o2 sensor - low voltage warning and an odd trace - any ideas

[miless2111s]

I wonder if you guys can help me? We have been struggling with a MIL light and error code which will not go away on my Honda Elise. We've installed a new o2 (primary) sensor and wired it up as per the fitment in the Civic Type R however we sill get the erorr.

Looking at the data log trace (see here: http://miles-lotus-adventure.blogspot.c ... -work.html) you can see a drop a few minutes after start up where the voltage does indeed drop to near zero and then recovers. Then 30-40 seconds later it thows the error and MIL light.

We wondered if this could be a loose wire / poor connection however the timing is pretty consistent in terms of when it happens and how long the recovery takes. I would have expected a rather more random timing from a loose wire. We were also unable to provoke the same trace by waggling leads and fiddling with connections.

The sensor is a new Honda item.

I wonder if there is some interaction or other effect which could be causing this (as the mechanical aspects appear to be ruled out). Notice that the s.trim line behaves the same in each case - dropping when the initial drop to zero happens and then going high only to recover to normal when the MIL kicks in and I assume the circuit goes to open loop.

The trace after the MIL light also seem svery unstable and settles gradually to the "normal" level of around 0,25v on the o2 sensor.

Any ideas / help you can give will be gratefully received! :)

Thanks

Miles

[Spunkster]

Exact errror codes, the calibration and datalogs showing the problem would be helpful.

[miless2111s]

Exact errror codes, the calibration and datalogs showing the problem would be helpful.

Thanks for the tip :)
I don't know that the normal protocol is however I will need to talk to Romain at EuroSpec before I issue his calibration - it is his work after all :)
I will get the error code tomorrow however I believe it was "P0131 o2 sensor circuit low voltage" ETA: confirmed.
I have tried to attach the data logs which sit behind the graphs on the blog.
I hope this is enough to help with any suggestions - if you need to know anything else just let me know :)
Thanks
Miles

[Hondata]

Your o2 sensor wiring is incorrect. You can always email the calibration if you do not want it to be public.

[miless2111s]

Turns out that the lean-rich / rich-lean thresholds were set wrong.

Were:
Rich-to-lean: 1.19V
Lean-to-rich: 0.2V

New Values:
Rich-to-lean: 0.70V
Lean-to-rich: 0.29V

Excellent explanation from Arno as to what was going on:
- the ECU at the moment never 'sees' the correct upper voltage on the
labmda sensor so ti doesn't start closed-loop cycling based on the
initial feedback

- after some (probably fixed) time it probably decides that the sensor
may be slow and it starts to lean off the mixture

- this you see in the graph as the O2 voltage drops from the displayed
0.21V to 0.0alittlebit and the fuel trim is also dipping. This is
good and normal!

- now it thinks 'aha! it is working!', so it wants to get to the 'rich'
peak and it starts increasing the fuel trim to get to it's upper
voltage level

And this is where it gets tricky..

It adds more and more fuel and indeed you see the lambda
voltage go from 0.0alittlebit to a little over 0.21V again (oh..
and the fuel status in the datalog changes to 'closed loop')

So far so good! This is also 'as it should be'

But.. This is where it goes pear-shaped for the ECU as it compares the
voltage from the sensor to the upper limit configured, so it can start the
'lean out' cycle again.

But... As you can see from the graph the lambda sensor is maxed out at
the 0.21V, but the ECU is still not reaching it's upper voltage from the
sensor, so it keeps adding more and more and more and more fuel.. this
is the fuel trim that goes waaaaaaaay up.

In the end it hits the maximum allowable fuel trim (+46% to -27%, also
configurable), *still* has not reached it's sensor voltage upper
limit as configured and basically it then goes "f*ck this.. it's broken!" and
turns on the MIL and goes back to open-loop running.

Everything after this point is basically 'nonsense' when looking at the
lambda sensor data as it's in open loop anyway..

Hope that helps anyone else who has that s.trim ramp and MIL light :)

Miles