2.0 TDI CBBB Roadster - Exhaust Gas Temperature Sensors (voltages?)

[markowen4]

I've been having so problems with my TDI 8J Roadster CBBB engine code. New to forum, so hello!

No MIL / Engine light or fault codes are coming up, Well, HVAC blower resistor, aftermarket aerial faults in VCDS, but nothing engine related. For a while now I get what's called a 'warm start', so after a short running period, or long period with a bit of cooling, the engine fails to turn over. Feeding a bit of throttle can get it going, as can turning off the AC, seats to give a few more Amps. The fact the throttle is by wire, suggests it relates to the following error on temperature sensors. I can live with the occasional, albeit embarrassing cough it does when starting, but recently it's going into limp mode and when warm (dash temp level) is out of limp mode. Occasionally falling back into it when on the motorway runs I've done recently and cruise disappearing and power dropping. No errors, no lights, no thing to go on...

I've been digging in VCDS at up at measuring blocks 99.2 (turbo prior), 100.1 (prior DPF) and 100.2 (pre catalytic converter) for the exhaust gas temperature sensors. After a 90 mi motorway run down the M1 they read 192 degC, 150degC and 162degC respectively. So... not right? I'll try and get a reading for when its cold, but don't have that to hand.

I've swapped the prior DPF temp sensor over for a new one, as easiest to get to and given it's just under 150k mi on the clock in my mind no harm in some new sensors for it. I've also the post DPF and Post Turbo sensors to change if I feel brave enough for the post DFP one and the fiddly nature of them. The new sensor didn't change the reading, unless 150degC after a run sounds right for a DPF?

Today I've been playing around with the multimeter with the sensors. With ignition on, I get c. 3.2volts supplied to the sensors. So first question, should this be c. 5 volts as supplied by the ECU? I appreciate that the ECU is measuring resistivity. But I'm presuming that it would supply c.5volts and this affects the measurement? Or completely faulty and what's measured in VCDS are default values the ECU is using?

There's continuity to ground from all sensor plugs (black, orange and brown), measured from the sensor pin brown wires to the negative post in the engine bay. Since all three are misreporting temperatures, my thinking is that either the signal ground and their connection to it is faulty. Or given my poking with the multimeter, the supply voltage rail, or some other small resistor/transistor is broken in the ECU.

The wiring diagram points to a common ground, 411 - Earth connection 2 (sender earth) in main wiring harness, but I get continuity. Is there a means of testing the ground further or is it that supply voltage to the sensors? It seems there's three separate wires at fault, and I'm not sure where that ground is or where those three ground lines join together and why not other faults if that's the case? So to me suggests the ECU?

Wondering if anyone's got some thoughts as to what to double check next, or does it point to the ECU (which I've replaced once already, so might suggest something...) and I should Jiffy bag it up and send it for diagnosis and repair?

cheers in advance

[Crasher]

Pressing the throttle has no effect on the fueling, all it doe is give the ECU a torque demand so I can't see how pressing it chnages anything. If it is definitely going into limp mode then it must leave a code so if not, it suggests a running issue the ECU is not picking up. The EGT's can be all over the place, don't take too much notice of them except when carrying out a service regeneration. Yes ECU's can suffer water damage but if the sensor reference voltage 1 or 2 fails, it flags a code, how was the last ECU replaced?

[markowen4]

I sent a donor ECU in with the old one to clone the details over. Was three years ago, ran well for a couple of years, but a second hand ECU.

Pressing throttle on a warm start is ancedotal but to me the torque response (am aware there's no direct connection to fuelling) it sends is matching a value/request that gets the engine to turn over, anecdotal though but seems to work, or could just be a couple of tries convinces it. So in my mind it sounds like the ECU is all over the place with what it thinks is going on at that engine temp.

There's definitely no limp mode flagged, but it dropping availability of cruise, drop in power, and not recording it is another issue that points to something being wrong. There might be likely other values all over the place but it's not always obvious as to what they should be, nor am I familiar enough to know what they should be.

I'm inclined to think water damage is a possibility as the drains under the scuttle on a roadster, or at least on mine, are a PITA. The ECU is at least on the high point under there, but most recently when I noticed them clogged and not draining was more recent than the warm start issues. Possible though, no idea if the high water mark was at the pins or ECU case.

TBH, I'm thinking shifting it as spares or repairs / part ex now, wanted to get another summer out of it, but looking less likely. If it wasn't an obvious ground or wiring fault I'm leaning towards getting rid. It's already drained my wallet enough.

Maybe one last roll of the dice though with getting the ECU diagnosed? Seeing what that comes back with, but likely yet another bill for a car I finally fell out of love with last weekend when rain started coming in the window seal (fixed now mind, but much swearing was involved at the time), basically posting on here to convince myself of one or the other option

[markowen4]

although water damage suggest a short or corrosion which as you say would throw errors.

This seems to be an issue that's getting progressively worse over time

Would EGTs being all over the place stop a regen? The values for that 'appear' ok, been a while since its last regen (3k km I think in VCDS) and was hoping recent run would do another, but not really able to get up to temps to trigger one?

The fact it's doing far too many short runs as well to the shops and no long runs (a sin with diesels) doesn't help.

Argh!

[Crasher]

I am having recurring nightmares with 8J roadster and water leaks. I fitted a new fuel pump controller which was water damaged but no apparent leaks and all around was dry, then a few months later the car broke down again and the new (genuine) controller had failed but was dry so we fitted a new one under warranty and the engine cut out again after 10 mins... We decided the pump was faulty and on a roadster this is a huge job as there is a cross brace across the top of the tank so it's exhaust of, subframe out and tank out, two days work compared to half an hour on the tin top! Then it broke down again yesterday, AA brought it in. I have looked and the controller is sitting in its own little puddle with the surrounding area dry and dusty, like ***! The plug terminals are rotten and the controller dead again. I have drilled a hole in the holder and I am going to cover the access panel with polyther and soak the new pins and controller with terminal protector.