Titania lambda : reason for high output?
Discussion
I've got a car in that's puzzling me somewhat.
It's on a cerbera, but seeing as the majority of knowledge is here I have put it in this section, and the lambdas are shared with other tvr models.
The issue is that the lambda sensors are outputting a range of 0 to around 1.8 volts. This is enough to send the ecu into a fault mode suspending lambdas control.
The previous garage put a new pair of lambdas in and I have done the same and it's the same story so not a lambda sensor issue (genuine NTK lambdas)
Both lambdas are affected equally so it's a global issue not just one lambda.
So far i have tried the new lambda sensors, and another ecu (easy to swap out ) .. same issue.
I have wired the lambda signals direct to the ecu .. same issue.
I have put the ecu grounds to the battery rather than the chassis earth point, and also put them to the engine block .. same problem.
Lambda heaters are getting battery voltage (around 13.6/13.7 volts engine running / alternator charging)
What could be causing the high voltage output? It's as if the lambdas are sensing a rich mixture, and then some more. But they switch back to zero with a good spiky nature like a good healthy lambda should, and the mixture is indeed being clamped to lambda=1.
It *seems* to be worse if the engine is getting up to operating temp, ie during the warm up stage where the lambdas are controlling the fuelling the peak output seems to be around 1.2 volts .. but you can watch this peak voltage get greater as the engine then heats up more. Not sure if that's significant .. other than lambda tip temp will increase with hot exhaust gases. But that still doesn't explain the behaviour.
any ideas folks?
It's on a cerbera, but seeing as the majority of knowledge is here I have put it in this section, and the lambdas are shared with other tvr models.
The issue is that the lambda sensors are outputting a range of 0 to around 1.8 volts. This is enough to send the ecu into a fault mode suspending lambdas control.
The previous garage put a new pair of lambdas in and I have done the same and it's the same story so not a lambda sensor issue (genuine NTK lambdas)
Both lambdas are affected equally so it's a global issue not just one lambda.
So far i have tried the new lambda sensors, and another ecu (easy to swap out ) .. same issue.
I have wired the lambda signals direct to the ecu .. same issue.
I have put the ecu grounds to the battery rather than the chassis earth point, and also put them to the engine block .. same problem.
Lambda heaters are getting battery voltage (around 13.6/13.7 volts engine running / alternator charging)
What could be causing the high voltage output? It's as if the lambdas are sensing a rich mixture, and then some more. But they switch back to zero with a good spiky nature like a good healthy lambda should, and the mixture is indeed being clamped to lambda=1.
It *seems* to be worse if the engine is getting up to operating temp, ie during the warm up stage where the lambdas are controlling the fuelling the peak output seems to be around 1.2 volts .. but you can watch this peak voltage get greater as the engine then heats up more. Not sure if that's significant .. other than lambda tip temp will increase with hot exhaust gases. But that still doesn't explain the behaviour.
any ideas folks?
It's the same titania sensor used in the rover V8 engines that's used in the cerbera V8 so yes the sensor is the correct one.
There's a modified version for the SP6 cerbs and early tuscans to cope with the heat of the sp6 install but Ive never had an issue using the standard NTK sensor on the V8 version.
There's a modified version for the SP6 cerbs and early tuscans to cope with the heat of the sp6 install but Ive never had an issue using the standard NTK sensor on the V8 version.
Given that we do sometimes have issues with new parts (I am thinking plug extenders and all manner of electrical components), have you tried a used set out of another car, ones you know work fine in that car. Or have you tried putting these new ones into a different car, for example your own Cerbera?
If the catalytic converter were faulty, what would the lambda voltages be?
If the catalytic converter were faulty, what would the lambda voltages be?
Edited by QBee on Tuesday 21st July 16:48
I've never seen one go that high, 1.4 tops for a highly rich mixture. I've seen shorts in the loom dump 12v on the ECU input (14CUX ) and it survived with just an error code. If the device is not faulty Id be looking at a current feed on a ground plane between the sensor and ECU that would causing the sensor ground to be at a higher voltage than the ECU ground, but I suspect Im teaching you to suck eggs.
QBee said:
Given that we do sometimes have issues with new parts (I am thinking plug extenders and all manner of electrical components), have you tried a used set out of another car, ones you know work fine in that car. Or have you tried putting these new ones into a different car, for example your own Cerbera?
blaze_away said:
Just an idea, take the cerb's lambda's and stick em on a known good running Chim or Griff RV8 and see what voltages you get on there.
Sadly the only other car I have in right now is a SEAC for an emerald ecu and that is a non-runner and doesn't have any lambdas anyway. I could run the lambdas in my own cerb but .. the dead SEAC is on my only usable ramp and I'm having to do the cerbera lambdas under my rolling road ramps (they are accessed from underneath) so swapping anything into another car is a bit of a non starter sadly.LimSlip said:
Titanium O2 sensors do not output a voltage, they change resistance according to O2 concentration. The maximum voltage across the sensor will depend on the reference current that the ECU provides.
I thought the heater circuit supplies power to the lambdas, and the conductivity of the sensor element (which varies with mixture ) then sends a small voltage out to the ecu. This is typically in the 0 - 1.2volt range but the output from 2 sets of lambdas is 0 - 1.8 volts and flagging a lambda fault in the ecu.David Beer said:
Cheap advert again joolz, surely you are the expert !
Not sure what you mean about advert, unless advertising that you are baffled by something counts as a positive. At least at the moment I don't look any worse than the previous garage, so that's something. I offer no fix no fee, so I won't be profitting from my incapability.Penelope Stopit said:
Was that voltage measured with the fault showing?
to be honest I don't know. I can't remember seeing the diagnostic screen showing anything other than normal running voltages, but I can data log the ecu later today and see if that shows up anything.blitzracing said:
I've never seen one go that high, 1.4 tops for a highly rich mixture. I've seen shorts in the loom dump 12v on the ECU input (14CUX ) and it survived with just an error code. If the device is not faulty Id be looking at a current feed on a ground plane between the sensor and ECU that would causing the sensor ground to be at a higher voltage than the ECU ground, but I suspect Im teaching you to suck eggs.
Hi Mark .. any input greatfully received right now. I'm guessing the ecu inputs on the lucas and mbe and others are well protected, they have to account for things like wires chafing through and 12v going straight into the lambda wires.cinquecento said:
Is it possible to take the belt off the altenator..or remove its 12v output. Maybe a leaking diode?
The alternator appears to be fine at the moment but I will data log the voltages at the ecu and see if anything shows up. Sadly I have no way of capturing instantaneous voltage spikes or similar voltages elsewhere as all I have is a digital voltmeter and that has a small lag / slowish sample rate.If I find anything else out today I will let you know.
Edited by spitfire4v8 on Wednesday 22 July 09:36
Edited by spitfire4v8 on Wednesday 22 July 09:54
spitfire4v8 said:
LimSlip said:
Titanium O2 sensors do not output a voltage, they change resistance according to O2 concentration. The maximum voltage across the sensor will depend on the reference current that the ECU provides.
I thought the heater circuit supplies power to the lambdas, and the conductivity of the sensor element (which varies with mixture ) then sends a small voltage out to the ecu. This is typically in the 0 - 1.2volt range but the output from 2 sets of lambdas is 0 - 1.8 volts and flagging a lambda fault in the ecu.
On the subject of alternator output, the voltage after startup rose to around 14.2 volts and then settled at around 13.8/13.9 occasionally going up to 14.1 for a moment ..
I had 10mins on the phone with NTK's technical department who couldn't offer a definitive answer other than to say it's a rich mixture and to look elsewhere on the car for the issue rather than at the sensor, which is fair enough given that 4 new and 2 good used sensors have all given the same outcome so far ...
I had 10mins on the phone with NTK's technical department who couldn't offer a definitive answer other than to say it's a rich mixture and to look elsewhere on the car for the issue rather than at the sensor, which is fair enough given that 4 new and 2 good used sensors have all given the same outcome so far ...
Hi Steve .. it's not that it's a permanently rich condition per-se .. the lambdas are cycling back to zero volts and spiking back up which shows the ecu can trim the fuelling back weaker and then richer in its natural cyclical routine. What's happening is that during the natural rich reading spike the lambda output is going over the normal 1.2volts you would expect to see and heading to 1.7volts which flags an ecu fault.
I can disguise the symptom by changing the lambda sensor limits in the ecu so that the high voltage doesn't flag an error, but that's not fixing the fault, which might get worse and my short-term fudge would be useless.
I can disguise the symptom by changing the lambda sensor limits in the ecu so that the high voltage doesn't flag an error, but that's not fixing the fault, which might get worse and my short-term fudge would be useless.
spitfire4v8 said:
Penelope Stopit said:
Was that voltage measured with the fault showing?
to be honest I don't know. I can't remember seeing the diagnostic screen showing anything other than normal running voltages, but I can data log the ecu later today and see if that shows up anything.As you mention, 14.1 volts down to 13.9 is good......if it stays at that or reduces when hot
I really was clutching at straws
What a pain of a fault
Gassing Station | General TVR Stuff & Gossip | Top of Page | What's New | My Stuff