Ignition diagnostics you will actually use: Part 1
Truglia is the owner of Car Clinic, a state-of-the-art repair facility in Mahopac, N.Y. He is ASE A6 certified with a M.A. from Columbia University. (Vehicles diagnosed and waveforms captured by G. “Jerry” Truglia, Kevin Quinlan and Adam Varney.)
Years ago, vehicles required annual tune-ups. Because there was so much money to be made, shops invested tens of thousands of dollars in “big box scopes” to diagnose all of the ignition problems that were common in the day.
Now, let’s fast forward 20 years. Tune-ups are once every 120,000 miles on many vehicles. Many repair shops don’t even use a lab scope. Suffice it to say, ignition diagnosis is becoming a lost art.
Ignition diagnostics is not that complicated. However, we want to give you techniques that you can use now to diagnose cars quickly and make money in the process. We’ll start simple and work our way up. This month we will cover ignition diagnostics without getting into interpreting the waveforms. We’ll cover that next time.
Quick ignition diagnosis strategies
Trick #1: The rubber hose trick [Figure 1]
What does it do? Identifies which cylinder has the ignition-related misfire.
How do I do it? Place a short piece of rubber vacuum hose between the coil pack tower and the spark plug wire. This allows for easy cylinder shorting to identify a weak or dead cylinder. Use a grounded probe or test light to attract the spark through the hose. Never divert the spark for more than three seconds at a time to avoid damaging the catalytic converter.
CAUTION: Never touch the hose (if you do, it will be a very shocking experience).
Trick #2: Getting coil-on-plug (COP) secondary waveforms [Figure 2]
What does it do? Allows you to diagnose if a misfire is plug or coil related.
How do I do it? Install an ignition jumper wire that you can make up using a spare spark plug wire. Simply hook up to the spark plug wire you put in series between the COP and the plug like the way you are used to, and you have your secondary waveform!
If you find out it is the spark plug to blame, make sure that the OE spark plug was installed. Do not get brainwashed by the parts store voodoo that you can use different metal or brand plugs. Just stick with what works. Also keep an eye on whether the plugs were torqued properly and the right amount of anti-seize is used, since globs of it can cause misfires. If you want to get really picky, seemingly good spark plugs might have different internal resistances. If you do an ohms check on each plug, do the comparison game. The one that sticks out like a sore thumb is causing the problem.
Trick #3: Quick ICM check
What does it do? Checks if the ignition module is to blame.
How do I do it? Try removing an ignition coil (one coil, two coils or all three), and connect a headlight. Use the alligator clips from the headlight and connect them to the coil pins (- and +).
This will load the circuit and make sure there is enough power and ground to fire a coil. Do not use a test light since there is not enough of a load. The test light may flash but that does not mean the coil will fire. The headlight requires 6-plus amps, while the test light only requires 3 to 4 milliamps to light.
This makeshift headlight tool has multiple uses. Use the headlight in the picture with a flasher to find a short in a vehicle. Just this time connect the alligator clips to the blown fuse terminals and get yourself a low current inductive amp meter or a compass to locate the short.
Trick #4: The ol’ spray bottle with baking soda
What does it do? Finds out if old spark plug wires are bad.
How do I do it? Try using some baking soda and water mixture (not corrosive like salt) to locate voltage leaks to ground. You can use a grounded screwdriver or a test light while spraying the mixture to be extra thorough, running the screwdriver or test light down the wire.
Make sure that both the screwdriver and test light have a good ground. Don’t forget to keep your fingers off the metal of the test tools.
Trick #5: Know the capabilities of your scan tool
Did you know that you can use a scan tool to locate a misfire?
The GM Tech 2 utilizes a misfire graph.
The Ford IDS has a Cylinder Power Balance Test that helps isolate misfires.
The Ford IDS and most scan tools fitted with generic OBD II displays misfire data using Mode 6. Mode 6 often has misfire count for each specific cylinder.
In this example, the OTC Genisys scan tool displays Mode 6 Ignition data from a Chrysler vehicle.
The Mode 6 example below is from a Generic/Global scan tool from ATS (Automotive Test Solutions). It highlights misfires in red and translates the Mode 6 into language we understand.
The ATS scan tool decodes the hexadecimal data and displays Mode 6 data into useful information that we can read and understand. Not all scan tools do. If the data is not decoded you will need to access the OE’s website to look it up.
Trick #6: Be acquainted with the common pattern failures
A couple examples of this stick out in my mind. Ford ignition coils up into the late 2000s are problematic. If one is bad, you legitimately have a good reason to just go ahead and change all of them. Use quality products. I have found BWD and Wells ignition package coils are robust. I prefer using these over the OE Ford ignition coils.
Another example: In the mid-2000s, four-cylinder Nissan Altimas’ sporadic no starts are caused by defective CKP and CMP sensors. The OEM made them out of plastic instead of metal, and they just crack and take oil in.
CMP and CKP diagnosis
These days it’s not as simple as the CKP controlling just spark and the CMP controlling just fuel. On some vehicles one or both sensors are responsible to the PCM to determine spark timing. So, if we want to diagnose ignition problems on today’s vehicles, we are going to need to know how to diagnose CMP and CKP sensors.
Whether you’re working on a distributor or distributorless ignition, camshaft and crankshaft sensors fall into one of three general types, depending on how they create engine speed and position sensor signals: Magnetic Inductive (AC voltage generators), Hall Effect, and Photo-Optic.
That’s it. Three different types. If you can learn how to test these three you can test all crankshaft and camshaft sensors.
You do not need to learn a new procedure for each make and model on the road. Keep a couple basics in mind.
The crankshaft rotates exactly two times when the camshaft rotates once. However, this will not always be reflected when comparing the two waveforms. Let’s take the three different sensors in order.
Magnetic Inductive Type
These sensors contain a magnet and generate an alternating current (AC) voltage. Used by most OEMs to measure the speed of various components, these sensors do not require an external power source to operate. The example below shows a GM crank sensor signal [Figure 3].
The sensor simply counts the notches on a tone ring.
If the ring is damaged or is a cheap aftermarket unit, this can throw off ignition timing and cause major driveability issues.
Use a scope to test these sensors with the engine running or cranking. Be prepared to adjust your scope setting of time and voltage depending on cranking or running RPMs. Any obvious defects in the waveform indicates and obvious problem. Sometimes, it helps to use a waveform library, like on iATN, to compare your waveform to a known good one. Also, if you are looking at an intermittent problem, simply scope Cam and Crank and record it in “movie mode.” The signal that drops out is killing the ignition.
Hall Effect Type (and Photo-Optic sensors)
Hall Effect and the old Photo-Optic sensors use DC voltage and each require an external ground and power supply. These sensors are as easy to test as 1, 2, 3.
1. Test at the sensor or module to verify power and ground.
2. Check for open-circuit reference voltage at the sensor or module.
3. Test for signal pulses using a logic probe or scope with the engine running or as it is cranked. When in doubt, consult a waveform library to see how cam and crank should match up.
Always check for codes or pending codes that may be set by a failed or failing crank or camshaft sensor.
Some engines won’t start without both sensors. Some will start with one sensor missing, but will run in a fail-safe mode.
2000 Chevy 2500 5.7L no-start
In this example, we get a van from a do-it-yourselfer who has installed his own engine. The van ran fine for months and then it barely made it home one day. Since then, the van has refused to start. The monster-sized van made itself nice and cozy in our bay, and decided it wanted to stay for a while.
We started with the basics. First, we tried starting in order to hear how it sounded. It just continued cranking, but refused to fire up. TSBs were checked. The battery was charged. Fuel pressure was at 63 psi and the fuel pump turned at 9.5 amps. The engine had spark. It had RPM (and a new CAM and CKP sensor.) This narrowed it down to a mechanical issue with the engine.
So, we decided to investigate whether or not there was a timing issue. Looking at cranking compression using a pressure transducer, everything mechanically seemed to be good including the timing. We also connected to the ignition coil and the number 1 injector. The injector and coil appeared to be firing correctly, but the ignition coil seemed to have a “double tap.” It appeared to pre-ignite before the main spark.
Engine compression (green) appeared to peak in time with fuel (red). Spark (blue) appeared to fire not always at the right time, as you can see in the circle.
So, we revisited the possibility that there was an ignition issue. The distributor assembly appeared to be installed incorrectly, so we called up the parts store and replaced it. However, we turned the key and the new one did not make a difference, so we returned it. To be honest, we did this with the cap and rotor, too. None of the parts made a difference, even though the old ones appeared to be in poor shape from all the fruitless starting attempts.
So, we had to explore other avenues that were not timing related, because the cranking compression appeared to reflect good timing and we could not find any problem getting spark.
We decided to check relative compression, and it was uneven. Relative compression is not as accurate as a real compression test so we pulled out the only plug that offered relatively easy access in order to perform a standard compression test and look at spark plug condition. Compression was good but the plug was fouled with fuel.
We then checked the engine oil and saw that it was flooded with gasoline. We confirmed this by lighting the dipstick on fire. The basics had to be covered first, because the cylinders were washed down. So, we changed the oil and the plugs, and used STP oil treatment in the cylinders to build up compression.
Sometimes, doing things like checking the oil right off the bat can save you a lot of time.
With confidence we turned the key again. The engine tried to start up now, but it now made a horrible metallic scream. It sounded like a bad flywheel, but we inspected the whole component and it checked out.
The engine then stopped starting because all this testing wore down the starter. So, we replaced it and the engine did exactly what it was doing before, screaming away and not starting.
In desperation, we looked at TSBs again and found one that dealt with “No, Hard, or Slow Start...Grinding or Unusual Noises During Crank.”
How did we miss this? We slapped in a new Delphi CKP sensor, hoping it would save the day. Again, no dice.
So, we were right back to square one. The only thing we found strange during the entire time was that ignition double tap. So, we double checked the CKP and CAM sensors with a scope and found no problems whatsoever with the sensors. Then, as luck would have it, we found a speck on the crankshaft’s reluctor/tone wheel. Was the timing being thrown off the whole time because of a tiny stupid spec being picked up by the CKP? Maybe it is the source of the double tap!
Here’s why: The CKP sensor counts the parts on the reluctor wheel that pop out in order to determine timing. If there is a speck in between, it might throw of its count. But, it could not be that easy. After cleaning it off, it would not start, sounding just like it did before.
Using help from a mobile tech friend of ours we were able to pinpoint our problem [Figure 4].
Here’s the Cam-Crank correlation on our truck. Notice the CAM waveform (yellow) beginning just after the CKP waveform (purple) [Figure 5]. Here’s a known good Cam-Crank correlation on the same van. Notice how the CAM waveform (red) starts exactly at the CKP waveform (yellow).
There was indeed a timing problem! It was time to inspect the timing chain.
When we took the timing cover off, guess what just flopped off without us exerting any force on it whatsoever? The crank reluctor/tone wheel!
We knew it was installed improperly on the remanufactured engine. So, we bought a new one from the dealer and was surprised to see that the part was much different. They did not line up at all.
As you can see in the photo, we are lining them up properly. The parts were obviously machined differently.
What is the moral of the story? Sometimes you can do everything right, check the Technical Service Bulletins, and do all the basic stuff; but sometimes you have to dig real deep.
Known good waveforms are very important, and there are some very good resources out there.
In Part II we will cover some more advanced material including real-world ignition tests, interpreting primary and secondary ignition waveforms, and some need-to-know theories on the modern ignition system you will actually run into.