This one stung a little. It's a Jeep thing, you wouldn't understand
I'd like to share one that threw an unexpected speed bump in front of my Jeep journey. I'll finish with a question at the end so lets get the conversation rolling.
06 Jeep Liberty 3.7 Customer complaint Check engine light is on with no driveability symptoms. Vehicle runs and drives great.
Of course, we begin with a code scan Here
I notice right away that there are 2 codes related to the upstream oxygen sensor on Bank 2. I already have the Scan tool out, lets review some data for the oxygen sensors.
So I start the vehicle and let it run for a few minutes while analyzing the behavior of all 4 oxygen sensors. I notice right away that B2S1 is stuck high at approximately 4 volts. This voltage includes the Bias voltage Chrysler likes to include on their oxygen sensor signals. In scan data, for ease on the eyes and brain, I chose to graph the 0-1v data that the scan tool allows you to view as well. This simply eliminates the bias voltage from the equation so you're looking at a typical 0-1v oxygen sensor sweep. This is much easier to easily digest the data. In the scan data here you can see how the sensor is stuck very rich. Also, notice my downstream sensor on that same bank is NOT rich. This is evidence that there more than likely is not truly a rich condition happening in the exhaust stream. I deciphered this data as the downstream sensor being very lean due to the trim correction caused by the stuck rich UPSTREAM oxygen sensor.
OK, So we potentially have a sensor gone wonky right? ( l love "wonky" by the way)
Well hang on, I have a heater code as well. Lets check that out as well. I unplug the sensor and perform a very quick resistance check of the heater circuit of the sensor. It is indeed open. I take a step further to confirm the driver of this o2 heater is not damaged by verifying it can carry current by supplementing the load with a sealed beam headlight. Heater circuit is confirmed intact up to the open element in the o2.
OK, so we have confirmed the sensor is indeed bad.
Lets back up a bit. Can something else cause my o2 signal wire to read incorrectly?
I've done this next step which I am about to explain many times in the past with success. I know there are many people who feel the Power probe is not the best tool in the world. I have a Power probe hook. And I love it to death. I, However, know its limitations, pros and cons. A major PRO, in my opinion, is the V-ref mode the tool has.
With this V-ref mode, you can supply a circuit with a voltage, in 500mV incriments, from .5-5v This is VERY handy when attempting to prove the circuitry and PCM processing of an oxygen sensor.
So, I unplug the oxygen sensor, I find a wiring diagram to determine which wire is my signal wire back to the PCM and I supply the circuit with V-ref voltage starting at 500 mV WHILE viewing scan data. I start at 500, the scan tool says 1.6v. I move to 1v, the scan tool still says 1.6v. This scan data pid does not change until I reach above 1.6 volts.
Ok, In my head, I assumed I needed to exceed the bias voltage provided by the PCM to see a change in scan data. So I get to 2v and scan data changes, I go to 2.5 and it changes, I go to 3v and it changes and so on and so on until I hit 5v max via my tool. Never once did the voltage provided by my Hook match the scan data BUT it did move. This tells me the circuit is in-tact and the PCM is capable of seeing a voltage change on the circuit its monitoring.
I KNOW I have a bad oxygen sensor for sure due to the open heater element. I'm thinking to myself Brian, sell the damn 02, quit over-analyzing everything and pull the trigger.
FINE. An oxygen sensor is sold and installed. Lets check our work.
I review scan data and immediately see activity on the suspect 02 sensor. NICE! Score for the good guys! It was hung at 4v prior so this is good. Looks much better I think we got this one nailed for sure. But wait, lets take a deeper look at that scan data.
Yes, my O2 has activity now, however the voltage does not match the other bank. The scan data you see has both raw voltage and also translated 0-1v voltage. Notice bank 1 is "hunting" 3 volts while my suspect bank 2 is hunting 3.6 volts. The signal is still rich! Rut row.
I have already confirmed wire integrity. I have replaced the 02 so it's probably good. The only thing left is PCM logic. Right? Maybe I have a bad ground? Can a poor ground affect only one o2 signal? Lets check anyways. So I verify all powers and grounds to the PCM are good and they are. Crap...
Am I missing anything? Condemning a PCM is one of my LEAST favorite things to do. I rarely have a "YEAH THATS IT!" feeling when my conclusion is made. Unless that PCM was hit with a baseball bat or better yet, a mack truck at 80 mph, I NEVER have a comfortable feeling in my gut that I've got this diag under control.
A PCM was ordered, Installed, VIN programmed, SKIM code installed and data reviewed.
PHEW! O2 Data is back to normal. Awesome!
My question to y'all. What would you have done differently? How do YOU confirm a PCM can process o2 signal voltage correctly?
So far, using a Hook has been flawless for me. I think, in this case, it still would have been IF I would have paid more attention to the bias voltage on this circuit.
THAT, is indeed, what failed. The guy who is providing that bias voltage, was not providing the CORRECT bias voltage that the o2 signal rides on. And his name, unfortunately.... :( is PCM.
On a positive note, the explanation was given to the customer with supreme finesse from start to finish. Not having the appropriate staff in the front of the house AND the back of the house can make jobs like this turn sour very quickly. The customer understood exactly what happened and was grateful her car was in good hands and fixed properly. Obviously, having the complete diag in the beginning makes situations like this easier BUT no ones perfect.
Great write up Brian! Although your diagnosis is spot on. I probably would've added just one more test. That is, to short the O2 signal wire to ground just to see if the data would concur. Just for grins. Nonetheless, good game plan. Maybe others can chime in on how your bias voltage was higher than it should've been. I'm also curious to the confirmation of the correct amount of bias voltage, just to add to the mental archive:) looking forward to more of your case studies!
Great write up . I myself also find it really hard when condemning a faulty ecu . I just don’t really understand fully what goes on inside the box . some of it I do . But this post just added more light to something I’m stuck in the dark in . Maybe soon I will reach the end of This tunnel lol ...
Did you compare the sensor voltage, tested with a scope to the voltage reported on the scan tool? This is something I do quite often when I am in doubt about the accuracy of the scan tool data
Albin, yes I did actually. I Should have included scope captures but I didnt have them saved. The true voltage on the signal wires did match what was displayed on the scan data.
when your testing with the PP-Hook wasn't showing quite what you expected, and you thought it was due to the bias, I would have tried it again on a different sensor's signal ckt to see how that one reacted..
Thanks! Out here in West Maui I don't ever see enough of anything to know the system rote, nor do I ever have another of the same car around, so anytime a car gives me the "gift" of more then one of something, (like WSS, or O2, or Injectors) I love to look at one that is NOT acting funny first then look at the suspect one,
Nice case study Brian. Very nice catch going the extra step to catch the higher voltage on bank 2.
Did you happen to catch the goal voltage request for bank 1 and 2 to see if the upstream sensors were meeting that goal voltage ?
If the upstream o2 had been failing for a while then possibly it changed the mixture at the cat enough where the downstream o2 was requiring a different goal voltage from the upstream o2 in order to maintain cat efficiency.
Did you reset all adaptive after the sensor replacement before checking the voltages? If so then the theory of the goal voltage causing your higher o2 reading on bank 2 is null and void.
If the adaptives were not reset it is possible that the old goal voltage was causing the higher sensor voltage.
Consider this. If the upstream o2 was falsely reporting rich due to the failed heater then the fuel trim correction would be lean.
So with a false lean request the downstream o2 would report low voltage due to the lean correction upstream.
So in order to raise the voltage on the downstream o2 to maintain cat effeciency the goal voltage of the upstream would be increased (rich request) which could have caused your slightly elevated upstream o2 voltage.
While using the hook is a good way to test the wires integrity I also wonder if somehow it may have damaged the module. It May be hard to prove it damaged it but it makes one think.
Since both of those wires have a bias voltage I’ve had decent luck checking circuit integrity without the use of an external voltage source. Like Mario stated short the wires to ground and see what the voltage is at the ground connection and what the scanner is reporting. If the scan data indicates something other than what is measured where the wire is grounded then it is time to check the voltage at the module connector.
The voltage at the module connector for the o2 wire should read the same as the scan data if not and the module powers and grounds test ok then the module is at fault.
If the o2 sensor return and ground can both be pulled to 0 volts by grounding them and they read 0 volts at the module connector then it would indicate the harness is ok. This is only true when there is a bias voltage on the wire.
Another test I like to run is to short the o2 ground and signal wire to each other and check the voltage at the sensor and on the scan tool to be sure they are both reading the ground voltage now.
Checking the voltage at the sensor while loaded (short to ground) and checking it at the module connector should also have a voltage drop test performed between the module connector and the shorted connection at the sensor connector.
Thanks for sharing your learning experience for the rest of us to learn from. We are all in this together.
While observing the first picture, the one showing the graphed O2 data, I was betting you'd find both Bank 2 O2 sensor connectors swapped. Looking at B2S2 (V) doing what B2S1 should have been doing, and the opposite as well, I was curious how you were going to explain how someone could have done what I thought. I see how well you methodically blew my first assumption out of the water (te-he). Good job Brian!
What did the scan tool pid read when the signal & signal return at the 02 sensor was jumpered (Terminal 3 & 4)
If you were concerned about a suspect bias voltage on one bank, I think one thing you could have done is simply unplug both upstreams and compare the open circuit bias voltages on the scanner. Bias voltages on O2 signal circuits will be present in a OL/failsafe event. You can measure it with your DMM with it unplugged, too. But keep in mind that your DMM is making it a complete circuit (albeit it is a massive internal resistor) and can show a slightly different voltage than the former method I mentioned. The only thing that you are really looking for here is consistency in the bias voltage readings from bank to bank in this scenario, though.
BTW, nice write up! I enjoyed reading it!