sti-spec-a

First noticed something was up when i switched on the ignition and the LambdaLink went straight to the top LED. I checked the output voltage of the O2 sensor on the PowerFC Commander and it was displaying 4.7volts and the engine was not even started. When i started the engine it slowly dropped to the normal operating range of 0.0v - 0.8v but after a short run along the motorway it slowly reduced to 0.0v constant on a light throttle. The voltage would not increase with any kind of increase on the accelarator so i presumed a dead O2 sensor.

I bought a new Bosch LS02 universal O2 sensor from the local motor factors for £60 and installed it within the hour. I switched on the ignition and the output voltage was 0.0v as expected and when i started the engine it slowly came up to opertaing temp wthin about 1 mile of leaving the house. In closed loop it would fluctuate between 0.0v and 0.7v, again as expected.

When giving the car a bit of load the top green on the lambda would stay solid but the max voltage on the Commander is showing a solid 0.74v. After a 20 mile drive to a friends to drop off parts i noticed that the LambdaLink was not showing closed loop any more so i checked the voltage on the Commander and it was reading 0.2 volts and dropping. After 10 mins of banging my head off the steering wheel it became apparent that yet another Bosch Universal sensor had "Gave up the Ghost" but this one had died after only 30 miles maximum, a cost of £2 a mile.

Pulled into the local Shell to fill it up and upon switching the ignition on it showed 3.5v on the Commander and top green on the LambdaLink

After some quality Pizza contemplation time i thought along the lines of the heat affecting the sensor in its current location, 95 STI sensor in headers, or that the 5v heater circuit of the sensor was somehow shorting over to the signal wire and giving the strange voltage readings.

Under the cover of darkness i proceeded to remove the new sensor and check the connections for any sign of shorting but all the connections seemed to be still tight and clean. I pulled the sensor out of the headers and checked it for signs of heat overload but it looked as fresh as did the day it was fitted, which was Tues evening. With the sensor dangling free from its mounting i switch on the ignition and it was showing 3.8v whilst in the atmosphere.

The only thing i could think of trying was to disconnect the 5v heater circuit to the sensor and take it for a run to see if it would produce any voltage. I took the car round the housing estate for a distance of about 300 yds and checked the output voltage on the Commander, 0.0v. It was stone dead so i turned round and headed for home but as i approached the entrance to my street i checked the voltage to see that it now reading 0.4v and slowly climbing. So i carried on towards my "Test road" to see how the voltage would react after i got some heat into it by means of the exhaust gas temp. After a distance of approx 2 miles the readings from the Commander were as expected for a normal sensor, engine temp was 65c and no boost had been used. Once the engine was upto opertating temp i gave it some beans and the voltage would show 0.86 on the Commander and top green on the LambdaLink.

To recap and summerise an extremely long winded post which took an hour to type due to my infamous 1 fingered technique

Brand new Bosch LS02 sensor lasted 500miles then died to 0.0volts
Another brand new Bosch LS02 sensor lasted 30 miles then died to 0.0volts and would only read max of 0.74v at Wide Open Throttle with the heater circuit connected
Bosch sensor now runs as expected readings, between 0.0v and 0.74v on closed loop and now 0.86 on WOT, on the Commander and Lambdalink with the heater circuit disconnected and the sensor operating as a single wire O2 sensor

Why are these brand new Bosch sensors only lasting a short time?
Why is the Bosch sensor now working as expected with the heater circuit disconnected?
Why was i only getting 0.74v on WOT with the heater circuit connected and now getting 0.86v at WOT with the circuit disconnected?
How much does the PowerFC rely on the O2 sensor for info input?
Can i run the car with only the signal wire connected, car is for weekend use only?
Why on the old sensor was my idle showing -490mm/hg on idle and with the new sensor it stablised to -505mm/hg?

I have an independent wideband monitor which is showing healthy AFR's through out the range
At idle the car is very smooth, feels better than std ECU, with very little or no vibration through the gearstick
Only slight negative i can feel or see is a slight undershoot on the idle when coming to a junction but nothing severe.

Grant

sti-spec-a

BTTT for any passing techies

I need to find out why it is doing it as i am not throwing new Lambda sensors at it every 30 miles

Grant

buzzard

are you certain the lambda sensors are in fact dead and not an intermittent prob with ecu/loom?

sti-spec-a

I have double checked the connections between the loom and sensor and everything looks nice and tidy

I was thinking along those lines because of the 5v readout which i believe is the same voltage as the heater circuit

I also cant fathom out why it showing 0.74v at WOT with the heater circuit connected and 0.86v at WOT when the circuit is disconnected

I am running the car just now with the heater circuit disconnected and on the std ECU it is showing no CEL fault so it must be seeing a signal which it is expecting once i get some heat into the sensor

I will get the local auto spark to check the heater circuit

The car is been final mapped on the 4th of Oct and i will ask Andy to switch off the O2 sensor

Grant