Sometimes You Get Lucky
By Jeff Bach
'No-start' on '03 Mazda leads to crankshaft problems.
I recently had an '03 Mazda 6 come to me from a local tire store and maintenance center. They had changed the oil in this car and apparently the filter didn't get tightened properly. The car happened to belong to the daughter of a guy who runs a used car lot and sends this chain store a lot of his business. The Mazda failed to start the day after the oil change, and the customer saw a puddle in the driveway. They had it towed to the tire shop where they worked on it extensively before deciding that it had somehow developed a wiring problem.
They send me their wiring jobs, which I enjoy doing. I was warned that the car already had new crank and cam sensors installed so they knew those weren't the problem - although the powertrain control module (PCM) was still setting a cam sensor signal code. Those of you who have had the pleasure of dealing with one of these 2.3 variable cam timing engines probably already know where this is leading. But for anyone who will encounter one in their near future, there may be some time-saving information in this article for you.
This thing would try to start and then die like it wasn't getting enough fuel. Since I'm such a current-probe head, the first thing I did was grab my probe and scope and clamp it around the feed wire to the coil pack. Figure 1 was the result.
The fact that I know I have a cam signal code and a new cam sensor along with this pattern put me in mind of a logic lockout from the PCM. This is true on many interference engines where the PCM compares the cam sensor and the crank sensor signals to determine the cam and crank alignment and decide whether or not to allow the engine to run by not allowing ignition, fuel or both.
The first instance of this condition I ran into was on an '86 Audi Quatro that had been given up on by several shops. I couldn't find much information back then but was able to buy a book for around $325. I believe that put me on to the signals that I was able to scope with my 20 megahertz, 2-inch screen analog scope. I was in heaven watching that screen.
The cam and crank signals for this Mazda are easily accessed at the PCM hanging right on the driver's side behind the dash. It even has the connectors looking at you so there's no need to remove it for access. I pinned the sensor plus/+ wires and cranked the engine, taking a snapshot of the signals with my scope.
The result is shown in Figure 2.
The information I was using to diagnose this vehicle showed oscilloscope patterns for both signals that looked very much like these but there were no patterns showing correct correlation. I couldn't find anything on these signal alignment patterns during my research either.
When I first started looking at the scope patterns I was getting from the cam signal, I thought that it looked a bit weak compared to the one in the diagram. I then realized that signals in the diagrams were taken at 650 rpms while idling. Since this is a permanent magnet-generated signal and greatly affected by rpms, I decided that my signal was good enough for 200 or so rpm cranking speed. Although I was still not sure signal alignment was the problem, I went ahead and pulled off the rocker cover and aligned the cam slots in back of the engine. I used a dial indicator on top of the piston to determine top dead center (TDC) since the crank pulley alignment marks can't be counted on due to the fact that the pulley is not keyed to the crankshaft. The slots at TDC weren't perfect but they didn't look like they could be out enough to cause this to be a timing issue. The timing alignment for the crankshaft is done by removing an access plug in the rear of the engine and inserting a longer bolt that would protrude into the block far enough for the crankshaft balance weight to hit it and thereby set the crank in the proper position to be timed. I used a mirror and a light to watch the crankshaft as the weight came into view at the beginning of the hole for the plug. I then realized that at this position, the piston was a full tooth on the crank sensor wheel before TDC. The cam slots were also now out of position enough to cause a problem. I loosened the crank bolt and realigned the cams using a holding tool in the slots, then reset the crank in the proper position and tightened the bolt. After reassembling the top end, I set the scope again and hit the key. This time the engine started up like it was just shut off warm.
Figure 3 shows the resultant cam and crank signals.
Notice that the cam signal gets usable only after the first primary firing pulse. This occurs when the second compression stroke after the PCM sees the No. 1 cylinder crank pulse.
I took another shot at a faster time base to give a better comparison of the cam and crank signals for future reference. Figure 4 shows the proper alignment of the cam and crank signals as the engine starts.
I would have loved to have had this picture at the start of this job.
Figure 5 is another shot of the engine from crank to start with the ignition primary current added.
As you can see, the velocity greatly increases, and the cam signal comes into view only after the first firing of the coil. My curiosity about this cam signal wouldn't let me forget it so I went back to the shop Sunday and played a little with adjusting the cam sensor distance to the cam to see just where the signal gets unusable. I was amazed to see how little of a signal the PCM will see and still run the engine.
Figure 6 shows the least level of cam signal I could get the engine to run on.
Figure 7 shows the first "no run" cam signal.
The threshold for allowing spark is pretty tight. I'm guessing that the sequence of events that took place (ending with running low on oil) caused enough friction in the cams to allow the crank gear to slip slightly (about 10 degrees) on the crankshaft. This is enough to move the cam signal out of phase with the crank signal such that the PCM will not allow continued spark or fuel. Showing these pictures to my customer and explaining what caused the "no start" gave me a sense of satisfaction rivaled only to that of showing off pictures of my granddaughters.
Jeff Bach is the owner of CRT Auto Electronics, an ASA-member shop in Batavia, Ohio. For more information on this topic, contact Bach at (515) 732-3965. His e-mail address is firstname.lastname@example.org and his Web site is www.currentprobe.com.
AutoInc. Web Site |
ASA Web Site |
Senate Judiciary Committee Reviews Insurance Antitrust Repeal |
Counterfeit Auto Parts: A Growing Industry Epidemic |
Update on Collision Repair Parts |
Listening to Customers Makes You More Money |
Guest Editorial |
Tech to Tech |
Tech Tips |
News Briefs |
Taking the Hill |
Around ASA |
Shop Profile |
Net Worth |
Stat Corner |