All for one and one for all… For some reason, the PCM
treated this odd little trio of O2 sensors like musketeers.
The All-too-familiar O2 Sensor
Just about every technician on the planet who has done much wrench work over the past twenty years would like to have a dollar for every oxygen sensor he or she has replaced, and most of us would be even richer if we had a nickel for every mile we’ve driven while attempting to verify a fix or locate a problem that revolved around an O2. These hot little sensors can be extremely useful to a scan tool junkie, or they can be the most annoyingly wretched little gremlins imaginable.
Since most multi-bank platforms have had an O2 sniffing the exhaust from each bank for more than a decade now, we’ve all learned to use it while chasing misfires, isolating leaking injectors, finding vacuum leaks, checking EGR operation, troubleshooting AIR systems, etc. Terms like Open Loop and Closed Loop became the order of the day more than twenty years ago, but I’ve known some techs who spent a lot of years to catch on to exactly what those words meant. It took nearly half a decade of slugging it out with O2’s before a lot of seasoned techs realized that a misfiring spark plug would cause an O2 to read lean and not rich, since the sensor doesn’t sniff fuel, but oxygen.
As for the GM engineers, they were intelligent enough to give us all a peek into the adaptive learning tables a long time before the other two American automakers did. The first time I heard the terms “Block Learn” and “Integrator”, I thought they were awfully mysterious. The first generation of 4.0L Jeep microprocessors named these figures something else, but after the federal J1930 regs took hold in the ‘90’s everybody was required to refer to the fuel adaptive figures as Long and Short “Fuel Trim.”
In addition to the oxygen sniffers providing feedback for each bank, the J1930 regs caused each vehicle to sprout a new one behind each catalyst. Big Brother wants to know how well the expensive converter is performing and since it stores oxygen, the O2 sensor is a natural for the job, and the downstream sensor(s) should switch a lot lazier than the upstream ones. As a result, we all had to learn which sensor was which, sorting out confusing new O2 designations. i.e., 1/1 (#1 bank upstream), 2/1 (#2 bank upstream), 1/2 (#1 bank downstream), 2/2 (#2 bank downstream).
The Need For Heat
An O2 sensor has to reach at least 600 degrees before it’ll work right, and if you’re doubting a particular sensor’s ability to sweep back and forth between the 0 and 1 volts it’s designed for, you can hook a voltmeter between the signal wire and its ground, then fire the tip of it with a propane torch. When you get the sensor is hot enough and it’s bathed in flame, the sensor output will sweep up to its highest available voltage. Moving the flame causes the voltage to plummet immediately. The PCM depends pretty heavily on this voltage to keep air fuel mixtures in line, thus becoming central player in the drama of keeping CO2 readings high and everything else low at the exhaust pipe.
Since the O2 has to be hot in order to work efficiently, most automakers started equipping each sensor with its own heater. At first the two heater wires were a simple fused 12 volt power lead with a hard wired ground circuit, but with the advent of J1930 and the desire to monitor the now all-important heater, the PCM took control of the O2 heater.
Some Chrysler PCM’s check their heaters in the middle of the night when we’re all sleeping by feeding a gentle 5 volts to the sensor signal wire and quietly firing up the heater to see if the reference voltage drops off.
Ford sensor heaters are a bit less spooky simply being continuously monitored by the PCM for current draw.
An Open and Shut Case?
How many times have you snatched an O2-related code, popped a sensor on the vehicle, done a quick test drive and pushed it out the door? Everybody has pulled that stunt at least once. “When I get an O2 code, everybody gets two sensors,” I heard one guy say, because in the early days he would change the sensor in one bank only to have the car come back with the other sensor flat lining a couple of weeks later.
The F150 in question here belonged to the Used Car Department (UCD) at my old dealer, and since I had a day off from teaching I showed up at the dealership to engage in a day of healthy professional development. The UCD tech had jerked two lean codes and three O2 heater codes, all pointing at some sort of weird O2 problem, and so he threw it into the driveability slot. In driveability, the guy pondered the codes, checked the PIDs, and found Long Fuel Trim figures at +25 percent on each bank. The Short Fuel Trim figures couldn’t decide what to do when the engine was off idle, bouncing here and there in wild gyrations.
One interesting PID we noticed was the O2 Heater Failure flag, which read a big fat YES on O2 sensor 1/1 (front sensor, #1 bank). We’ve all seen faulty O2 heaters, and usually the sensor itself turns out to be the problem, which appeared to be the case this time.
Heater Fail PID
It was a telltale sign that there was something wrong with the #2 upstream sensor heater.
And since this was a UCD vehicle and we didn’t want to deal with a nasty recurring MIL, both sensors were replaced, the adaptive tables dumped, and the test drive went beautifully. One crucial error came to pass when we got so caught up in watching the beautifully performing Fuel Trim readings and the pleasant darkness of the “Check Engine” light that neither of us thought to re-check the heater failure monitor PIDs. Since the MIL was extinguished and the Trims looked good, we parked the truck after a twenty-mile drive and moved on.
Big Bad Boomerang
The MIL winked back on while the UCD guy was driving the truck for something else, so I pulled it back into the service bay. The three heater codes had returned, P0135 (HO2S HTR-11), DTC P0141 (HO2S HTR-1/2), DTC P0155 (HO2S HTR-2/1), as well as a P0171 (HO2S 2/1 too lean) and P0174 (HO2S 1/1 too lean). Since cool O2 sensors generally flatline and throw lean codes, I opted for an O2 heater check.
Using a homemade tool which consisted of an old O2 sensor connector with a 12 volt bulb wired into two white heater wires, I disconnected O2 sensor 2/1 and installed the tool there because that sensor was the most accessible. With the key switched on, the 12 volt bulb should light, and this one didn’t. On Jeeps the engine has to be started before the heater circuit will fire up, and since the PCM can opt not to energize the sensors, when in doubt I generally start the engine on Fords just to be sure. Still no light.
What I didn’t know at this point was which leg of the heater circuit was open. The nifty thing about having the tool installed was that it brought the terminals up to where they were easily reached with a test light.
With a 12 volt bulb wired to the two white heater wires on an old O2 sensor connector, checking O2 heater circuit operation (or the lack of it) becomes a lot easier.
Pulling up the wiring schematic, I verified that each O2 heater circuit is powered by the EEC relay on this unit (which also fires up the injectors and everything else the PCM controls), so at least one wire feeding each sensor should have 12 volts with the key switched on. The schematic showed the EEC power coming to the sensors through different connectors, so there wasn’t likely to be an open circuit on that side of the sensor. A test light check for heater power confirmed my hypothesis. It was the ground that was lacking, and this particular O2 heater was fed by PCM pin 94.
Grabbing the Breakout Box (BOB), I disconnected the PCM and connected old BOB to the harness, leaving the PCM disconnected. This effectively takes the PCM out of the equation, which prevents confusion when checking harness circuits for shorts to ground, power, or other wires. With HO2S 2/1 plugged in, I should read 12 volts coming through the heater to BOB pin 94. My test light read a nice healthy 12 volts. (incidentally, using a test light instead of a high impedance meter to check the wiring harness is safe as long as the PCM is disconnected) Now it was time to disconnect the bulb from the homemade HO2S tool. If things were as they should be, the light should wink out. Bulb disconnected: Test light still on. We had a short circuit somewhere.
With the test light connected to BOB pin 94 (HO2S 2/1 heater control circuit). Homemade O2 tool installed. With the PCM out of the loop, the test light should wink off when the tool bulb is removed. It didn’t.
I decided to connect the PCM to the breakout box connector just to see if the PCM would decide to ground the control circuit. It didn’t. Leaving the PCM connected and rechecking the PCM pin assignments again, I found PCM/BOB pin 93 leading to HO2S 1/1 and pin 95 feeding ground to the only downstream sensor, which is designated HO2S 1/2. Since the downstream sensor connector was easier to access than the connector for the other upstream sensor (1/1), I opted to move my light to pin 95. I had my friend Donnie disconnect the downstream sensor (1/2) and the test light winked out.
Not Feeding Volts
Disconnecting the downstream sensor with the test light connected to BOB 96, the light winked out. The HO2S 1/2 circuit was working normally.
The HO2S 1/2 circuit was working properly. With the test light probe resting in BOB pin 93, I worried the HO2S 1/1 connector loose while Donnie watched the light. It winked out, but now we found that pin 94 had also gone dark. It appeared that the wires leading to pins 93 and 94 were shorted together somewhere.
Tracking It Down
Between the PCM and the engine is nestled a big connector the Ford wire harness gurus numbered C101. Most Ford pickups since 1992 have had this connector, which marries the 12A581 harness to the 9D930 body harness.
Disconnecting C101 isolates the 12A581 engine harness from the 9D930 harness… When the O2 heater circuits normalized after this connector was separated, we realized the problem had to be on the 9D930 side of the harness.
On the 1996 F150, this harness connector is found between the air cleaner and the master cylinder, and all three heater circuits leave the PCM, travel through the C101, each one of the three feeding heater ground to a different HO2S. EEC power vanished from pin 94 when C101 was disconnected, but for test purposes we back fed 12 volts into the EVR solenoid EEC power terminal. Since the BOB was out of the loop with the C101 disconnected, we researched the C101 pinout and rechecked the circuits in question at the female side of the C101 connector. When each respective sensors was unplugged, the test lamp was extinguished, pointing us away from the engine harness and toward the PCM side of the circuit.
With 12 volts back feeding the EEC circuit at the EVR, we checked each sensor circuit at the engine side of the C101 connector.
With the C101 disconnected, the short circuit between pins 93 and 94 disappeared. This isolated the problem to the few inches of wire between the male side of C101 and the PCM. With C101 still disconnected, but with BOB plugged in between the PCM and the PCM connector, we measured a 2.6 ohm short between pins 93 and 94.
Bingo! With the BOB connected to the PCM and the vehicle harness disconnected, we read a hard short between PCM pins 93 & 94. The PCM drivers were compromised, possibly having been damaged by the first shorted O2 heater. Furthermore, since theO2 heater circuit readings it was getting didn’t add up, the PCM opted to shut all three of the heaters down and throw DTC’s for every one of them. I haven’t seen this strategy spelled out in any of the books.
Measuring from 94 to 95 produced infinite resistance, as did 93 to 95. Disconnecting the PCM caused the short circuit to evaporate. The F150 would need a $280 PCM.
An interesting side note to this story is the fact that the PCM detected the short and knew enough to shut all the heater circuits down. Also interesting is that while it only showed a fault on sensor 2/1, it was throwing codes for every O2 heater on the truck, and since the heaters were shut down, the sensors were largely unresponsive unless a long test drive was under way. On a long drive, the exhaust stream would heat the sensors to the point that the fuel trims looked good, but the UCD guy’s test drive somehow satisfied the drive cycle criteria the PCM needed to re-illuminate the MIL. The original 1/1 sensor was long gone by this time, but its internal heater may well have been shorted, which could been the causal factor in the destruction of the PCM’s internal circuitry.