Two Asian Adventures Seasoned dealership mechanics love working on their store brand vehicles because they usually know exactly what parts fail and what adjustments need to be made – and they have learned to do the work efficiently so as to beat book time. I’ve worked in that arena too, and I love it. But for those of us who work on any and every kind of vehicle, learning experiences abound. Factory shop manuals, Alldata, and Identifix, and Mitchell on Demand are all great resources, but there are times when they come up short, and in the end, solving the problem is our responsibility because the customer is paying us to make their problem our problem. A really good technician will keep going even when those sources of information are exhausted and nothing that’s happening under the hood even seems to make sense. Once again, there is no honor in giving up the fight when we’re going head to head with an uncooperative ride. This month we’ll talk about a ’96 Nissan Maxima and revisit the ’94 Dodge Stealth. Different automakers do things in totally different ways and this experience illustrates that. Stealth Redoux The 1996 Dodge Stealth that was the centerpiece of this previous April’s Motor Age Garage article ran like the proverbial sewing machine for several months and thousands of miles until it died a mysterious death one day and rolled back in to my shop on the back of a tow truck. To begin with, we found that it was only firing one of its three coils while spinning, with numerous induction and exhaust backfires. This particular car won’t talk to anything OBD II that we have (don’t you hate that?), so without a window into the system, we initially formulated our attack around the information we could glean from Alldata. This Mitsubishi ignition system uses cam and crank signals to determine which coils to fire first, with the PCM using three internal PNP transistors to drive three NPN transistors in an igniter assembly mounted just below the ignition coil array, with each transistor in the igniter triggering its own coil. It didn’t take long to break out the scope – what we found was that the PCM was only triggering one of the three coils. Using a low impedance test light connected to the positive battery terminal, with the PCM disconnected, we could “tap test” the appropriate backprobed terminals at the igniter and trigger each of the transistors to fire any of the three coils at will. A quick search on Identifix revealed just one hit on a Mitsubishi 3000 that matched our problem. The Identifix member postulated that if the cam and crank sensors aren’t in sync or one signal is missing, only one coil will fire. Both the cam and crank sensor connectors are easily accessible on the timing belt end of the engine just outside the timing cover. With our scope connected, we found a nice clean square wave crank signal but no cam signal at all. Paydirt? It appeared to be. What naiveté! One way or another, a cam sensor was in order, with the $100 sensor replaced, we had a crisp sensor signal, but the Stealth’s symptoms remained exactly the same. Since we had both cam and crank signals, and (according to the post we read) the sync between those two sensors was supposed to be a possible cause, the next logical conclusion would be that maybe the cam and crank were mechanically out of sync. Rotating the engine until the cam timing marks lined up on the front sprockets, we found the rear cam sprockets a couple of teeth out of time – the cam sensor blade is mounted on the rear exhaust camshaft sprocket, and if the Identifix post was right, that out-of-sync camshaft could cause a one-coil firing no-start. If so, we had a smoking gun, but how did this happen, and did the cam sensor fail at the same time the belt made its move? Not likely. This exercise was getting thicker and more misleading by the hour. I had Eric put the marks all where they were supposed to be and we reconnected the o-scope. Both the cam and crank sensors were producing neat square waves, but were they normal? I had no idea. The Identifix hotline guy (I contacted them) firmly said that if the cam and crank sensor signals were present and in sync, the PCM had to be at fault. Well, that sounded right – after all, the PCM wasn’t even trying to fire two of the coils, and the one it was firing was ‘way out of time. It kind of seemed to make sense, but a replacement PCM didn’t fix a doggone thing. The pain of that bad decision felt like we had just experienced a root canal. With the Identifix folks back on the line, I asked them what the cam/crank pattern was supposed to look like, which was a request I should have made earlier. We knew at this point that there was no way the two signals could be out of sync, at least not for mechanical reasons! The I-fix boys faxed me part of a bulletin that showed what the cam/crank sensor traces should look like, and while the cam sensor looked just like the illustration they sent, the crank sensor only had a third as many 5 volt switches as it was supposed to have. That explained why the PCM was only firing one coil, but what the heck could cause the crank sensor to produce such a beautifully symmetrical signal that was so totally wrong? Had the blade lost two of its teeth? Messed up Maxima Johnny is one of the toughest first semester students I’ve ever seen, but he’s no stranger to a wrench. He literally grew up in a salvage yard. This Maxima had previously been worked on by a back yard guy who was supposed to have replaced the water pump, but this one was towed in as a no-start. The diagnosis was quick and easy – the cam sensor was dead on this one too. A new sensor got the engine started but it barely ran and there was a lot of noise from the timing chain area, the radiator had a long split in it and the with the cooling system full and the engine running, the water dribbled from the water pump weep hole behind the A/C compressor at the rate of about a pint a minute. Like the Mitsubishi-powered Stealth, this was a DOHC V6, but the engine spins in the opposite direction and the camshafts are driven by three splash-oiled chains instead of a rubber timing belt. The water pump has a chain sprocket on it and is driven by the long timing chain. The water pump housing was the same heavy red varnish cover as everything else in there, so it had obviously never been replaced. Furthermore, the long chain’s tensioner plunger was missing and a fixed chain guide was broken, thus the chain noise. Along with the new cam sensor, the Maxima got a radiator, a water pump, a new fixed guide, a new tensioner (complete with the long aluminum and nylon slide), five fresh quarts of oil, and a filter. The chain was supposed to have three painted links, but we could only find two, so we had to creatively follow the illustration when reinstalling the chain. There was a bit of in-the-cover rattling on initial start until the oil-driven tensioner ratcheted itself out and took up the slack. At that point the Maxima’s 3.0L ran smoothly and silently with no leaks. Stealth Sensor With the sensor removed an in my hand, I could tell that the magnet had come apart – there was no hard evidence of the blade contacting the sensor, but it was hard to say exactly what happened. We did notice that there was quite a large fragment of the magnet clinging to the blade. That particular blade was magnetic because of the large piece of magnet, thus when the magnet passed by the Hall Effect Unit, the ill-timed oddball crank signal was produced. The very confused PCM reacted to the screwball crank sensor pulse according to its software program, fired the one appropriate coil based on the cam sensor signal, thus the induction and exhaust backfires. It took a replacement crank sensor to get this one going, and after we finally put our hands and educated eyeballs on the guilty part, everything was crystal clear. The mystery was solved. R.W.M.
Posted on: Fri, 06 Sep 2013 09:37:59 +0000
Trending Topics
Recently Viewed Topics
© 2015