It had apparently been everywhere twice, and it still bucked and snorted when she dug in her spurs on takeoff.
I had a two-week break from teaching last summer and decided to spend some of that time back at the Ford dealership. I'm constantly working on ways to remain current in the field for fear that my students will be learning from someone who's fallen behind the times.
I wanted to help out without stealing labor hours from the two regular driveability guys, so I started out slowly, working as a consultant and advisor and helping out where I could. By the third day, the service advisors were throwing tickets at me like I was a regular; but there was plenty of work for all of us, so the other guys didn't mind. I saw some interesting stuff I could really sink my teeth into.
Buckin' BroncoOne example was an interesting little Bronco II that had a relatively simple write-up: "Unit skips while driving down the road." This is a fairly common complaint, but it was no ordinary problem. The SUV ran better than I expected on the test-drive. The vehicle was smooth and responsive on hard acceleration, but when it got warm enough to drop into closed-loop fuel control and started using oxygen sensor input to adjust the injector pulse, the tailpipe started belching soot and I experienced the bucking and jumping the owner wanted to correct.
Closed-loop fuel control is as old as oxygen sensors, which first appeared in the mid-'70s in Europe. One of the interesting things about a vehicle entering closed-loop is that you should never feel it happen. If you feel it, something is wrong. And if you punch a Wide Open Throttle (WOT) and feel the system drop into open loop, the engine will come to life again instead of running like a slug. When you level out to a more sensible throttle angle again, the engine generally will run well for a second or two until you feel it drop back into closed-loop once more. Then the sluggish bucking will return.
Armed with a scan tool, anyone who understands Electronic Fuel Injection can watch the PIDS and determine when closed-loop takes place. My problem was that Fords didn't have datastream capabilities in 1987, so I headed back to the shop to rendezvous with my old friend, the Service Bay Diagnostic System (SBDS).
Mental analysisOn my way back to the shop, I considered my first course of action. I could check fuel pressure; older fuel pressure regulators fail fairly regularly and cause the pressure to bounce up near 100 psi. However, when that happens, a WOT kick won't clear up the problem. What I experienced on my test-drive was more likely to be related to the relationship between the Powertrain Control Module (PCM) and the oxygen sensor.
Connecting the SBDS to the VIP test connector, I ran a "Key-On, Engine-Off" (KOEO) test and pulled a code 41 from memory. There were no hard faults. Next I ran a "Key-On, Engine-Running" (KOER) test and got another code 41. This didn't necessarily mean the oxygen sensor was bad; there could be an oxygen sensor heater circuit problem, a shorted or open signal wire, or, most interesting to me, a bad oxygen sensor ground.
Having a look at the graphThe SBDS machine has a breakout harness that connects between the PCM and its connector. Because I had entered the Bronco II at the beginning of my diagnostic session, the machine knew the connector pins. There are aftermarket units available that will perform this same job on most vehicles from 1981 to 1999.
I selected several pins and let the machine build a graph. See Figure 2. The O2 signal appears in yellow, with voltage levels on the left and a timeline across the bottom in standard graph style. Note the odd O2 voltage displayed on the screen. Instead of nice humps between zero and 1 volt (V), the screen shows a negative voltage reading of -1.2V raggedly making its way across the bottom of the graph. I have seen strange figures like this before, and they generally indicate a bad ground somewhere. The body-to-block ground can cause screwball negative O2 readings that will clear up when the alternator is disabled, but unplugging the alternator did nothing to correct this particular reading. I decided to check the O2 sensor ground.
Well-grounded inspectionAfter the Ford 2.9L engine had been around for a while, I began noticing a recurring problem in the service bay, particularly on Rangers and Bronco II units with automatic transmissions. There is an orange ground wire that leads from PCM pin 49 to the back of the passenger side cylinder head. It should be clamped between the dipstick tube bracket and the head. The circuit inside the PCM depends very heavily on this ground in order to properly read the oxygen sensor signal.
However, some technicians will forget to reconnect that ground when reinstalling the transmission. Black smoke problems and odd O2 sensor voltages are common with a loose or disconnected ground.
Incidentally, the Ford 60-pin PCM connector had five grounds leading to it until 1989. Pins 20, 40 and 60 are connected to the negative battery post or the vehicle body near the PCM. Pin 16 is a ground from the TFI module mounting plate at the distributor, and finally, Pin 49 is the oxygen sensor ground.
In 1990, a four-wire sensor supplanted the three-wire oxygen sensor. PCM pin 49 feeds a ground out to the oxygen sensor from the signal return circuits inside the PCM. Take note that the ground on four-wire sensors goes out to the sensor instead of coming into the PCM from the back of the cylinder head. Late '80s V8 engines with two O2 sensors actually have two O2 grounds, one to each head. Older oxygen sensors have just one wire, with the ground provided at the point where the sensor screws into the exhaust.
On visual inspection, I found an oxygen sensor ground wire that had been connected to the wrong spot, which was a bolt on the upper intake. There are simply too many gasket and bolt points between this spot and the O2 sensor - located in the exhaust header pipe - for the PCM to get a good O2 sensor ground reading. Another ground wire is supposed to go under that particular bolt, but that one had apparently been attached somewhere else. Some technicians will put the orange oxygen sensor ground wire here to avoid fighting with the confined space at the rear of the head, but that connection can be accessed fairly easily through the wheel well, and I took the necessary measures to move the ground to the right spot.
Better readings, but...The SBDS 60-pin graph looked a lot different after I moved the ground, but there were still problems. The nice switching humps I had been looking for were there now, but the sensor was switching in the wrong voltage range.
Notice that the switch on the yellow O2 sensor trace was switching between -0.2 and +0.3. The PCM liked the signal a little better, but the engine still ran ragged because the O2 voltage never broke into the rich range above +0.5V. A quick look at the oxygen sensor revealed a fairly new-looking sensor, so I decided to look for further ground problems.
More ground concernsAt the negative battery terminal, I found a universal-style terminal of the type that I have no use for. Twenty minutes later, I had replaced it with a good solder-on terminal and the secondary grounds were securely soldered into the terminal with the battery cable. However, the SBDS graph kept looking worse and worse.
Another KOER test produced yet another lean code, and I decided it was time to replace the oxygen sensor. I've seen sensors get out of sync with what's proper. They will read within a range that would be acceptable if it were higher on the scale, but the whole pattern of sensor operation will be out of line with what the PCM wants and the engine runs accordingly.
Finally fixedA new O2 sensor made the little Bronco run like a new horse. Pulling up a new 60-pin graph confirmed it. The range was proper, and the switching was normal. This is the pattern that a driveability guy likes to see.
My final test-drive went very smoothly. The Bronco II performed like a Mustang compared to the old nag feeling it had when I first drove it. And it gave me some satisfaction to know that I had managed to make the repair surgically, without changing several hundred dollars worth of hardware.