Every once in a while I get a Diagnostic Dilemma that takes me back to the basics of how modern engine management systems work.
This month’s happened to be a 1992 Jeep Wrangler equipped with the 4.0L engine and manual transmission.
Of course, a vehicle of this vintage and configuration has to be about as basic as modern technology can get, right? Well, there are always exceptions.
I’ve been getting a number of diagnostic cases on ’90s Wranglers because, with a minimum of on-board electronics, they’re very adaptable for off-roading.
Many enthusiasts install lift kits, air-locking differentials, winches, skid plates, higher numerical ratio axle gears and other off-road paraphernalia. Stroker kits also are available to increase the displacement of the 4.0L.
The cylinder head also can be modified to increase compression and breathing for high-altitude off-roading.
The fuel-injected 4.0L is also ideal for traversing high mountain trails because it compensates for major changes in altitude and doesn’t starve for fuel or flood out on steep inclines.
As with many other cases, this month’s “Diagnostic Dilemma” was referred to me by a client shop.
The owner had just installed a 4.0L engine modified by a very reputable local engine rebuilder.
To his dismay, he found that, although he installed the engine “according to the book,” it would crank, but not start. I assumed that there was more to this problem than meets the eye, and accepted the assignment from my client shop.
Base-Lining the Engine
When dealing with a new engine that won’t start, it’s important to “baseline” all of the essential functions.
In years past, I’ve had product problems like a brand-name timing chain set coming out of its box with the timing marks incorrectly drilled into the cam sprocket and a new camshaft with the sprocket locating pin drilled in the wrong position.
Stuff happens and, if you’ve rebuilt many engines, you understand what I’m talking about.
Because many Chrysler-based engine management systems are sensitive to battery voltage, my first step was to recharge and test the battery.
With that out of the way, I found that the engine would crank and fire the cylinders, but fall just short of actually starting, which was the symptom described by the owner.
Although I verified the rotor position in relation to the #1 terminal in the distributor cap, I knew that I would probably have to re-visit this issue due to some peculiarities in Jeep cam and crankshaft position sensor synchronization.
I also verified TDC by attaching a good vacuum gauge to #1 cylinder with a compression gauge hose that had the Schrader valve removed.
Being new, the engine was hard to rotate by hand, so I used a remote starter switch to bump the engine past TDC.
If TDC on the harmonic balancer is correct, the vacuum gauge will rapidly show pressure as the piston approaches TDC.
The pressure will rapidly drop to zero at TDC and rapidly pull vacuum as the piston descends on the power stroke.
This method isn’t perfectly accurate, but it will show within a few degrees whether the harmonic balance ring has slipped on its hub.
Last, I spot-checked compression pressure on #1 cylinder to verify that the valves were seating and that basic camshaft timing was correct.
See Photo 1.
The next step was to verify spark.
Because I had a scan tool attached, I went to Chrysler’s Automatic Test Mode (ATM) and “buzzed” the coil to check spark output.
While I was at it, I also activated #1 fuel injector. I couldn’t really hear the fuel injector activate, so I used a fuel pressure gauge and fuel injector testing tool to verify that the injector was injecting fuel into the intake port by watching the fuel pressure drop as the injector was activated.
All through the above testing, I verified the presence of spark at the spark plug and, with the aid of a ’noid light, verified the presence of fuel injector pulse.
But I did notice during the spark test that the engine would sometimes produce an irregular spark.
See Photo 2.
A few months earlier, I had another ’92 Wrangler that produced the same kind of erratic CKP pulse during cranking.
The owner had reversed polarity on a pair of jumper cables and perhaps over-voltaged the system, which ruined the ECM and evidently damaged the CKP sensor.
In the current case, I attached my lab scope to the CKP harness and discovered a very irregular CKP waveform while cranking the engine.
Here again, I can’t make assumptions on new engine installations because the wiring in the CKP might have been broken during removal.
Prior to installing a new CKP sensor, I used a remote mirror to inspect the flywheel timing shutters as I bumped the flywheel around with a remote starter switch.
To my dismay, I discovered that the engine balancing shop had drilled two 3/8” balance holes about 5/16” into the edge of the flywheel next to the series of four timing notches machined into the flywheel.
This inspection turned out to be the key to solving this “Diagnostic Dilemma.”
Although the new CKP sensor brought the engine to life and all seemed well for the moment, I did notice a surging idle condition.
Cleaning the throttle plate and idle air control valve offered no remedy.
If you’re like me, you do more scan tool diagnostics nowadays than you do lab scope diagnostics.
While I can’t spend much time discussing lab scope technique, I do want to emphasize that it’s important to understand the capabilities of your scope when testing high-frequency signals from crankshaft position sensors.
Most of the time, I use a lab scope to verify simple things like sensor inputs and actuator outputs. With that said, many techs become quite involved with scope testing various technical issues including cam/crank synchronization.
Many post some very excellent waveform cases on the International Automotive Technician’s Network’s (iATN) waveform library.
If you’re into scope testing, it’s always a good idea to store (or have access to) known-good waveforms, especially those dealing with cam/crank sync issues.
You are required to be an iATN member to have this access, but the wealth of information contained in the forums and waveform library is well worth the money spent.
In this case, I accessed some known-good waveforms from the iATN waveform library.
The Jeep 4.0L flywheel has three sets of four timing notches located at 120° intervals on the flywheel.
Each set of four notches creates four magnetic pulses in the CKP sensor to indicate each cylinder position.
The camshaft position (CMP) sensor located in the distributor generates a single large square wave indicating where cylinder #1 compression stroke begins.
Groups of four pulses from cylinders #4, 1 and 5 are grouped together when the 5-volt reference is pulled down to zero.
Cylinders #3, 6 and 2 are grouped when the reference rises to 5 volts.
Notice the irregular display in Photo 2.
This could be caused by a bad CKP or it could be caused by a bad setup on the scope.
Because some scopes don’t have the sampling rate of others, the 200 millisecond (ms) time base on this capture might be too long.
A phenomenon called “aliasing” can occur, which provides a faulty waveform pattern, such as the one in Photo 3.
Because idle speed generates a higher frequency than cranking speed, the scope’s time base must be reduced.
In this case, I reduced the time base to 100 ms for idle conditions. See Photo 4.
This waveform shows that the four pulses representing cylinder #3 aren’t within the 5-volt reference, indicating that the CMP isn’t synchronized with the CKP.
That part I already knew. But surprisingly, I counted five pulses on cylinder #2 instead of the expected four.
So, did this waveform indicate that the CKP was reading the balancing drill-outs next to the cylinders #2 and #5 timing notches? And, if it did, what would be the effects on spark timing for those cylinders?
Going back, I advanced the distributor rotor one tooth clockwise to sync the CMP with the CKP.
After re-syncing the distributor, the CMP/CKP synch was displayed correctly.
But the question remained as to the effect that the extra pulse was having on #2 and #5 cylinder spark timing.
See Photo 5.
After thinking a while about devising a method of displaying those effects for the benefit of the vehicle owner and the engine machinist, I attached my “ancient” Snap-on 1665 scope to the ignition and produced this display (Photo 6) of dwell time at 1,610 rpm:
Notice that, at 1,610 rpm, cylinders #2 and #5 produce nearly double the dwell of the remaining cylinders.
During a subsequent test-drive, the Jeep exhibited a rolling idle speed and began to jerk and misfire at engine speeds over 3,000 rpm.
The likely explanation is that, when the dwell angle reaches 60°, the primary circuit is no longer being opened by the coil driver located in the PCM.
Because the circuit must be opened to produce a spark, cylinders #2 and #5 are no longer firing or are firing erratically. In the bar graph display, cylinders #2 and #5 are producing about 46-48° dwell at 1,610 rpm.
It doesn’t take a great leap of the imagination to believe that another 12-14° of dwell are being added at 3,000 rpm.
When total dwell angle exceeds available dwell angle, the ignition quits making sparks.
Duty Cycle and Dwell Angle
Modern techs usually think in terms of duty cycles and on-times.
This photo (Photo 7) of my Fluke 88v DVOM displaying min/max on-times shows the coil primary being turned on a minimum of 45.33% and a maximum of 77.00%.
Veteran mechanics think in terms of dwell angle.
Dwell angle is simply one revolution of the distributor (360°) divided by the number of cylinders (6), which equals 60° available dwell for each cylinder.
When on-time exceeds 50% on a single-coil ignition system, the coil begins to overheat due to excessive cycling. So the on-time or coil saturation time on this Jeep is kept down to about 12° dwell angle or 20% on-time (12° divided by 60° equals 20%) at idle speed.
Dwell angle increases as the engine accelerates to increase coil saturation time for maximum spark output.
Examining the Flywheel
A close look at the flywheel explains how the CKP sensor was reading the drilled balance holes. See Photos 8 and 9.
It easy for veteran mechanics to understand the relationship between dwell angle and spark timing.
As the rubbing block on the ignition contact points wears down, the dwell angle or on-time is increased. In other words, the points close earlier and open later, which retards base ignition timing.
So, most master mechanics adjusted new points with slightly more air gap and a few degrees more base timing than specified.
As the rubbing block seated into the distributor cam, the dwell angle and base ignition timing came into
Similarly, the extra CKP pulse on cylinders #2 and #5 evidently increased coil on-time on those cylinders.
Since the drilled balance holes were located at the leading edge of cylinder #2 and #5 timing notches, the PCM turned the coil on early.
The PCM turned the coil off as designed when the last timing notch passed by.
The difference between the leading and trailing edges evidently caused the on-time of the coil to increase to the point that the coil would no longer turn off long enough to create a good spark. At least that’s my theory and, until somebody proves otherwise, I’ll probably stick to it…