RC Low-pass filter for O2 sensor

Thread Starter

john2k

Joined Nov 14, 2019
85
OK so I placed the extra resistor across the lines just like the drawing and just like how I tested indoors with the battery. It is still reading 1.27V that makes no sense. So I disconnected the connector on the cars connector that goes to the ECU and with the car running and the O2 sensor completely disconnected I measured the blue and white wires using a DMM and from what I understand the voltage reading is supposed to read directly from the O2 sensor as it heats up from the cats. I was getting a very low reading but then when I plug the connector back on to the car connector then the voltage reading on the DMM shows as 1.27V. so this high voltage only reads when I connect it to the car. So disconnected the connector again and measured the two pins on the ECU plug and no reading. When I connect the connector up again then I get 1.27V again. What could be wrong?
 

Reloadron

Joined Jan 15, 2015
5,363
That makes absolutely no sense. The ECU in this case is not supposed to source voltage. A heated up O2 sensor generates a voltage proportional to the O2 it senses. That signal or actually voltage level is routed to the ECM for reading. Beyond that I have no idea how you can be seeing what you are seeing. It's like where the hell is it coming from? Weird? Open connection somewhere? There are ways that a signal could be injected but that's where things start getting a little complicated.

Ron
 

Thread Starter

john2k

Joined Nov 14, 2019
85
It makes absolutely no sense to me either. From what I understand the two black cables are supposed to be heating cables. But the white and blue cable is supposed to give a voltage reading back to the ECU based on the heat inside the catalytic converter which heats the sensor which then generates a voltage based on that temperature right? But where could this random voltage be coming from? Before I added the resistor and capacitor it wasn't doing this, do u think I should try and remove it completely and wire the two wires back normally and measure voltage and see? If it doesn't read the strange voltage then something with the RC filter is causing it right?
 

Reloadron

Joined Jan 15, 2015
5,363
Something like that. Once at temperature the O2 sensor just measures un burned oxygen in the exhaust. They generate a voltage proportional to how much un burned oxygen there is. Nothing to lose in taking it back to the beginning as something just isn't right.

Ron
 

MisterBill2

Joined Jan 23, 2018
5,802
Something like that. Once at temperature the O2 sensor just measures un burned oxygen in the exhaust. They generate a voltage proportional to how much un burned oxygen there is. Nothing to lose in taking it back to the beginning as something just isn't right.

Ron
Something like that. Once at temperature the O2 sensor just measures un burned oxygen in the exhaust. They generate a voltage proportional to how much un burned oxygen there is. Nothing to lose in taking it back to the beginning as something just isn't right.

Ron
Are you certain that the oxygen sensor is not a temperature dependent resistor, or semiconductor, and that the ECU is not supplying a bias to it? That would explain the 1.27 volts. This is the explanation that makes sense to me.
 

Thread Starter

john2k

Joined Nov 14, 2019
85
Something like that. Once at temperature the O2 sensor just measures un burned oxygen in the exhaust. They generate a voltage proportional to how much un burned oxygen there is. Nothing to lose in taking it back to the beginning as something just isn't right.
OK, so on Saturday i think what I will try is completely taking out the RC filter and wiring the white and blue wires back together directly and measure what the voltage reads. I was just thinking to test if I can actually send a specific voltage to the ECU and read it with the scan tool. So what I was thinking is maybe get a AA battery with resistors to make the output equal half the voltage of the AA battery and then connect that output to the ECU end of the connector seeing as I am going to cut the two wires again anyway and then see if the ECU reads the voltage that the AA battery is sending? If that works, is there not a way to make a device to just constantly send the correct waveform voltage to the ECU and just cut the two wires from the O2 sensor off and cap it off? What does the heater wires do for the O2 sensor? Below is a section from the wiring diagram for my car that shows the 4 wires going to the O2 sensor, I am working with the Bank 1 sensor 2.



Are you certain that the oxygen sensor is not a temperature dependent resistor, or semiconductor, and that the ECU is not supplying a bias to it? That would explain the 1.27 volts. This is the explanation that makes sense to me.
How would I determine this? The oxygen sensor has two black wires which are some sort of heating wires I am told but I don't know much more. Diagram posted above. Can you kindly explain how I can test this theory? Thanks
 

Reloadron

Joined Jan 15, 2015
5,363
This is a good read on O2 sensors. Like so many other sensors they vary by manufacturer. My wife has a sweet little 99 Tahoe which uses 4 sensors with her V8. Two upstream and two downstream of the catalytic converters. My Denali with a 6.0 Liter V8 uses two downstream and that's all. Some are heated using an element and some are not and rely on EGT (Exhaust Gas Temp) for heating. There is a delay before the ECM starts reading. That allows either type to get warm.

Ron
 

Thread Starter

john2k

Joined Nov 14, 2019
85
Thanks Ron, i'll have a read of that. What do you think of my using a AA battery with a resistor to bring the battery voltage down to half and them try putting the battery output to the ECU and see what the ECU sees? If it sees what the battery is measuring with a multimeter then it means the direction of the voltage measuring is correct and that something is going wrong with the RC filter etc. I will remove the RC filter and do some tests on Saturday. Is there not a way to just make some sort of device that can give a constant correct waveform voltage output in the form of a + and - outut which I can then plug in to the ECU connector and disconnect the two going from the O2 sensor?
 

Thread Starter

john2k

Joined Nov 14, 2019
85
I would try it. I doubt you will hurt the ECU.
I'll give that a try too. Because if the O2 sensor is supposed to produce a voltage between 0.1 and 0.9 and then send that as signal and ground to the ECU, then surely sending a voltage created by a AA battery reduced to under 1V should essentially give the ECU a reading as if it was getting a reading from the O2 sensor. Ofcourse the AA battery will just be a fixed voltage and wont have a waveform so other issues will be detected if it works but at least then I can carry on trying to figure it out.
 

MisterBill2

Joined Jan 23, 2018
5,802
OK, so on Saturday i think what I will try is completely taking out the RC filter and wiring the white and blue wires back together directly and measure what the voltage reads. I was just thinking to test if I can actually send a specific voltage to the ECU and read it with the scan tool. So what I was thinking is maybe get a AA battery with resistors to make the output equal half the voltage of the AA battery and then connect that output to the ECU end of the connector seeing as I am going to cut the two wires again anyway and then see if the ECU reads the voltage that the AA battery is sending? If that works, is there not a way to make a device to just constantly send the correct waveform voltage to the ECU and just cut the two wires from the O2 sensor off and cap it off? What does the heater wires do for the O2 sensor? Below is a section from the wiring diagram for my car that shows the 4 wires going to the O2 sensor, I am working with the Bank 1 sensor 2.





How would I determine this? The oxygen sensor has two black wires which are some sort of heating wires I am told but I don't know much more. Diagram posted above. Can you kindly explain how I can test this theory? Thanks
If the oxy sensor is a temperature dependent resistor then the resistance between the black and white wires, pins #29 and #30 will change when it is hot. But if a voltage appears between those wires while they are disconnected from the ECU, then it is not a resistor. And you may have done that check already. One more thing, based on that drawing, is are you sure the polarity is correct? The black to white looks funny to me.
 

Thread Starter

john2k

Joined Nov 14, 2019
85
I haven't had a chance to try the things I suggested I would try yet as we've been in lock down because of what is happening with the Covid19. Once we get back to a bit of normality, I will try this and update you. In the meantime, everyone stay safe.
 

atferrari

Joined Jan 6, 2004
3,930
I haven't had a chance to try the things I suggested I would try yet as we've been in lock down because of what is happening with the Covid19. Once we get back to a bit of normality, I will try this and update you. In the meantime, everyone stay safe.
Sorry but, could you revise why you started considering to add a filter? Like it or not that is a modification of the original design. Am I right?
 

Elfather

Joined Apr 25, 2020
1
There is no need for a diode. The design is a simple LP (Low Pass) filter. While it will only pass lower frequencies and filter the spikes what the ECU sees will be less than the O2 sensor sends, that may be a downside waiting for 5 RC Time Constants. I have seen it done using capacitor values from 1.0 uF to the 470 uF shown in your drawing. So in this case:

"The time also called the transient response, required for the capacitor to fully charge is equivalent to about 5 time constants or 5T. This transient response time T, is measured in terms of τ = R x C, in seconds, where R is the value of the resistor in ohms and C is the value of the capacitor in Farads".

In this case 470 seconds for one time constant now we get 470 * 5 = 2350 Seconds or about 39.2 min. That's a bit long. You may want to try some lower value capacitors like 10 uF and see how your results actually look. Some of this also depends on the O2 sensor in question too.

Ron
I'm planning on adding a 1Mohm resistor and a 4.7uF capacitor like the picture below on a o2 sensor that generates a voltage from 0 to 0.9v. The plan is to remove highly frequent voltage changes and allow the lower slow voltage changes to happen. Basically frequently changing voltage reports ineffecient catalytic converter so with the resistor and the capacitor apparently this allows the slower less frequent changes but smoothens out the faster frequent changes. This diagram is a tried and tested one based on youtube and the person who did it has had it over a year. My only question is, is there a need to put maybe a diode across from the + and - just like the capacitor?

Hi

I have a question I have a honda accord 2004 2.4l Manual, I have the code P0141 I have put several sensors and it does not work and the line is fine but I do not have the catalyst. I can do those steps so that the witness goes off?
 

Reloadron

Joined Jan 15, 2015
5,363
Greetings. You just want to make sure of how that make and model actually work. All makes and models are not the same and don't always return the same signal(s). My wife has a classic little 99 Tahoe which uses 4 sensors, two upstream and two downstream of the catalytic converter. All of the harness wiring to the engine control module is important too and some of it is exposed to some harsh conditions. Dome O2 sensors include heaters and some don't just using the hot exhaust gas temperatures.

Ron
 
Top