Trying to diagnose a minor glitch in engine performance, especially when there are no Diagnostic Trouble Codes (DTCs) to provide some type of guidance, is one of the tougher challenges a drivability tech faces. On the opposite end of the spectrum is the no start complaint. This should be one of the easiest to figure out, shouldn’t it? Yet for some, this is as tough a troubleshooting chore as any other. And it all usually centers on two things: lack of system knowledge and lack of a logical process.
Ready to improve both?
Open Up The Funnel
A common mistake is to jump to premature conclusions as to what the cause of a problem might be. These assumptions often result in a case of tunnel vision that prevents even the best of techs from reaching a successful conclusion. Ever find yourself under the hood, scratching your head and pondering what possibly could be wrong, only to have a co-worker come over and spot the problem in seconds? “Can’t see the forest for the trees,” is a good saying to keep in mind when starting off your diagnostic quest.
To avoid this trap, initially consider every potential cause as suspect. With perhaps dozens (if not more) of possibilities, it becomes evident that you need a process that will help you sort through them all in as efficient a manner as possible. That leads me to the next common mistake I want to discuss: testing without purpose.
Any test you make on a vehicle system should be performed with an end result in mind. The first task at hand is to thin that list of suspects as much as you can, and that calls for a general procedure that covers a lot of territory in one, easy as possible to perform, test. You might follow your first general test with a second or even third to reduce that list of suspects to a more manageable size. Once done, you then can focus in on each of the remaining possibilities using more precise, or pinpoint, tests to isolate the root cause of the problem you’re chasing down.
Instead, many of us tend to dance around. We perform a general test, then a pinpoint test and then we back up with yet another general test. We test whatever comes to mind first, hoping we’ll stumble on the answer. Often, the result is that we pass right by it without even realizing it.
Let’s apply all this to a typical no start complaint.
What Kind Of No Start?
In my reality, no start complaints usually arrive at the shop on the back of a tow truck or already are waiting in the lot having been towed in while the shop was closed. It certainly isn’t unusual to see nothing more than the words “No start” written on the work order and not much else. At this point in the game, every possibility under the sun is suspect.
A no start condition easily can be broken down into two general types, though. An engine that won’t turn over for whatever reason is one type, while an engine that cranks but will not run on its own is another. I can perform a quick, easy general test to isolate which variant I’m dealing with.
First, I turn the key to Start and try to start the car.
In this case, the engine’s starter motor engages normally and the engine spins over at a normal cranking speed but does not run or even make the attempt to start running.
Troubleshooting is not a matter of only science and measurement. Every sense you have plays a role. Can you hear the starter motor engage? Do you smell fuel? Do you see any reading on the tachometer as the engine is spinning? Do you feel the engine spinning over normally? These are only a few examples of the information you should be taking note of while you are verifying this customer’s complaint. Are you actively aware, though, of these things when you’re faced with a no start diag?
The key turn is the first general test and noting what happens, in as much detail as you can, is the collection of the test results. After you have the results, it’s time to apply them logically to the list of suspects and narrow the field. From there, you can decide if another general test is needed to trim the selections even further or if you can move to a pinpoint test to pin down the culprit.
Is it necessary to test the battery or starting system, based on our first general test? Of course not. The engine engaged and turned over at a normal speed. Cranking it over is not the issue. Yet some of you do exactly that, burning time and getting no closer to the cause.
Do I suspect any mechanical issues with the engine? Sounds like a reasonable question. If the engine were binding it may not turn over at all, but I would have heard that in the tone of the engine as I tried to crank it over. Perhaps there is a loss of compression that is preventing the engine from starting. After all, an engine needs three things to run: compression, fuel and spark. But again, wouldn’t I hear that in the tone of the engine while cranking?
And who said “antitheft”? Possible, but doesn’t that usually allow a car to start and run for a few seconds before dying and/or indicate that the system is activated with a little flashing light on the dash?
OK, I heard that. Somewhere way in the back. Who asked if there was an rpm reading on the tachometer while the engine was spinning?
Moving Onward
As it turns out, there was no movement of the needle on the tachometer while the engine was spinning. On many vehicles, the loss of the Crankshaft Position Sensor (CKP) can prevent the vehicle from starting. Because this same sensor often is the source of any rpm signal used by the instrument cluster, the lack of needle movement could indicate a problem in the CKP or its circuit.
Would you perform a general test or a pinpoint test next? I’m voting for a pinpoint test next, using a scan tool to focus on that suspect CKP.
Specifically, I’m going to see if there are any codes stored related to the CKP sensor. But I’m not going to be surprised if there aren’t any. It is not unusual for the Engine Control Module (ECM) to miss a crank sensor failure. Do you know why? Depending on the programmed testing strategy (which you can learn more about by reading up on Theory and Operation in your service information system and/or by reading the conditions/criterion for setting a CKP failure code), when the CKP sensor signal is lost, the ECM simply thinks the engine has shut down.
I’m also going to take a look at live data with a particular interest in what the CKP Parameter Identifer (PID) is reporting. While this signal might be the source of the rpm reading at the instrument cluster, that signal could be taking the long way there through other modules on the network. I want to know specifically what information, or lack of information, the ECM is receiving.
Accessing codes first, I do indeed find a CKP circuit fault code stored in the ECM’s memory. In addition to the CKP code, though, are similar codes for the Camshaft Position Sensor (CMP), Throttle Position Sensor (TPS) and Barometric Pressure Sensor (BARO), all defined simply as “circuit fault.” Not the test results I expected, but with this information it wasn’t necessary to continue any further with the scan tool.
Do you agree?
A Change In Direction
When I began this test, my purpose was to verify my suspicions that the CKP sensor was the culprit. But the presence of additional sensor codes was a test result that altered that original hypothesis. The CKP sensor still was on my list of potential causes, but now I had one more to add. One, I thought, would shortly prove to be the root cause of the problem. One that one more pinpoint test would verify.
With this many similar codes stored, isn’t it logical to consider what they have in common with one another? At this point, you might find yourself returning to the service information and doing a little homework. Pulling up the Engine Performance wiring diagram, it was easy to see that all of these sensors had a common denominator.
(The vehicle in question is a 2005 Chrysler Pacifica, by the way, with a 3.5-liter engine. If you don’t know what I found yet, I encourage…no, I challenge you to pull up this same schematic and see if you can find what I did before you finish reading this story.)
All of these sensors share a 5-volt power supply fed to them by the ECM. This is their “Reference” voltage and without it, no signal could be sent back to the ECM. The pinpoint test I next needed to perform was to see if the five volts were present. Because the TPS sensor was the easiest to access, I backprobed the appropriate wire in the TPS connector and measured the voltage present with my Digital Multimeter (DMM).
I read 0.31 volt.
Now I know what the root cause is. No reference voltage to the sensors critical to the operation of the ignition and fuel injection systems. The remaining question to repairing the problem is figuring out why. Is the ECM at fault, not supplying the voltage, or is something on the 5-volt line shorting it out downstream?
Typically, when a loss of reference voltage is experienced, a shorted sensor on the 5-volt line is the cause. This also is easy to find. Simply monitor the reference voltage while you unplug the sensors sharing the line. How do I know that? I read the Theory and Operation of the system I am troubleshooting. I read the Technical Service Bulletins (TSBs) that might be related to the problem I’m facing. I attend training when I can, read the trade magazines, follow the forums on sites like iATN.
I continue to educate myself and grow my skills.
In this case, the first to be unplugged was the TPS sensor location I started with. No change. The next easiest to access was the BARO sensor. Unplugging that sensor immediately restored the 5-volt reading I was looking for on my meter. Because I didn’t need the BARO sensor to start the engine, a quick turn of the key followed by the immediate start of the V-6 powerplant verified the problem was found.
Recap
This story was based on an actual vehicle that I assisted a coworker with a few years ago. He spent more than a few hours trying to figure out what was causing this relatively simple no start complaint. But he did it following a not so uncommon process that I think many of you may be familiar with. He relied on the “silver bullet” diagnostic method.
He pulled the codes and found the same data I did. But his first general test was a search of the Identifix database. There, he found multiple hits on failed CKP sensors shorting out the 5-volt line as well as a few that listed the ECM as the culprit. The CKP was cheaper, so he threw that at the car first. When that didn’t work, he went to the boss to order the ECM. Instead, the boss asked me for a second opinion. The rest you know.
Don’t get me wrong. I value resources like Identifix, iATN and others as informational sources. But I abhor their use (as I’m sure they do) as a source of “fast fixes” without confirmation testing that you do, indeed, have the same problem as those 341 (or whatever the number happens to be) other “confirmed fixes.” Unfortunately, there is no record of how many repairs are made and become, as this one would have, a confirmed waste of time and money.
Regardless of how simple the problem, or how obvious the solution, might appear to be, save the silver bullets for hunting the Boogey Man. Professional technicians test, they don’t guess.
Subscribe to Motor Age and receive articles like this every month…absolutely free. Click here
