Intermittent MIL Light

[CerbWill]

Guys I'm at the end of my tether with this problem. My Cerb 4.2 keeps giving me an intermittent MIL light and logged faults for Lambda 1 & 2, seemingly because the voltage from the sensors exceeds 1.8V. It started after I'd had the intakes & fuel rail off to repair a cracked cam cover. So far I've:

Replaced the lambda sensors.
Replaced coil packs. Took the intakes & fuel rail off again to change the packs, so I checked out the heat barriers & changed the o-rings to ensure a good seal.
Replaced HT leads.
Checked ignition using ignition lead testers that flash if there's a spark.
Checked 12v supply to lambdas.
Checked continuity of signal wire from lambda sensors to ECU.
Checked continuity to earth.
Replaced lambdas again with brand new NTK sensors.

Any ideas most welcome!

[An Marcach]

Given that you've checked and changed all the obvious bits I'd be starting to look elsewhere for a damaged wire or bad connector.

Now I appreciate that trying to trace a wire in a Cerbera isn't easy but I'd start with the connectors on the bits you've changed with wires on them and work outwards from the area where you did the initial job.

You might find it's a bad connection on the ambient air sensor or something like that..

On the few occasions I've had the MIL light come on it was always due to a connection.

[LincsCerb]

Hi Will
As your cerb is a 96 you may have the connector in the main wiring loom under the expansion tank. It's a common place for moisture to get in and can cause lots of problems.
Otherwise are you sure the lambdas haven't got crossed over and are connected to the wrong bank.
If you need any help just let me know and I'll come across.

[CerbWill]

Thanks Martin, typically after having done this for a couple of weeks as I tried various things to fix it I've driven the 10 miles home without a single flash of the MIL light, lets see what tomorrow brings. I expect it'll come back, but the intermittent nature of it is pointing more towards a wiring fault that component failure in my opinion.

[Supateg]

CerbWill said:

Thanks Martin, typically after having done this for a couple of weeks as I tried various things to fix it I've driven the 10 miles home without a single flash of the MIL light, lets see what tomorrow brings. I expect it'll come back, but the intermittent nature of it is pointing more towards a wiring fault that component failure in my opinion.

Could be the warm weather had dried a connector out?

[CerbWill]

Maybe, although I'd have thought engine bay heat would have done that sooner. However, another 10 miles done to work this morning with no MIL light.

[Englishman]

I've had a similar issue on the Cerb for a long time with an occasional lambda 1 fault. You change something, its ok for weeks and then it happens again!

I left the Cerb outside overnight as sealing the garage floor only for the MIL light to come on when started. Went off soon after, but having to stop at lights on an upward hill started it again. Once heated up has been fine though. So looks like a little moisture condensing somewhere creating an intermittent connection, made worse by tilting.

The lambda has been replaced twice, the connectors renewed, any other ideas???

[Tanguero]

I used to get this regularly on mine until I changed the lambda sensor connectors for properly waterproof ones. I have never had it since.

Edited by Tanguero on Wednesday 1st July 22:40

[Englishman]

I replaced the TVR ones with new TVR ones a while ago but wonder if moisture is still getting in. What connectors did you use?

[Tanguero]

I used AMP Supaseal, the ultra heavy duty ones.

[CerbWill]

And it's back on today. Time to get the engine loom out of the car for a thorough check over I think. Might as well change the sensor connectors while I'm at it.

[CerbWill]

I think I've fixed it. The wiring to one of the connectors had been repaired before and didn't look in great shape. Anyway, new AMP Superseal connectors and all seems well. The peak lambda voltage is now much lower, about 1.30-1.35v. I guess that's still enough to tell the ECU the mixture is rich?

[Supateg]

CerbWill said:

I think I've fixed it. The wiring to one of the connectors had been repaired before and didn't look in great shape. Anyway, new AMP Superseal connectors and all seems well. The peak lambda voltage is now much lower, about 1.30-1.35v. I guess that's still enough to tell the ECU the mixture is rich?

1.7v triggers a fault; looks ok as long as the reading is switching
Aides app is great for showing analogue switching.

[CerbWill]

They switch fine. Before the voltages went high enough to trigger a fault warning they'd be up at about 1.5v for a rich reading, so I wondered if the slightly lower voltage reading I'm getting now would make any difference, but it seems not.