Unfortunately they are difficult to find here in Sweden...

Well, thats where I am at the moment. Are there anyone out there with similar experience? Any ideas of next steps would be appreciated!

Regards Niklas

[blitzracing]

"Then it gets stuck in either Lean or Rich and stays there until I kick the throttle pedal or restart the engine"

Which way? A misfire will give a 255 lean reading.

[FlipFlopGriff]

I don't have a detailed response but you could try diconnecting the ECU for 30 seconds or more and see if it resets itself.
FFG

[blitzracing]

Dont forget if you do this to let the car idle for 2.5 mins at full temp to allow the long term fuel trim to recalibrate. If you dont do this, the short term fuel trim (lambda) readings may be skewed as the ECU tries to compensate for the long term trim being reset to zero.

[Scania]

blitzracing said:

"Then it gets stuck in either Lean or Rich and stays there until I kick the throttle pedal or restart the engine"

Which way? A misfire will give a 255 lean reading.

Not 100% sure but I would say it is more often Rich (=value 255X) than Lean (=value 0X). What I do not understand is that the values on the Ecumate appear to be the same whether the lambdas are connected or not.

[Scania]

FlipFlopGriff said:

I don't have a detailed response but you could try diconnecting the ECU for 30 seconds or more and see if it resets itself.
FFG

Tried this also but no difference, thanks.

[Scania]

blitzracing said:

Dont forget if you do this to let the car idle for 2.5 mins at full temp to allow the long term fuel trim to recalibrate. If you dont do this, the short term fuel trim (lambda) readings may be skewed as the ECU tries to compensate for the long term trim being reset to zero.

The car has been on idle for long time so it should not be a problem.

Do you think that there could be a mechanical problem behind? E.g. faulty injectors or other problems?

[blitzracing]

The ECUmate reports back what the ECU is trying to do to get the fueling correct- this is not the same as the lambda voltages, but the response to the voltages. You may just be introducing electrical noise to the ECU if the lambdas are disconnected that the ecu is responding to- the voltages are small- typically 0 to 1.2 volts. The fact that the ECUmate is showing both rich and lean may not be an issue, as the ECU can overshoot as it corrects the mixture so you can readings both ways once in a while, If its stuck rich or lean that that is an issue- is that what happens?. Also the ECU may well be trying compensate for an underlying error (ignition or fuel) and making bigger fueling corrections. You need to take a step backwards and look at the bigger picture- does the engine sound like its misfiring? What colour are the plugs across all 8 cylinders- really sooty ones are often a misfire. If its injectors you will see marked difference between plug colours.

Edited by blitzracing on Thursday 1st August 16:15

Edited by blitzracing on Thursday 1st August 16:17

[Scania]

"If its stuck rich or lean that that is an issue- is that what happens?" Yes it is.

Thanks for the advice, it makes sense to go the direction you suggest. Being away from home at the moment it will take a week or two before the next try. I will post some feedback then.

[EGB]

Mine had high CO and HC readings at MOT because no one plug lead had fallen off. The Crypton was sucking up neat 99 octane fuel. Check plug leads including HT to coil, plugs, Ig amp, dizzy etc.

[ChimpOnGas]

Additional checks:

1) Fuel pressure

2) Manifold/plenum vacuum

3) Fuel pressure regulator operation

If you have an induction leak, you will have insufficient vacuum in the manifold/plenum.

If you have insufficient vacuum in the manifold/plenum the fuel pressure regulator return port will not be opened sufficiently.

If the fuel pressure regulator return port is not being opened sufficiently you will experience excessive fuel pressure at the injector rail.

If you have excessive fuel pressure at the injector rail your injectors will be overcome causing over fuelling.

All this normally happens at idle when the vacuum in the manifold/plenum is at it's lowest value.

Even a small induction leak can be enough to mean the fuel pressure regulator return port is not being opened sufficiently to maintain the correct fuel pressure at the rail.

Or it could be as simple a dislodged or split vacuum pipe to the FPR.

Start by fitting a fuel pressure gauge to the Shraeder valve on the injector rail, you're looking for 36/37 psi.

Its also worth noting an overly lean mixture can give very rich symptoms, this sounds counter-intuitive but if the lean condition is causing a misfire the unburnt fuel will result in a very high HC reading at the tail pipe.

• Check your fuel pressure with a fuel pressure gauge

• Check your manifold/plenum vacuum with a vac tester

• Check your FPR by artificially applying vacuum to the FPR using a vac tester

I'm not saying any of this is your problem, but you need to include these checks as part of your diagnostic process.

[blitzracing]

This is worth a read on stuck lambda voltages:

http://www.g33.co.uk/images/PDFS/14cux%20faulot%20...