 | ||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||
|
'Diving Too Deep'Posted 10/18/2005By Jeff Bach
"Well, we know how often it turns out to be the computer, don't we?" I said. Since he's had this car, he has had the computer replaced by different shops three different times for various problems. A new computer has not corrected the problem. Tim had the car towed to us. I began my diagnosis with the usual research of the problem on iATN for commonalties. I turned up a common thread in the ignition switch. Also, this model seemed to be plagued by shorts in the engine wiring harness. I saw on the "guess list" that Tim gave me that the ignition switch had previously been chosen. I decided to treat the car as just a run-of-the-mill no-start.
The module controls the spark timing until the engine gets to 400 rpm. The PCM then takes over control of the timing. I tested the power and ground circuits for the module, which looked good to me at "key on." This led me to my next suspect, which was the crank sensor signal. I hooked up the scope to the purple wire, which is the signal wire for the crank sensor, according to the wiring diagram. Seeing the image in Figure 2, I either had a bad crank sensor, crank sensor wire or a pin fit issue. I tested the wiring and pin fit with an ohmmeter and a test pin, which checked good, and replaced the crank sensor. I hit the key and it started. "Another one bites the dust," I said as I backed it out of the bay. It suddenly stalled and wouldn't restart.
I rechecked the crank sensor wiring. No problem. I started trying to theorize some possibilities based on the data I'd collected so far. "Maybe you're looking too deep," said Vern, my service manager. Vern is a big guy with a ponytail. He's an ex-marine, a biker, and looks kind of like Mark Warren when he grins. I weigh Vern's words carefully. He's been right before. I hate it when that happens. I started thinking maybe the module was somehow grounding the crank sensor signal and went as far as substituting it for a known good one. I ended up with the same result, no reference, no crank sensor signal. I decided to look directly at the crank sensor waveform from the plug at the module to see what the module was seeing. I unplugged it at the module and cranked the engine. The crank signal was picture perfect. Now I'm getting that weird feeling you would get if you were walking on unfamiliar ground somewhere and suddenly realized that the ground beneath you was moving. You wouldn't be sure where to step or back up to next.
Key on again, I have system voltage. Crank, crank, crank. Again, it doesn't hit a lick. As soon as I stop cranking I look at the voltmeter again and it's reading 12.3 volts. I pushed the MIN MAX button to see the image in Figure 4. "Man, that's some meter you got," I hear Vern say. "What'd you do to push the smile button?" (Not that he's been paying attention to what I'm doing.) The low (MIN) reading had dropped to 3.95 volts. Now I can see where I went wrong. I didn't check the coil feed voltage under load before. I tested the ignition circuit straight out of the switch and got a steady 12 volts cranking. I double-checked at the coil, thinking maybe I moved the harness at the switch and changed something, but it was still sticking minimum readings in the 3.-something voltage range on the meter. My next stop was the C100 connector down below the brake booster. But when I moved the intake snorkel and looked down, I saw this inconspicuous-looking fuse stuck in the harness just before the connector (see Figure 5).
I cut the fuse holder out and soldered in a wire. I turned the key, and the thing fired right up ... problem solved. I was once again back on solid ground and not kicking myself too bad for overlooking the minimum dynamic voltage drop. Tim was tickled to have the car running and felt it was safe enough to let his daughter start driving it again. Now Vern teases me when he sees me in one of those "I'm-not-getting-anywhere" moods and holds the meter up to me and says, "Here - which button was it again?"
|
|||||||||||||||||||||||||||||||
Copyright (c) 1996-2011. Automotive Service Association®. All rights reserved.
 | ||||||||||||||||||||||||||||||||
Don't you hate it when someone tries to tell you you're on the wrong track and you trudge ahead anyway, only to find out later they were right? That's pretty much what happened to me on this '99 Cavalier that belonged to a friend's daughter. He lives on the other side of town and didn't want to bother me with this simple problem. He had spent quite a bit of time on the car with a chain store and was now being told it was about 90 percent sure the problem was with the computer. He was beginning to become (rightfully) skeptical. It was at this point that my friend, Tim, called me for a second opinion.
I hooked up the Tech II and checked for codes. There were none. Next, I thought I'd do my usual General Motors fuel pump relay two-second test. I turned the key on and made sure the fuel pump relay was being powered up for two seconds by the power control module (PCM), then cranked it for a few seconds to see if the relay stays powered up. This was my quick check to see if the PCM is seeing a reference pulse. The relay current dropped after two seconds during cranking. The PCM must not be seeing the reference pulse, which is sent to it by the ignition module based on an input from the crank sensor. Figure 1 is the reference signal (ppl/wht) I was getting during cranking with the scope hooked to it.
"Now what?" I thought. I pushed it back in and started over. Again no codes, and no fuel pump relay drive after two seconds of cranking. I collected the waveform in Figure 3 from the crank sensor wire.
I plugged the sensor wire back in and cranked the engine once more. The signal was gone again. Once again I started checking the basics. This time, the scope is hooked to the crank sensor and reference and a voltmeter was hooked up to the ignition power and ground.
Sure enough - it's the pink wire feeding the coil. As it turns out, this car had had some trouble with an intermittent short in the ignition circuit. One of the shops that had looked at it for the short problem previously had installed a fuse in the pink ignition wire to the engine compartment in an attempt to isolate at least one leg of the spider plant circuit that sprouts from the ignition fuse. The fuse holder had corroded and was dropping the voltage across it under the load of the coil primary charging. Now it all makes sense.
