Repairs Performed:
Throttle body cleaned; ignition module, coil pack, IACV and ECM
replaced
Dave’s Diagnosis:
The car didn’t exhibit any problems during a test drive, so this appeared to be an
intermittent malfunction. The vehicle owner was contacted to see if we could get any
additional information, and when asked when the problem first started occurring, the
response was “about 3 weeks ago.” One of the students pointed out that a
hot-weather trend had started about — you guessed it — 3 weeks ago.
Back into the car for another test drive with the air conditioning
on, and when we slowed to a stop, sure enough, the SES light flickered and the engine
died. Next, we connected a scan tool, armed the snapshot mode and drove the car again. We
coasted to a stop, the light flickered, the engine stalled and we captured the scan data.
Back in the shop, we analyzed the data, frame by frame. Nothing at
first, but then we saw it: the Knock Sensor line read: NO-NO-NO-NO-YES-NO-NO-NO-NO. We
compared the Timing Advance line and there it was: 17-17-17-17-0-17-17-17-17. We found the
problem, but what was causing it?
We’re fortunate enough to have a dynamometer in the shop, so we
tried to recreate the conditions in the shop. It didn't take long to find our root cause.
When the car accelerated, the engine jumped up about 3 inches. When the car decelerated,
the A/C compressor disengaged momentarily, the engine rose another inch, the A/C
re-engaged and the engine settled back down with a thud. Oldest problem in the world: a
collapsed engine mount insulator.
When the engine dropped, the mount would contact the frame and the
resulting vibration was transmitted to the ECM as detonation. The bottom line was the
ECM was trying to correct a problem that didn't really exist!
Why didn't the original technicians find the bad motor mount? Two
reasons: a bad mount tends to be less apparent in a manual transmission-equipped vehicle,
and more importantly, they didn’t test drive the car with the A/C on. In other words,
the car never exhibited the symptom while they were diagnosing it! Talk about
shooting blind!
The Vehicle:
1986 Buick Century, 3.0L. equipped with PFI, DIS and Automatic
Transmission
The Problem:
Runs poorly, SES light on, DTC 34 (MAF sensor circuit voltage low)
Repairs Performed:
2 MAF Sensors and ECM replaced
Dave’s Diagnosis:
This one took the class about 2 minutes to diagnose. When the engine
started it ran terrible, but what got our attention was how quiet the exhaust was. Too
quiet. A walk to the rear of the ear revealed that nothing was
coming out of the exhaust pipe. The oxygen sensor was removed, the car restarted. A
mushroom cloud of carbon blew out of the O2 sensor port, the car ran great (loud but
great) and no SES light: Plugged converter.
The trouble code and SES light were the ECM’s way of screaming for help. As the
exhaust backpressure increased, the airflow into the intake manifold decreased and the RPM
just didn’t correspond with the calculated load.
Too often, a technician focuses on being a diagnostician and forgets to
be a mechanic!
The Vehicle:
1989 Dodge Caravan 3.0L. equipped with PFI, Distributor Ignition and
Automatic Transmission
The Problem:
Surges at freeway speeds after 30–40 minutes of operation; no stored
trouble codes
Repairs Performed:
Distributor cap, rotor, ignition wires, fuel filter, TPS, MAP and O2
sensor replaced, fuel pressure and volume tested okay
Dave’s Diagnosis:
The first step was to recreate the problem. We attached a scan tool,
armed the snapshot mode and headed for the freeway. Sure enough, after about 35 minutes
the van began to surge: Symptom confirmed.
Back at the shop, the captured scan data revealed normal parameters,
but we noticed that when the surge began, the O2 voltage was biased low and the injector
pulse width was increasing. The computer was responding correctly to the oxygen sensor
input, so the indication was a mechanical problem. Something was leaning it out.
Next, we checked the fuel pressure. It held steady at 48 PSI, right at
specification. Accelerate the engine and the pressure responds accordingly. Disconnect the
regulator vacuum hose, 57 PSI. No problem there.
We left the van idling and rechecked the specs in another manual. We
returned to the van and did a double-take: The pressure now read 42 PSI. As we watched in
disbelief, the pressure almost imperceptibly continued to drop and after 15 minutes read
36 PSI. The fuel pump was slowing down!
As with any aging electric motor, the longer a worn fuel pump runs, the
more current it draws and the slower it spins. While this might not be critical on a power
window, it sure makes a difference when it involves fuel supply. |
