I'll frequently 'stack' the bugs I plant. One simple bug-hunt I put together dealt with my 1997 Olds 88 trainer vehicle. I clipped one of the legs off the fuel pump fuse (it would seem to be fine when checked with a test light), then removed each and every spark plug and bent the gaps completely shut.
The hitch is that one can only polish troubleshooting skills to a certain rough edge by planting bugs on trainer vehicles. With that in mind, I always leap at the chance to give my students live troubleshooting work as long as the vehicle and the customer are qualified.This qualified customer had driven his C1500 long enough to run the odometer past 176,000 miles. Everybody who's familiar with the venerable old Chevy 350 knows of more than a few of these tough old powerplants that have gone a lot farther than this without needing repairs. The engine had been using some oil and was smoking out the pipes, but other than that, it ran fine. Rather than rebuild the existing engine, the owner decided to buy a remanufactured long block from a local parts supplier. He had swapped out engines before, and he did a reasonably good job on this one. The problem was that he simply couldn't get it to run right.
First look
While the owner had driven the vehicle to our shop, it was obviously in pretty bad shape. When we pulled it in, hydrocarbons were belching from the pipe, and the engine would hardly run. The student I had assigned to the job, Larry, quickly found the MAP sensor vacuum line disconnected and plugged. But as we plugged a Tech I into the ALCL connector and did some tests, it appeared that the owner had disconnected the vacuum line because the truck wouldn't run well enough to drive otherwise.With the MAP vacuum connected, Larry noticed that the engine wouldn't accelerate at all; it would only bog and spit back on snap-acceleration. At first it appeared to be an ignition timing problem. Just for grins, Larry grabbed the distributor and yanked the timing up about 20 degrees and the engine seemed to snap-accelerate normally under those circumstances. A quick check with the timing light indicated that the mark on the balancer and the rooster comb on the timing cover didn't match, because the timing had to be set at about 20 degrees BTDC to get normal acceleration. What I needed to know was the exact location of the number 1 cylinder top dead center (TDC) in relation to the balancer mark and degree comb.
Finding TDC
One of the first things I taught my students in Engine Repair class last semester was how to build a homemade tool that enables them to find exact TDC on any cylinder of any engine. Some tool manufacturers also sell these. Punching the center out of an old 14mm spark plug, I tapped the body with 3/8 USS threads and screwed a five-inch piece of all-thread rod into the new threads. With this tool installed in any spark plug hole, carefully turning the engine with a socket on the balancer bolt will bring the piston up until it gently contacts the all-thread rod. Making a mark on the balancer at this point and then turning the engine back the other direction until the piston contacts the all-thread rod again will pinpoint another mark on the balancer. Halfway between these two marks has to be TDC for that cylinder.When Larry started to turn the engine with his ratchet and socket, the engine wouldn't move. All he was doing was tightening the bolt. A closer look revealed that the balancer wasn't pulled up against the crank gear. Larry kept tightening the bolt until the balancer had moved about 5/8-inch farther onto the crank. Now we could finish our search for TDC. The existing TDC notch on the balancer turned out to be dead on target, so we installed the plugs, disconnected the spark bypass wire and set the base ignition timing to spec for that engine, which is zero.
When a vehicle comes to me running as poorly as this one, I like to look at the datastream. But beyond recording the trouble codes on paper, I don't generally pay too much attention to them until I've checked everything else. We obviously had codes related to the disconnected MAP sensor along with several others. Now with the MAP sensor reconnected and the timing set, we disconnected the battery to dump the PCM memory.Almost normal... sometimes
It annoyed us that, as our diagnostic process continued, the engine would go through intermittent spells of running almost normally. It seemed to pick those times very strategically. Once I simply touched one of the injector connectors, and the engine seemed to clean up its act and run great until the next switch-off and restart. After that time, repeated checks of the injector connectors and wires failed to duplicate this anomaly. During one of its normal 'moods,' the oxygen sensor voltage would sweep up and down for nice healthy double-digit cross count numbers, and the truck would have plenty of power. At one point I took a test-drive on a stretch of highway, and the only noticeable symptom was a high idle, which seemed to be a direct command of the ECM as IAC steps were anywhere from 100 to 145.This high idle fooled the owner into thinking a throttle body gasket would cure the concern. Well, he replaced it, and it didn't help. Switching the ignition off and restarting, the engine would return to its original rotten state of affairs. It would start and stall, start and stall, or start and idle at about 1,100 rpm, only to die when the gear selector was placed anywhere except 'Neutral' or 'Park.' Fuel pressure was steady at 13 psi even with the engine doing its worst.
An ECM he had installed from an aftermarket supplier wouldn't communicate properly with the scan tool, so I found another one in the electronics lab. I checked his PROM ID number against the school's information service database and installed a known-good ECM with his original PROM and Cal Pack. Now we had good communication. The engine was still idling fast, but the IAC numbers indicated the fast idle was commanded by the PCM.
Oxygen sensor... Check!
On one attempted test-drive, the truck ran so badly that we didn't make it out of the parking lot. We rechecked the trouble codes and found codes 43 (Electronic Spark Control (ESC) and 44 (lean exhaust). Engine vacuum was strong, but the oxygen sensor was frequently at flatline, even when the ECM was in Closed Loop. IAC steps were bouncing around between 25 and 145, depending on what sort of mood the beast was in. Block Learn was at about 145 and Integrator was hovering near 175.Disconnecting the oxygen sensor with the signal at flatline and measuring the voltage right out of the sensor, we saw a big fat zero - no signal at all. Removing the oxygen sensor and examining it, we saw no hydrocarbon fouling such as would be present if the ECM were correcting for a false lean condition. During a propane torch test with a voltmeter, we discovered it was able to sweep comfortably up to about 750 millivolts, but since these one-wire sensors are so cheap I decided to replace it.
Knock sensor... Check!
This unit is equipped with the stand-alone spark control unit that shares a mounting plate with the EGR control solenoid above the passenger side valve cover. This ESC module receives a knock signal through the blue wire from the sensor that is screwed into the water jacket under the left cylinder head.There are four wires connected to the ESC module: 12-volt power, ground, the knock sensor signal and the wire that delivers the knock signal to the ECM. The voltage on the signal wire rests on about 10 volts when no knock is present and drops in direct proportion to the intensity of the knock. As it was, we could monitor the 10 volts, rap on the block near the sensor with a small hammer and watch the signal drop below 7 volts. The sensor seemed to be working fine and the wire connector was in good shape. On this particular engine, the Knock Sensor PID would always change from "NO" to "YES" at about 1,650 rpm under load.
Out of time
One thing about teaching that I find to be somewhat troubling is that state requirements for curriculum time prevent my students from working on problems like this every day. We had to put this truck on the back burner until the following Wednesday. The owner opted to take the truck back and drive it until we could look at it again. This gave me time to check with my 'guru.'I spoke with an old friend named Mike who I worked with at an independent shop way back in '77. He told me that aftermarket reman Chevy V8 engines frequently arrive with their valve adjustment too tight.
When we got the truck back, we snatched the valve covers and one of my more experienced students went through all the valves to check the adjustment. Sure enough, he had to back off every nut nearly two turns before the clicking started.
The truck ran noticeably better after the valve adjustment, and a test-drive went well. But when I switched the engine off and restarted, we were back to the same old symptoms: The truck would bog and spit on acceleration. Stall speed was down around 1,700 rpm with the transmission in gear and the brakes locked. The oxygen sensor would flatline, and the engine would stall every time it was started or it would idle at 1,800 rpm.
According to one source I found, the minimum air rate should be set so that the engine idles at about 450 rpm with the IAC completely extended. Because I couldn't get the engine to idle well enough to find 450 rpm, I set the idle so the TP sensor read 0.50 volts.
Breakthrough
The next day, I told the class that the problem simply had to be in the throttle body. We had noticed that spraying some carb cleaner into the throttle body past the injectors would improve the engine's performance. I had already checked the injector pulse width for proper pattern and duration with the oscilloscope.Finding another throttle body proved to be fairly simple. We borrowed one from an '88 Suburban trainer vehicle, used a pigtail to adapt the TP sensor connector and started the truck. It ran like a dream. Removing the injector/fuel pressure regulator assembly from the '88 Suburban throttle body and installing it on the '91 Chevy was a simple fix. The truck had plenty of power, the oxygen sensor switching looked good, and the Block Learn and Integrator numbers were hovering around 128. In 'Drive' or in 'Park,' it idled smoothly, and the truck ran well on the road after throttle body part replacement but it would still stall occasionally at idle until we replaced the PROM and CAL pak in the ECM. I later found out that the truck had the stalling at idle problem even before the engine was replaced.
At this point, I still have no idea what was wrong with the original throttle body or why it waited until the engine was replaced to start causing problems. I'll let you know if I find out.
