Rover Gauge question

[Plan B]

After a bit of fiddling around managed to get RG working (thanks Mark!!). RG helped diagnose the issue that was making the engine cut out at idle and being very lumpy at low revs (turned out the TPS was permanently showing 100% even at idle and on cleaning the harness connector to the TPS problem was resolved). However I still have an unanswered question. All seems well on the RG display except before the engine gets to operating temperature the short term lambda trim is off the scale showing +100% on the RH and hovering around +/- 5% on the LH. This is then expectantly flagging a RH lambda error. When the engine gets up to temperature all appears to normalise - the short term trim then goes to zero on both sensors. and stays that way. (The long term trim is always showing zero for both sensors.)
Why would the RH lambda report such a difference from the LH when the fuelling to both banks should be about the same?

[Plan B]

Bump! Any suggestions?

[Plan B]

Bump! Any suggestions?

[bobfather]

Process of elimination, swap the lambda sensors over, if the fault switches side then it's a faulty sensor

[chris52]

I'm no Rovergauge expert but have been using it recently to try and find a fault in mine and found that if you disconnect a lambda the reading on RG for that Lambda goes full +100% so I would suggest that maybe you have a faulty lambda or faulty wiring to the lambda. Also when the car gets hot they should be sweeping back and forth from 0 but will eventually go static if left idling for a while. So maybe your car is going into limp mode when your engine is warm and cutting off the Lambdas. This is only a theory from my limited experience of RG
Chris

[Plan B]

When the engine is warming up the LH lambda is swinging between -10% and -5%. When up to full temperature the LH moves to around zero and hardly moves from there. The RH initially goes +100% and flags an error but when up to temperature settles back to zero. I'd be surprised if this was switching into limp. RG is still showing fuelmap 5 and power when driving seems fine.
I'll try switching LH/RH sensors as suggested. If I remember correctly to do that I need a specially adapted (butchered!!) socket.......

[bobfather]

Yes, lambda tool is an unusual size, can't recall the AF but a long socket chopped so that the cable can slot through

ETA 17mm AF

Edited by bobfather on Sunday 16th April 18:05

[Plan B]

I'm. being a lazy boy - Just ordered a Draper lambda socket thru Amazon Prime. Delivered tomorrow.

[bobfather]

Plan B said:

I'm. being a lazy boy - Just ordered a Draper lambda socket thru Amazon Prime. Delivered tomorrow.

I thought the Draper sockets were all 22mm. This one needs 17mm AF

[Steve_D]

If the readings even out when warm then it is likely a lazy lambda. Some start to give slow switching prior to failing completely.

If you truly have found a lambda socket of the right size then let us know as I want one. My modified socket only just works as there is not enough depth inside and the hex only just engages.

Steve

[Plan B]

Just checked and yes it is 22mm. Oh sh**. Have now ordered 17mm deep socket and will unleash angle grinder on it.

[ClassicChimaera]

Steve_D said:

If the readings even out when warm then it is likely a lazy lambda. Some start to give slow switching prior to failing completely.

If you truly have found a lambda socket of the right size then let us know as I want one. My modified socket only just works as there is not enough depth inside and the hex only just engages.

Steve

I used a fairly cheap socket and had to drill it out slightly to get a decent bite on a tight lambda.
Grinding a groove weakens the socket badly, I just clamp some mole grips on it. JD

[Englishman]

Rather than physically changing the lambda's over, why not just swap the connectors just to see if the fault moves or not first?

[bobfather]

Englishman said:

Rather than physically changing the lambda's over, why not just swap the connectors just to see if the fault moves or not first?

That would prove a cable fault but not a sender fault.

[Englishman]

bobfather said:

Englishman said:

Rather than physically changing the lambda's over, why not just swap the connectors just to see if the fault moves or not first?

That would prove a cable fault but not a sender fault.

Maybe I'm missing something, but surely if the fault moves to the other bank, as seen by RoverGauge, it would show either a cable fault or a lambda fault. If nothing changes then the lambda is most likely fine and the fault is further up the chain?

[bobfather]

Englishman said:

Maybe I'm missing something, but surely if the fault moves to the other bank, as seen by RoverGauge, it would show either a cable fault or a lambda fault. If nothing changes then the lambda is most likely fine and the fault is further up the chain?

Trying to determine whether the fault is genuine fuelling issue or a false indication caused by a faulty sensor. You must move one from the other. Swapping the cables won't identify one from the other. After swapping the sensors If the fault swaps sides it's a sensor fault, if it doesn't swap it's a fuelling or ignition fault

Edited by bobfather on Sunday 16th April 20:19

[Englishman]

Swapping the connectors does the same test in trying to identify if one of the lambda's is faulty, just as swapping them over does as you suggested above - just saves the effort in physically swapping them over.

[bobfather]

Englishman said:

Swapping the connectors does the same test in trying to identify if one of the lambda's is faulty, just as swapping them over does as you suggested above - just saves the effort in physically swapping them over.

Swap the cables and the fault swaps sides, what does that prove? If there's a genuine fuelling issue the fault will swap sides, if it's a faulty sensor the fault will swap sides. Swapping the cables gives the same outcome for both scenarios so swapping the cables won't point to either fuelling or sensor. How will swapping the cables prove whether it's fuelling or sensor. If you only swap the cables then you'll only show cable or ECU errors. You can't identify differences between fuelling and sensor if you keep them paired, splitting them allows you to follow and identify the error location

Edited by bobfather on Monday 17th April 09:21

[stevesprint]

James

Swapping just the lambda plugs over will confuse the ECU as it will try correcting the fuelling for the wrong bank and cause havoc.

Completely swapping the probes & plugs over will prove if a probe is faulty becuase if faulty the fault will change banks, I know from experience.

Good Luck
Steve

[blitzracing]

If the switching is stopped at idle- just give the car a rev and check it restarts cycling again. I dont know for sure why this happens, but Ive seen it quite a bit. A lazy probe will still switch, just slowly- so this maybe what you are seeing at idle. If the probe has died totally - ie the output is zero, then the long term trim shifts bit by bit until it reaches 100% as the ECU tries to correct the mixture, but cant as there is no output. It will eventually throw an error code, but it takes some minutes. The ECU wont throw an error code if the probe switches at any point in this period, so it wont trip a fault code at all easily. You can always back up RoverGauge by popping a test meter across the probe output wires- black and white and see if you have any voltages- its in the range of 0 (lean) to 1.4 volts (Rich).

Good article here on the lambda cycling:

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