can you test Lambda sensors ?
Discussion
My renault megane is playing up, and been a carb and dizzy man this electronice sensor stuff is all new ! i understand the principles of what the sensors do, but as the car is running poorly and this morning cut out altogether i wondered how you go about checking if the sensors are working properly. I dont have home diagnostics so is there any other way to avoid just changing sensors and hoping to get lucky, or is it a trip to the dealer time do you reckon ?
I didn't write this but,
Testing Lambda Sensors
Lambda sensors can be tested either on or off the car. Both methods have their advantages. Testing the sensor off the car enables you to check the sensor itself without interference from other components. Testing the sensor on the car shows you what is going on in the real-life situation.
Before testing the sensor on the car you must make sure that the engine is thoroughly warm (leaving it to idle won’t do – take it for at least a 10 minute drive) and that it has no misfires or leaks in the exhaust system between the engine and the lambda sensor.
Warning
For both tests you need a good quality digital multimeter with a high impedance (at least 1,000,000 ohms/volt – most digital multimeters meet this requirement). Do not use an analogue meter – they will not give accurate results and can damage the sensor. Many digital multimeters don’t react quickly enough to the rapid changes in output from the sensor and so I use a simple home-made tester which has LEDs to indicate voltages between 0 and 1.2 volts. (Separately I will put up details of how to build this).
On Car Test
The lambda sensor used on the your car has 3 wires – two white heater wires and one black signal wire. Generally it is best to tap into the wires using wire piercing test probes which makes a tiny hole in the insulation. I make the connection at the car side of the plug where the leads from the sensor join the main loom because the wires attached to the probe are sheathed in a very hard heat proof insulating material which is very difficult to puncture. If when you switch on the ignition you get a reading of 12+ volts you have tapped into the heater power lead by mistake.
For this test you must leave the sensor connected to the ECU. Connect the positive multimeter probe to the sensor output wire and the negative probe to earth. With the engine running if all is well the voltage shown on the meter should flicker up and down the scale changing at least twice a second between about 0.1 and 0.8 volts -- the average reading should be about 0.4-0.5 volts but it should never be steady at this voltage. Flick the throttle open and the voltage should flicker, rise for a second or so then fall again. This indicates the ECU is receiving input from the sensor and is responding properly.
A steady voltage of between 0.4 and 0.5 V is a sign that the ECU is running in open loop mode . Either the engine is not fully warm or it may indicate a bad connection or a faulty water temperature sensor or an ECU fault. It could also indicate a problem with the sensor heater so check the heater is getting power and is properly earthed.
Workbench Testing
Use a good digital multimeter. Clamp the sensor in a vice or use pliers or vice-grip to hold it. Clamp the negative multimeter lead to the case and the positive to the output wire. Use a propane torch and apply the inner blue flame tip to heat the fluted or perforated part the sensor. You should see a DC voltage of at least 0.6 within 20 seconds. If not, the most likely cause is open circuit internally or lead fouling. If OK so far, remove from flame. You should see a drop to under 0.1 volt within 4 seconds. If not the sensor is probably silicone fouled (coolant leak?). If still OK, heat for two full minutes and watch for any drop in output voltage. Sometimes, the internal connections will open up under heat.
Testing Lambda Sensors
Lambda sensors can be tested either on or off the car. Both methods have their advantages. Testing the sensor off the car enables you to check the sensor itself without interference from other components. Testing the sensor on the car shows you what is going on in the real-life situation.
Before testing the sensor on the car you must make sure that the engine is thoroughly warm (leaving it to idle won’t do – take it for at least a 10 minute drive) and that it has no misfires or leaks in the exhaust system between the engine and the lambda sensor.
Warning
For both tests you need a good quality digital multimeter with a high impedance (at least 1,000,000 ohms/volt – most digital multimeters meet this requirement). Do not use an analogue meter – they will not give accurate results and can damage the sensor. Many digital multimeters don’t react quickly enough to the rapid changes in output from the sensor and so I use a simple home-made tester which has LEDs to indicate voltages between 0 and 1.2 volts. (Separately I will put up details of how to build this).
On Car Test
The lambda sensor used on the your car has 3 wires – two white heater wires and one black signal wire. Generally it is best to tap into the wires using wire piercing test probes which makes a tiny hole in the insulation. I make the connection at the car side of the plug where the leads from the sensor join the main loom because the wires attached to the probe are sheathed in a very hard heat proof insulating material which is very difficult to puncture. If when you switch on the ignition you get a reading of 12+ volts you have tapped into the heater power lead by mistake.
For this test you must leave the sensor connected to the ECU. Connect the positive multimeter probe to the sensor output wire and the negative probe to earth. With the engine running if all is well the voltage shown on the meter should flicker up and down the scale changing at least twice a second between about 0.1 and 0.8 volts -- the average reading should be about 0.4-0.5 volts but it should never be steady at this voltage. Flick the throttle open and the voltage should flicker, rise for a second or so then fall again. This indicates the ECU is receiving input from the sensor and is responding properly.
A steady voltage of between 0.4 and 0.5 V is a sign that the ECU is running in open loop mode . Either the engine is not fully warm or it may indicate a bad connection or a faulty water temperature sensor or an ECU fault. It could also indicate a problem with the sensor heater so check the heater is getting power and is properly earthed.
Workbench Testing
Use a good digital multimeter. Clamp the sensor in a vice or use pliers or vice-grip to hold it. Clamp the negative multimeter lead to the case and the positive to the output wire. Use a propane torch and apply the inner blue flame tip to heat the fluted or perforated part the sensor. You should see a DC voltage of at least 0.6 within 20 seconds. If not, the most likely cause is open circuit internally or lead fouling. If OK so far, remove from flame. You should see a drop to under 0.1 volt within 4 seconds. If not the sensor is probably silicone fouled (coolant leak?). If still OK, heat for two full minutes and watch for any drop in output voltage. Sometimes, the internal connections will open up under heat.
My only additions/corrections to the above are as follows.
Most sensors now have 4 wires & you will need to use the black & grey if it is a 4 wire sensor.
VERY occasionally you even get 2 wire sensors, in this case use the black & chassis earth.
There are other oddities, but 2 3 or 4 with black grey & white in the mentioned combinations are the most common.
Secondly, in the on-vehicle test, if your meter isn't fast enough you may see 0.45 as a constant reading, you will only see (VERY QUICK) changes in voltage if it is quick enough.
Most sensors now have 4 wires & you will need to use the black & grey if it is a 4 wire sensor.
VERY occasionally you even get 2 wire sensors, in this case use the black & chassis earth.
There are other oddities, but 2 3 or 4 with black grey & white in the mentioned combinations are the most common.
Secondly, in the on-vehicle test, if your meter isn't fast enough you may see 0.45 as a constant reading, you will only see (VERY QUICK) changes in voltage if it is quick enough.
Also even a good volt meter has a 4Hz refresh rate ie it refreshes 4 times a second. A lambda needs to switch from 0.2 to 0.8volts at least once a second more is better. A volt meter is not really an ideal tool for the job. A scope is what you need but most people will not have access to that and a volt meter may give you miss leading information.
You are absolutely right that a scope is the tool for the job, alas 99% of us don't have one & the 2nd poster above was talking about using a DVM to test so I was commenting on his post with a bit more information to hopefully help if that is OK?
On top of that as you yourself say the refresh rate of a good DVM can be 4Hz or occasionally better, whereas the Lambda sensor update rate can be down to 1hZ.
As 4 is much faster than 1 then yes it will suffice in the absence of a scope!
Yes a DVM CAN give misleading information, but then so can a poorly set up scope.
On top of that as you yourself say the refresh rate of a good DVM can be 4Hz or occasionally better, whereas the Lambda sensor update rate can be down to 1hZ.
As 4 is much faster than 1 then yes it will suffice in the absence of a scope!
Yes a DVM CAN give misleading information, but then so can a poorly set up scope.
Understanding the shortcomings of the tools may stop someone playing parts darts because they think the tool is giving good information. A good lambda can switch twice a second so a multi meter could take the first reading at say 0.5 volts on the rise it's second reading on the fall will be 0.5 volts it's third reading on the second rise will again be at 0.5v and the fourth reading on the fall of the second switching point at 0.5 volts. The multi meter will show this as a buggered sensor as it will look like the sensor is stuck around 0.5volts. I'm not saying don't use a multi meter if that all you have just think before you do and understand how it works and don't throw away a part that is working very well.
Understanding how a sensor works as well will also help you stop changing parts that don't need to be replace. I have seen more fault codes for lambda sensors that are caused by environmental reasons than a faulty sensor.
Oh and by the way a if the scope is set up poorly and not giving you good readings to test a lambda sensor thats not the fault of the tool is it now? Tool as in scope not operator
Understanding how a sensor works as well will also help you stop changing parts that don't need to be replace. I have seen more fault codes for lambda sensors that are caused by environmental reasons than a faulty sensor.
Oh and by the way a if the scope is set up poorly and not giving you good readings to test a lambda sensor thats not the fault of the tool is it now? Tool as in scope not operator
Gassing Station | Home Mechanics | Top of Page | What's New | My Stuff