Vehicle details: Rover 45
Engine code: K16
Year: 2004
Symptom: MIL on but no symptoms,
Author: Rob Smith | Doncaster, South Yorkshire, UK

Rover 45 | Rover 45 with MIL on, P0170 Fault Code


This car came with the MIL (Malfunction Indicator Lamp) illuminated but no reports of any running issues.

The first action was to read the PCM (powertrain control module) for the stored code/s. There was only one stored:

Diagnostic trouble codes (DTCs)

P0170 - Fuel system adaptions out of range

The car was then road tested and live data via a scan tool was watched for any anomalies, one thing that stuck out was that the up-stream O2 sensor, with my foot hard to the floor on the throttle pedal, it was hitting a clean 0 V, on return from road test I put a test plan together.

I started by checking the fuel pressure, which was on spec, one thing I noticed at this point was it was still on its original fuel filter. The car was on its second owner from new and had low mileage for a 2004, so I replaced it as a matter of course.

Re-road testing after the filter change made no difference to the up-stream O2 sensor 0 V reading, I contemplated checking fuel flow at this point but decided to hook the PicoScope up and see what was occurring with the O2 signal. The reasoning for not checking fuel flow was the car never hesitated or exhibited any lack of fuel symptoms.
This was the first capture I took which shows the O2 sensor behaving normally.

After this capture I then carried out a test forcing the O2 sensor lean and rich, this is when the fault showed up.

As you can see in the above trace, Channel A (the O2 sensor) clearly drops to sub-zero readings. It was clear now that I was looking at CSD (Characteristic Shift Down).

Characteristic shift down

An oxygen sensor compares the residual oxygen content to an ambient air reference; it does this through the sensors wires. If the sensor ceramic cracks and the exhaust gas seeps through the crack the reference side becomes contaminated. The oxygen sensor then will start to work in reverse because the reference side will have less oxygen than the exhaust side therefore generating voltage in the opposite direction. This is seen as a negative voltage. The sensor can still switch as the average voltage approaches zero and continue negative. This symptom can clear up of its own accord, but if the ceramic is cracked the Characteristic Shift Down will no doubt return.


A new O2 sensor was fitted; the car was road tested and the O2 sensor behaved as normal hitting around 900 mV with the throttle on the floor.


Another case where true real time data is seen by the PicoScope compared to live data from a scan tool, as simple as this study may read and the outcome turned out, the PicoScope really gives diagnostics another dimension and I wonder without one what the next step might have been ?


8 comments | Add comment

April 01 2013

@Kim Anderson
Thanks for the comment.

I never looked at the fuel trims as i trusted the DTC to be telling the truth, but had i looked i would have expected the fuel trim to be maxed out as in the description of the DTC ‘Fuel system adaptions out of range’.

@Scott Thomas
Thanks for the comment.

To force a lean/rich condition add either air or propane etc into the manifold.

Thanks to all others too for your comments.

Scott Thomas
March 29 2013

Great article showing the power of using a Pico-scope and not relying just on live data. Just wondering Rob, how did you go about forcing the O2 sensor signal into the full rich / full lean modes? Regards Scott

March 29 2013

Good report, thankyou.  I learned something new today.  I was not familiar with the cracked ceramic substrate although I have witnessed negative voltages.

carlos alberto moraes
March 29 2013

Este codigo p0170 aparece muito no brasil nos carros linha citroen picasso 2.0 sempre relacionados a sonda lambda. MORAES

Kim Andersen
March 28 2013

Good job and your approach to this problem - seems straightforward.

Do you had a look at the fuel trims as well - if they were out of range. I’m asking because it would be nice know - if this type of error would be seen in the fuel trims.

washington davy masunga
March 28 2013

This was well executed,I do encourage today’s auto electricians to do a basic instrumentation and control course to understand sensor operating characteristics at varying temperatures and to interpret scantool or scope readings etc

March 28 2013

Excelente aporte, muchas gracias

Dave Hill
March 28 2013

Nice work Rob.

A great example of the need for scrutiny of the “actual” sensor values, rather than total reliance on the scantool’s data.

I am always suspicious when I see ZERO volts on a Zirconium Dioxide O2 sensor. seen on a scan tool screen & this will prompt me to follow up with a Pico.

I have to thank James Dillon for teaching me about the cracked reference chamber as the cause of this fault.

Keep up the good work Rob!

Your email address will not be published. Required fields are marked *

Case study: Rover 45 with MIL on, P0170 Fault Code