 | |||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||
|
Shooting in the DarkPosted 5/16/2000By Jeff Bach
Say "Scope Triggering" in a crowd of auto techies at a convention or a technical conference and watch the ears perk up. Not long ago there seemed to be very little interest in this subject, but as more technicians are getting used to using their lab scopes, triggering is becoming more of an issue. After all, it's the next level of expertise that needs to be achieved after learning the scales. Trigger options vary widely on scopes and should be (in my opinion) a major consideration when considering purchasing a scope. Manual trigger options should be easily accessed and at a minimum allow you to trigger from an adjustable voltage level, on a plus or minus slope. Additional options that I find useful are trigger from A or B channel, trigger from a single event, time-delay trigger, and events delay. When you consider the ever-increasing complexity of the systems we are required to troubleshoot when they are hard failures, it's no wonder shops get frustrated with problems that are intermittent. This is an area where a good triggering technique is most helpful. I have seen this scenario more than a few times with intermittent stall complaints: "First they said it was the fratestator, then they said the rollersnober was bad and that made the wicklefoggle go out, now they say I need a new cangamangler and it still does what it did when I first took it in. I think they are just shooting in the dark." Sometimes in this profession, that's the only time them critters come out. A lot of stalling complaints can be classified into this category. The first example I have is from a '92 Buick Century. The customer said it usually stalled while idling when the engine was warm. First step: Verify the customer complaint. I didn't have a clue whether it was fuel or ignition related, so I started by putting the current probe around the fuel pump wire and watching for current loss at the time of stall. To do this, I set the scope trigger to about 20 percent under the normal current and on the minus slope. This means I won't see a signal, perhaps, unless the fuel pump current drops more than 20. Fuel pump current on this car is normally around 5 amps (see Figure 1). When the car finally stalled, I had no pattern until a couple of seconds after the car died, which is normal as the ECM shuts the fuel pump relay drive signal down when it loses rpm. I had verified fuel supply and now wanted to see ignition (Figure 2). By monitoring the ignition primary with a current probe and the scope set to "wait for trigger" with the voltage level at about half the level of the signal (about 400mV), the trace won't repeat if more than half of the signal is lost. This leaves the last good signal left on the scope screen and shows that I have lost primary current. The wide last primary current signal has me leaning toward a signal generation problem rather than a power or ground loss. The next thing I wanted to see was how the crank sensor signal looked when I lost primary current. Figure 3 was the result. The crank sensor signal disappeared with the primary current and a new crank sensor fixed it. When diagnosing an "intermittent stall" or "quits at times" complaint, I try to pin the problem first to either fuel or ignition by using a current probe (probes) to monitor one or both and setting up the scope to capture the stall if it occurs. The '94 Seville in Figure 4 was stalling at times for no reason. While driving it with the current probe on the primary circuit, triggering from halfway up the positive slope, I saw that I was losing ignition when it died. I wanted the next shot to show whether or not I had lost EST when the primary quit. Triggering from the ignition primary and monitoring the EST from the ECM, I could see that the EST signal was dropping out. Since I now knew I was losing the EST signal, I wanted next to see if I still had a reference signal when the EST went. I set the scope to trigger on the EST signal and monitored the REF on channel two. Figure 4 was the result, which led me to replace the ECM that fixed this one. The stall capture in Figure 5 came from an '83 Olds with a V6 that had an engine installed recently. It came to me with a complaint of the carburetor running lean on cold start and that's what it felt like. Each time the engine would start it would run until the choke pull off opened the choke a little, then just die like the choke was pulling open too far. After doing some checking (closing the choke tighter) I realized that this car was losing ignition after starting. I triggered from the primary current (red wire feeding the distributor) and as you can see (Figure 5), I lost primary current but still had a reference pulse. I suspected the module but wanted to be sure, so I triggered from the primary again and monitored the EST. This is the signal the ECM sends to the module to fire the coil after the engine starts. Note that the primary circuit (Figure 6) stays energized and never discharges. This is because the module fires the coil when the falling edge of the EST signal drops below a predetermined level. Also note that the EST stays almost 1 volt above the low ref while the engine dies. I now had reason to doubt my suspicion of the module. I turned toward the ECM ground. Triggering from the primary again and grounding the scope at the ECM ground, I noticed that as the engine stalled I got one more primary pulse when the EST dropped below the ECM ground level (Figure 7). It was about this time that my brother, Eric, noticed that the EFE relay on the fender would click when the engine stalled. He unplugged it and the problem seemed to disappear. I triggered from the relay current while watching the EST on the other scope channel and captured the image in Figure 8. At the moment the EFE grid current appeared, the EST signal shifted below ground and the engine stalled. Notice the EST pulses getting closer together, then widening after the EFE current signal. The shift in the EST signal was directly related to the EFE current. The next shot I wanted was the ECM ground compared to the battery ground while triggering from the EFE current (Figure 9). The ECM ground was shifting up about 600mV. This confirmed my theory. I found the loose ground on the back of the cylinder head. The next intermittent stall was a "Randomly quits while driving down the road" complaint. Restart time would vary from 10 minutes to overnight, but the vehicle would invariably start by the time a wrecker got there. The customer wanted a new fuel pump because someone told him it was getting spark when it wouldn't start and that if you sprayed carburetor cleaner into it, it would start. After walking back to the shop from my first extended road test while monitoring fuel pump current, which looked fine when the engine died, I got some lunch and had Eric take me back for some more hunting. This time I set the scope to trigger from the crank sensor signal on one channel, while monitoring fuel injector and ignition current with a current probe on the other channel. The car started and ran fine back to the shop. The next day on the way to lunch ... "Bang!" "Got one" (Figure 10). You can see the crank sensor signal just quits in mid-screen. To me, having a scope with the ability to trigger on this event and capture a shot like this (when and if it just happens to act up) makes me feel all tingly inside. I got the car started again after only ... a little while. I wanted another shot of this with some more injector signals since this shot showed me that the crank sensor was the problem, but I couldn't see - from the picture - why the car stalled. I still had ignition current pulses and the next injector pulse should occur three ignition pulses down (just off screen). I reset the scope with a longer time base and drove the car again. Figure 11 gives a little better picture of why I lost fuel. The interesting thing on these cars with dual signal crank sensors is that you can still get spark when one of the signals quits. The injector signal is what goes (Figure 12). You should see an injector pulse for every three ignition pulses. If I have to go hunting in the dark and I have a choice, I want a gun that has a scope with adjustable power, infrared and a laser sight. Thermal imaging would be a nice luxury along with automatic target acquisition and remote firing. Doesn't sound very sporting does it? Imagine overhearing a conversation by a couple of technicians talking over coffee in the break room saying, "Hey, Charlie, did you see that job Tuvi got? "Yea, Bert, I heard the lady say, 'It stalls sometimes out of the blue then starts back up after sitting a while and may run on all day.'" "You know he's got one of them scopes with variable voltage, time and events delay, + or - slope, manual triggers." "Yeah, I heard he can trigger from either channel, and he's got a current probe too." "Poor intermittent crank sensor, ... didn't stand a chance." "Nope, don't seem right."
|
||||||||||||||||||||||||||||||||
Copyright (c) 1996-2008. Automotive Service Association. All rights reserved.
 | |||||||||||||||||||||||||||||||||
