SomeBuckaroo Posted February 6, 2023 Share Posted February 6, 2023 Moses, I finally got around to 'scoping up my TJ for further diagnosis. I've attached pictures of various traces. A few notes: CH1: Cyl 1 Spark Plug Wire (via capacitive probe) CH2: Spark Coil Wire (via capacitive probe) CH3: MAP Sensor Signal My capacitive probes are cheap and I think the probe on CH1 (Cyl 1 plug wire) may be degraded. It works well enough to trigger the 'scope, and the probe on CH2 seems to do a fine job picking up the coil spark train. OR: is this degradation actually evidence of an issue? I've replaced the cap & rotor several times with no change to my symptoms. I should have swapped the probes to see if the weak signal followed the probe. Also note the noise on CH3 (MAP sensor signal), seemingly uncorrelated with injectors or spark plug firing. Is that amount of noise normal? I wonder from where it's coming, how much is present on the other sensor lines, and what effect it has. I wish I had reference signal traces of a known-good 4.0L. Regardless, it does appear that the MAP sensor has sufficient bandwidth to capture manifold vacuum fluctuations within the firing order. Pretty cool! I've tried to choose filenames that describe the activity being shown. I'm happy to do additional rounds of scope work, adjust the time-base, delay the trigger to capture each cylinder individually, etc. I do know that features of spark waveforms can be used for troubleshooting, but I am not at all versed in that area. Let me know if any of this helps! Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted February 8, 2023 Author Share Posted February 8, 2023 After re-reading my post, let me add some questions and clarifications: 1.) The MAP sensor "noise" to which I refer are the large spikes with a period of approx 4.5ms. Maybe this is from the alternator's PWM field regulation? It's very regular when present, which looks like hysteretic voltage regulator action to me. 2.) The coil secondary waveform on CH2 seems to randomly vary in average value. This is especially evident when comparing the flat portion of the waveform between firings, which does not maintain a consistent level. And the transition from spark firing to the flat inter-spark region is usually abrupt, but sometimes has a curve to it. I'm hard-pressed to believe this is instrumentation/measurement error. Any idea what's going on there? 3.) The MAP sensor waveform represents intake valve action for the cylinder currently in the intake / compression stroke, not the cylinder currently firing (according to the spark secondary waveform above it). This took me a while to realize! The firing order is 153624, so (correct me if I'm wrong) e.g. when the spark waveform shows #1 firing, the MAP waveform directly below shows the end of #5 intake stroke (each successive cylinder is 120 degrees behind the previous). Just wanted to point this out for future readers who are nose-deep in engine sensors for the first time like I am. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted February 8, 2023 Administrators Share Posted February 8, 2023 SomeBuckaroo...Pleased you're working with the oscilloscope! Glad you clarified, I was about to answer your original comments referring to the photos, and the additional information is helpful...The offset "correlation" could be due to a simple factor: the spark plug firing lines reflect spark in timing advance mode while your MAP probe or indexing channel could be picking up either the voltage (+) or ground (-) from either the camshaft or crankshaft position sensor. 2000-up Coil-On-Plug engines actually share the 5V (+) camshaft position sensor lead with the MAP sensor. The '98 wiring system may be picking up a similar reference or at least feedback. As a D.C. system, completing the ground or reading the positive supply voltage will show as an "open" unless the +/- circuit is complete. While your PWM/hysteretic theory is certainly a consideration, it's not likely at play here. (A voltage reading of that sort would not be a consistent, repeated single spike with uniform height.) The MAP line is consistent and appears at the same timed interval with uniform voltage. It also is in relative sync with the #1 cylinder firing line. To verify what I'm suggesting, confirm that the MAP voltage line is consistently just a few degrees from the spark firing line for #1 cylinder. This would reflect the normal spark timing advance with the engine at idle. You can even verify the amount/degrees with your timing light: break down the offset degrees in your scope lines, they should match with the number of degrees of spark timing advance seen with the timing light. The crankshaft position sensor indicates a true and constant TDC for #1 cylinder, as this signal comes from the flywheel. Unless the camshaft is painstakingly "degree'd" during engine assembly, the camshaft position will deviate slightly ("retarded") due to timing chain slack/wear. A camshaft position sensor signal in a distributor engine like yours will also reflect the distributor drive gear wear. The camshaft position sensor is monitoring the degrees of valve timing error between the crankshaft and the camshaft. When the cylinders' spark firing lines parade with the engine running, these firing lines are X-degrees of advance from TDC. You can alter the engine speed to get a shift in spark timing advance (confirmed with your timing light) then see whether that shows up on the scope as the number of degrees of separation between the MAP probe line and the #1 cylinder's spark firing line. This will test the theory...Keep in mind that the camshaft position is 1/2 the crankshaft degrees, so this could even break down whether the MAP channel probe is picking up the crank position sensor signal or the camshaft position signal. Your timing light reads crankshaft degrees while the camshaft position sensor is picking up camshaft degrees. The crankshaft rotates twice (720 degrees) for each 360-degree rotation of the camshaft. To your point, if this theory has the MAP channel reflecting the crankshaft or camshaft position as opposed to the #1 spark firing position, the spark line adjacent to the MAP probe spike is number one cylinder firing. If the parade is reading correctly, you can move to the right to pick up the other cylinders in the 1-5-3-6-2-4 firing order. To confirm, remove a spark plug lead (with either insulated spark lead pliers or the engine shut off!) and see whether that cylinder drops out of the firing sequence. I see some firing symptoms worth discussing. Before doing so, if it's possible to flip your voltage spikes to the upside instead of facing downward on the scope, I can share what the firing lines represent. I'd like to see an image of the six cylinders in parade plus a stretched view to a clear single cylinder (#1 cylinder, preferably). I'll explain what you're seeing. Despite modern ignitions and PCM spark timing management, the firing line and other features are similar to scope patterns we read on breaker point ignition systems over fifty years ago. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted February 9, 2023 Author Share Posted February 9, 2023 I think I understand your theory on the MAP spikes, and using timing advance is a very clever way to verify and potentially single out the interference to cam or crank sensor! I did have time this afternoon to invert the channels and grab a parade plus each cylinder individually. My scope has enough memory to capture and store the entire parade and "zoom" into each cylinder at high resolution (as opposed to setting the trigger delay and "manually" capturing each cylinder over the course of several minutes). I figure a one-shot capture is important since it means external conditions remain relatively constant in the ~100ms it took to capture the parade. I've attached all images here. The image files are named for the cylinder # (e.g. cyl_5.jpg is the 2nd event in the parade). Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted February 9, 2023 Administrators Share Posted February 9, 2023 SomeBuckaroo...Interesting patterns...I would drop the voltage baseline to expose the full height of each spark plug firing spike. The aim is to compare spike heights and voltage accordingly. Height is not only about ignition condition, it also reflects relative cylinder compression, a rich or lean burn condition, cylinder fill/vacuum, effects of spark timing and the overall condition of your running engine. With EFI/MPI (individual port fuel injection), we have the ability to assess and compare the rich/lean burn condition of each cylinder. This can help clarify how each fuel injector functions. Carbureted engines did not have that advantage. We need to read the spark spikes and their voltage heights to make these observations. Have you tested or changed your spark plug wires or the coil high tension lead between the coil and center of the distributor cap? Aside from firing line height, I'm concerned about these ignition components. The cap and rotor are "good"? The cap has brass contacts? Run a standard ohms resistance test on the individual spark wire leads (end to end) and the ignition coil high tension lead (spark cable from the coil to the distributor cap). Twist or coil the wires gently while running these tests...Also run an ohms resistance test on each spark plug (out of the engine and isolated) for comparisons. Spark plug quality is all over the board these days, even when new. What kind of spark plugs are you running? I'm seeing distinct ignition pattern differences between cylinders. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted February 10, 2023 Author Share Posted February 10, 2023 Hey Moses, I've attached the shots showing the entire spark firing spike heights. However, watching them in real-time, they are not consistent at all per-cylinder. They jump all over the place. I took a short video to show what's going on:https://youtu.be/JIem4c3dJRo Over the past few years I've gone through several sets of spark plugs, wires, rotors, and caps. I've even changed all at the same time. I never noticed a difference in the symptoms. I have also measured spark plug cable resistance in the past and found nothing out of the ordinary against recommendations online (I can't recall how many ohms per inch of spark plug cable is acceptable, but I did look it up at the time). However, of course I will follow your advice and re-check everything, probably this weekend. I believe I'm running the usual Autolite or Champion copper spark plugs at this time; I'll pull each and verify. Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted February 10, 2023 Author Share Posted February 10, 2023 I pulled the distributor cap, rotor, and spark plug cables. I've attached pics of all - straight off the jeep, no cleaning. The plug cable for cyl #4 is different from the rest, it's a blue Belden cable. The rest are no-name black cables. (I damaged #4 a while back and replaced with what I had laying around). I measured the resistance (and length) of the spark plug cables. As you suggested, I twisted/pulled on them while measuring. In all cases the resistance fluctuated by several K-ohms, but never more than that (nor did any go open-circuit). Here's the table of my measurements: Cable | Resistance | Length (inch) | kohm/foot Coil | 4.1k | 10.5 | 4.7 #1 | 5.5k | 12.5 | 5.3 #2 | 6.4k | 15.0 | 5.1 #3 | 4.4k | 9.5 | 5.6 #4 | 3.6k | 9.0 | 4.8 #5 | 6.2k | 14.5 | 5.1 #6 | 5.5k | 12.0 | 5.5 Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted February 12, 2023 Administrators Share Posted February 12, 2023 SomeBuckaroo...This is the oscilloscope vantage you needed. Great scope settings and photos! The spark firing lines look reasonably even. Regarding the spark firing lines within the firing order, there is some variation. To make sense of whether this is an acceptable range, please share the voltage readings for peak spike height per cylinder in the firing order. Firing heights and resistance/voltage do not seem grossly different per cylinder, but we're keeping your rough idle symptom in mind. What is the lowest versus the highest voltage reading per cylinder? Also, blip the throttle and watch where the spark firing lines go. Watch this carefully and note the height of the firing spikes. See whether they remain somewhat uniform height. Keep the valve fluctuation in mind (i.e. compression and combustion efficiency as reflected in the spark firing lines). See whether an engine speed change reveals signs of poor combustion. Resistance (spark lines) should go up with the snap throttle, but do the firing lines rise uniformly? You will need to lower your screen view to see these changing spark line heights. The spark firing lines are high kilo-voltage readings. Set the scope for accurate and clear readings at this voltage. You'll be viewing six-cylinders simultaneously in parade. This viewing should align with your leakdown and compression checks. Your spark cables do meet OEM/FSM resistance range (3,000-12,000 range ohms per foot). From what I see, there's no reason to suspect a fuel injector issue, which you verified with the fuel trim readings. The oscilloscope has provided another dimension to the overall tune "picture" in real time running state. Very valuable. As for the cap and rotor, they're a mess and need replacement. If these pieces are still apart, I'd also like to see the spark plug tips with the plugs laid out in their cylinder order. Measure the spark plug gaps and note the oscilloscope spark line and coil oscillation view for each spark plug in the firing order. This is the kind of insight a scope provides. Regarding how to "read" the spark firing lines, this is a tradition look at how an ignition functions: 1) The high spike is the spark plug firing; 2) the relatively straight or up-sweeping jagged line is the spark firing duration and dissipation; 3) the squiggly extended line is the ignition coil oscillation and 4) the long, flattening line is the "dwell" period for when the coil builds voltage for the next cylinder to fire. There is a slight irregularity, though not an "Ah, ha!" moment, in your latest scope reading. The firing line height and duration of the spark can be impacted by the spark plug condition and the spark plug gaps. Wider or narrower spark plug gaps will affect the height of the firing line (reflecting the degree of resistance). A lean or rich fuel mixture will also affect the height of the spark plug firing lines. This reflects the combustion efficiency, which can suffer from fuel injector issues, a dirty or sticky idle air control valve, an erratic Throttle Position Sensor, a clogged exhaust or cat, or abnormal fuel trimming. I don't sense a glaring mechanical issue like intake/exhaust valves unseated or not sealing. From the oscilloscope vantage, we're only seeing combustion efficiency, read as spark resistance, i.e. the height of the spark plug firing lines. This screen view is not a mechanical assessment, yet it provides an insightful look at the running engine's ability to provide compression and normal combustion. The obvious tell all with the oscilloscope would be running an in-cylinder pressure transducer test of the running engine. You would test the in-cylinder pressure, preferably at each cylinder, and compare the results, looking for telltale scope pattern differences between cylinders. This would be the oscilloscope test for accurate valve timing events and valve seating. On that note, in-cylinder pressure transducers can be expensive. Pico and others make them. Your scope would likely work with any transducer, which simply converts pressure changes into voltage readings. The in-cylinder transducer must have a fast read/ms rate and enough accuracy to pick up the four-stroke phases of a running engine. A far less expensive alternative is a tailpipe pulse pressure test with a pulse tester. I bought the CRUZ pulse sensor at eBay, and it works: https://www.ebay.com/itm/184415845508 At the eBay sales page, the British Columbia seller provides a photo of the device reading a running engine from the tailpipe. You'll see how useful this pressure pulse sensor can be for determining exhaust pressure pulses. (Another scope channel can pick up #1 cylinder firing as a marker for identifying the pulse waves by cylinder firing order.) With your concern for valve seating or spring issues, this would be a vital tool and alternative to the more expensive in-cylinder pressure transducer. I red highlighted (above) some tune and idle related items to consider. The idle air control valve and its throttle body port get clogged over time. This can be an inexpensive service item at Amazon. (Read reviews. I installed an off-shore IAC valve on the XJ 4.0L that has held up so far. Buyer beware, though, we get what we pay for.) TPS is another duty cycle item that affects idle smoothness. At your TJ's mileage, I would replace the idle air control valve, clean the port in the throttle body and consider the age/mileage on the TPS. NTK makes a quality TPS at a reasonable price: https://www.amazon.com/gp/product/B07F48C2RS/ The scope is formidable and less expensive than a high end scan tool. The two tools in tandem make troubleshooting and diagnostics simpler and far more accurate. If I were to consider one tool over the other, for versatility and pinpoint testing (independent of the PCM/ECU/ECM data stream), the oscilloscope would be my first choice. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted February 12, 2023 Author Share Posted February 12, 2023 Thank you for the feedback on the cap & rotor - I will replace those today, along with the spark plugs. Is there a certain plug you recommend for the distributor-equipped Jeep 4.0L? I pulled the plugs, measured gaps, and took the pictures below. The plugs are all NGK ZFR5N (my earlier post was mistaken!). My 0.035" feeler had a consistent fit slight drag in each. In the past I have replaced the TPS, cleaned the IAC housing in the manifold, and more recently replaced the IAC itself (all Mopar parts). I will order the TPS you linked above and replace when it comes in. I did recently observe its signal on my scope over its range of travel and did not see any indications of wear to the internal potentiometer (e.g. signal discontinuities), but I certainly do have experience with those issues being sneaky. Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted February 13, 2023 Author Share Posted February 13, 2023 I captured a video of the entire spark parade during several throttle snaps:https://youtu.be/y4Rm8htZ660 I took rough measurements of firing spike heights. My Hantek HT25 probes attenuate by 10,000:1; I configured the scope such that each major vertical division represents 5,000 volts. I tried to set a cursor (horizontal dashed-line) to be helpful, but realized after the fact that it was not placed quite right. Regardless, here's measurements from several frames of the above video: Spark Firing Spikes (Kilovolts) "Idle 1" and "Idle 2" are measurements taken during idle "Snap 1" through "Snap 3" are measurements taken during brief throttle snaps "Max diff" is the maximum difference between lowest & highest cylinder (KV) Cylinder | Idle 1 | Idle 2 | Snap 1 | Snap 2 | Snap 3 | 1 | 18 | 21 | 16 | 16 | 16 | 5 | 17 | 17 | 15 | 15 | 16 | 3 | 17 | 17 | 14 | 13 | 21 | 6 | 18 | 28 | 14 | 15 | 15 | 2 | 16 | 24 | 15 | 17 | 17 | 4 | 21 | 22 | 15 | 15 | 15 | Max diff | 5 | 11 | 2 | 4 | 6 | Generally, my observations are that the spikes are significantly lower during a throttle snap than at idle. They also appear to be somewhat more uniform in height during the throttle snap than at idle. I can't really discern any characteristics of each particular cylinder - the spikes seemed to all vary a lot regardless of cylinder. I had considerable difficulty reliably triggering on Cyl 1, so I am considering using the camshaft position sensor signal (CMP) to trigger since it ought to be less noisy than the capacative pick-up of Cyl 1 plug wire. I will first capture CMP and Cyl 1 spark to establish & document the timing relationship. For additional info, I hooked up my timing light - and found idle timing "off-scale" BTDC: the timing mark was further BTDC than the extent of the scale on the timing cover, and was often nearly obscured by the fan pulley. The timing also jumped around considerably. I verified with my OBDII scanner that the ECU indeed reported approx 18* BTDC at idle (so the harmonic balancer hasn't significantly slipped on its rubber isolation ring). Is this timing advance and erratic behavior expected? I took a video showing the timing mark's behavior at idle (I used a red paint-pen on the crank pulley scribe, and I marked up the video with a white arrow pointing to the action):https://youtube.com/shorts/xYFxykaLYQw Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted February 14, 2023 Administrators Share Posted February 14, 2023 SomeBuckaroo...I really like the spark plug condition and burn! NGK is among my favorites, I run NGK on all my dirt Honda XR motorcycles and most often in the 4.0L Jeep six. While the bikes (carbureted with a wide altitude range of riding) do better with iridium style plugs, that's not a necessity for your 4.0L Jeep application. The current V-Power replacement plugs are doing a great job. There is some wear on the center electrode tips but unless you want to replace these plugs due to mileage, they should still be firing okay for now. (Check and compare resistance with your ohmmeter if concerned.) Likewise, the TPS on order could be held as a backup until the Mopar TPS gets symptomatic, a matter of time. For guaranteed idle stability, the TPS and IAC valve need to perform consistently...Your call. As for your deep dive into firing line voltages and the snap throttle experiment, see my comments below in red italics...Nice work here, I like your scope adjustments and attenuation in order to see the full spark line heights. 4 hours ago, SomeBuckaroo said: I captured a video of the entire spark parade during several throttle snaps:https://youtu.be/y4Rm8htZ660 Watched the video several times, very helpful. Did you change the cap and rotor before this test cycle? There's nothing extraordinary or grossly troubling. The spike heights are normal, created as load increases and cylinder pressures rise with the quick/snap throttle opening. As the throttle closes, the lines drop below the idle baseline then fluctuate as the idle restores, all of which is normal. This is a reduction in engine load/cylinder pressure plus the fuel injector, fuel trim and ignition timing functions (essentially the EFI "fuel-and-spark" engine management)—again, nothing alarming. On older carbureted engines, the closing throttle and immediately after the snap would show enrichment (lower spark lines) due to the "venturi effect" within the carburetor. With venturi effect, while the closed throttle does not call for added fuel, it gets fuel anyway. Aside from no load and dropping cylinder pressures, the richer mixture ratio fires easily with less voltage resistance, and this reduces the amount of spark voltage needed. The spark firing lines are shorter. By contrast, EFI can cut off fuel as soon as the throttle shuts, eliminating richer mixtures during deceleration and throttle closure. I took rough measurements of firing spike heights. My Hantek HT25 probes attenuate by 10,000:1; I configured the scope such that each major vertical division represents 5,000 volts. I tried to set a cursor (horizontal dashed-line) to be helpful, but realized after the fact that it was not placed quite right. Regardless, here's measurements from several frames of the above video: Spark Firing Spikes (Kilovolts) "Idle 1" and "Idle 2" are measurements taken during idle "Snap 1" through "Snap 3" are measurements taken during brief throttle snaps "Max diff" is the maximum difference between lowest & highest cylinder (KV) Cylinder | Idle 1 | Idle 2 | Snap 1 | Snap 2 | Snap 3 | 1 | 18 | 21 | 16 | 16 | 16 | 5 | 17 | 17 | 15 | 15 | 16 | 3 | 17 | 17 | 14 | 13 | 21 | 6 | 18 | 28 | 14 | 15 | 15 | 2 | 16 | 24 | 15 | 17 | 17 | 4 | 21 | 22 | 15 | 15 | 15 | Max diff | 5 | 11 | 2 | 4 | 6 | Your attenuation and visible patterns are well done...The most reliable read of the snap throttle was your first one. The time between the snaps was too short; the engine management needed to stabilize. If you monitor fuel trim with your scan tool at the same time, you'll see what I'm suggesting. Allow the engine idle to completely stabilize between each snap test. As for your results, I see this differently. Watch closely: The snap causes an instant increase in the voltage demanded; the lines increase in height at the snap. This is due to cylinder pressures ramping up and the added firing voltage needed to provide adequate spark. Generally, my observations are that the spikes are significantly lower during a throttle snap than at idle. They also appear to be somewhat more uniform in height during the throttle snap than at idle. I can't really discern any characteristics of each particular cylinder - the spikes seemed to all vary a lot regardless of cylinder. The other way around: Snap raises the lines. You're catching the rebound as the throttle closes, which drops the spark firing voltage below the normal idle voltage. At an idle, the system quickly restores to the idle spark voltage. Can you stretch out the scope pattern to reveal a wider view as the spikes occur? Or can you stretch out time in your video software edit (create "slow motion") while keeping the audio active and in sync with the stretched video track? The audio will sound slow and drawn out, but it will align with the spark firing line positions. This might be easier to follow. I know what to expect at the snap, so it's easier for me to discern that the lines go up. To be clear, the spark or firing resistance goes up during the snap. Therefore you see the lines momentarily shoot upward. The drop in firing line heights when you unload the throttle reflects either lower firing resistance or momentary fuel enrichment—or both. Review your video or play with this more. In general, a lean mixture will raise the spark firing lines because a lean mix requires more voltage to fire. A relatively richer mixture fires more easily. The baseline for optimal gasoline combustion (stoichiometric) is 14:1 Air/Fuel. For emissions purposes, modern EFI engines are tuned to run as lean as possible except when accelerating (wide open throttle or WOT) or when under load. For adequate performance, a balance gets struck between spark retard to prevent detonation and optimal fuel mixtures for meeting tailpipe emissions requirements. To see this at work, read your lines with the engine first idling cold (enrichment cycle) in O2 sensor open loop mode and below 140-degrees F coolant temperature. Spark firing lines may be lower than when the engine is warm and in closed loop. This would reflect a slightly richer mix or what a "choke" once did on carbureted engines. Since you have baseline "normal" idle spark firing lines established, see whether this holds true. The wild card is that cold spark plugs and cylinders create resistance to combustion, so spark voltage requirements and firing lines may go up. Try this, anyway. I had considerable difficulty reliably triggering on Cyl 1, so I am considering using the camshaft position sensor signal (CMP) to trigger since it ought to be less noisy than the capacative pick-up of Cyl 1 plug wire. I will first capture CMP and Cyl 1 spark to establish & document the timing relationship. That's a good idea. You'll be at a fixed #1 cylinder timing line that aligns with the distributor shaft. This should be relatively close to TDC at the crankshaft. The engine is nearly new with a new timing chain and sprockets. For additional info, I hooked up my timing light - and found idle timing "off-scale" BTDC: the timing mark was further BTDC than the extent of the scale on the timing cover, and was often nearly obscured by the fan pulley. The timing also jumped around considerably. I verified with my OBDII scanner that the ECU indeed reported approx 18* BTDC at idle (so the harmonic balancer hasn't significantly slipped on its rubber isolation ring). Is this timing advance and erratic behavior expected? I took a video showing the timing mark's behavior at idle (I used a red paint-pen on the crank pulley scribe, and I marked up the video with a white arrow pointing to the action):https://youtube.com/shorts/xYFxykaLYQw This is all very normal for electronic fuel and spark management. If your crankshaft position sensor, MAP, O2 and other sensors are feeding information correctly, the engine is attempting to minimize emissions while in an unloaded condition. Advanced timing raises manifold vacuum...Under normal conditions, with the engine loaded in Drive with an automatic transmission, the idle timing advance would change. Overall, the PCM has spark timing curve algorithms the are constantly changing with sensor input. The spark timing curves are unlike the fixed curves in conventional distributors. Our 1999 4.0L sixes do not have a knock sensor and rely on sensor-fed information and fuel-and-spark management programming to prevent detonation/ping. Our distributors set in a fixed position. (There is no provision for "adjusting the base timing".) The distributor housing and rotor/shaft each index in fixed locations. The PCM gets a #1 cylinder TDC reference from the crankshaft position sensor. All timing adjustments and changes take place electronically. (There are no centrifugal or vacuum distributor advance mechanisms!) The rotor and cap contacts feed spark to the plug wires while the degrees of advance and retard get assigned by the PCM...2000-up Coil-On-Plug engines eliminate the distributor altogether. Spark is triggered directly by the PCM to a coil at each cylinder. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 5, 2023 Author Share Posted March 5, 2023 Moses, apologies for the delayed response and thank you for the detailed information! The spark plugs, cap, and rotor were indeed brand new for that round of tests - I should have noted that in the post. Your explanation of spark firing heights makes sense, and I see how I was watching the post-snap trace rather than during the snap. I ended up ordering the pulse sensor, as well as an in-cylinder pressure transducer ("Rotkee" brand, from Ukraine), and some new capacative probes for the spark plug wires (I'm having difficulties with my current cheap Hantek set). I'm looking forward to another round of diagnostics when those new tools arrive. I suspect I'll be able to capture an instance of my brief "rough idle" symptom with the pulse transducer, which will allow a visual comparison of that behavior to adjacent periods of "smooth" idle. I'll post the new diagnostics here, hopefully in a week or two. Thanks again! Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted March 6, 2023 Administrators Share Posted March 6, 2023 Exciting, SomeBuckaroo!...Pleased that you purchased the in-cylinder pressure transducer and reliable probes. We could see the spark cable/high tension cable noise with the Hantek cables. For an accurate picture of cylinder pressure(s) and how they relate to a rough idle, you can 1) back probe the Crankshaft Position Sensor to set a true TDC for #1 cylinder and 2) probe the #1 spark plug cable to show #1 cylinder firing and identify which stroke the piston is on. If you parade the firing order on one channel and the in-cylinder transducer on another, you can look for a relationship (if any) between the idle roughness and the test cylinder's running pressure changes and compression. Since your scope has four channels, you can probe the crankshaft position sensor on one channel, clamp #1 spark cable on a second channel, use the in-cylinder transducer on a third channel and use the fourth channel to probe the camshaft position sensor. (Select and set up the channels as you want them to display on the scope.) Variations between the crankshaft position and camshaft position will indicate any timing chain slack. The camshaft position probe will also help match up the camshaft degrees with the cylinder pulses. We can discuss camshaft/valve timing, lobe wear and lifter issues if you see fluctuations between the tested cylinders. The camshaft, lifters and chain are new. There should not be an issue here. Since the #1 cylinder (or whichever cylinder you test) will not have ignition with the in-cylinder transducer in place, you are strictly looking at fluctuations in real time pressures within the cylinder. Since the cylinder is not firing, there will be no combustion effect or analysis. You are isolating and testing the cylinder's pressure changes, not the combustion process. For a look at ignition function while you test cylinder pressures, a spark plug simulator can test the spark cable and ignition function with the engine running. This will demonstrate ignition performance while you run the in-cylinder pressure test. It will also prevent damage to the ignition...Example of a spark plug simulator: https://www.amazon.com/Stens-750-018-Ignition-Designed-exclusively/dp/B001OK6MIO/ You may discover mechanically caused roughness (like poorly seating valves) as you test each cylinder. If the valves are not seating or there is irregular piston ring seal, you will see fluctuating, unstable pressure readings. In any case, while you test a cylinder, the 4.0L inline six engine is running on five cylinders. The purpose of the test is to discover abnormal pressure pulses and determine peak (ready to fire) pressure in the cylinder. I would run the test at each cylinder, record the findings or save screen shots. Compare the cylinder readings. The highest pressure cylinder is often your benchmark. The in-cylinder compression test will illustrate variations in intake or exhaust stroke pulses, show the compression pressure rise on the compression stroke, and pinpoint the maximum pressure or any pressure fluctuations in the live cylinder. At normal idle speed (which the IAC should establish even on five cylinders), you will be observing 720 crankshaft degrees (two rotations of the crankshaft) to complete the 4-stroke phases. The crankshaft will be turning 650-700 rpm. Set your time interval readings to provide a clear, six-cylinder firing parade across the screen. Then zoom into a view of just the test cylinder. If pressure pulses match per cylinder and compression is nearly equal and sufficient per cylinder, with no signs of valve leaks or wide differences in peak compression, the roughness is likely tune, O2, MAP, IAC, fuel trim, injector, vacuum or air leak (intake or exhaust) related. We're speculating for now. Your pinpoint test tools are coming! Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 11, 2023 Author Share Posted March 11, 2023 The pulse sensor came in (but the in-cylinder transducer is still in transit), and I had more fun with the oscilloscope today. I probed and am triggering on the camshaft sensor which is far more stable than triggering on #1 spark wire. I did include a channel for #1 spark for ease of correlation (and the other spark events are just barely visible on that channel as well, which helps even more). For sensor orientation, I verified that decreasing pressure (sucking on the hose!) produced a decreasing voltage waveform on the scope. I connected the pulse sensor to the intake manifold via some vinyl hose I had laying around. Video of intake manifold measurements: The scope traces are: camshaft sensor (the big square wave), #1 spark (superimposed on the camshaft trace), and then of course the pulse sensor. Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 11, 2023 Author Share Posted March 11, 2023 ... and a video of the same pulse sensor connected to the tailpipe (I just shoved the vinyl hose up the tailpipe several inches and clamped it in place with minimal force). https://youtu.be/m-DCnoyy2hI Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 11, 2023 Author Share Posted March 11, 2023 My thoughts on the manifold vacuum video: It's clear that the same "basic waveform pattern" would be expected to show up 6 times per firing order, and indeed it does here. How the valvetrain actions across the 6 cylinders contribute to that basic waveform is not immediately clear to me; I'd have to sit down and diagram it out. I'll use cylinder spark events (as seen on the 'scope trace) only as a reference "marker" for manifold waveform features - I don't mean to imply the "sparking" cylinder is the cause. There are definitely consistent differences in the "basic waveform pattern" as it is repeated through the firing order. This is certainly caused by consistent differences in airflow via valvetrain operation. I see two instances of clear negative "double-peaks" (just after #1 spark and #3 spark) which I don't clearly see in the other 4 patterns. Probably more interesting is an _inconsistent pattern_ in the positive-going peak just _prior_ to #6 spark - it seems that a double-peak pattern _comes and goes_ as the video progresses. That whole pattern has less defined features compared to the other cycles. Patterns that _come and go_ are a red flag! The overall height/amplitude of the sensor readings do vary somewhat over the course of the video - as does the idle, which can be seen in the varying length of the camshaft square wave cycles. I don't know how "rock steady" a 4.0 in good condition should idle. I suspect the variations I'm seeing are downstream effects of some valvetrain issue as visible in the manifold waveform around #6 cylinder spark. (My thoughts on the exhaust measurement video are forthcoming ...) Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted March 12, 2023 Administrators Share Posted March 12, 2023 SomeBuckaroo...Let's start by considering your test patterns. It's important to separate the roles of the camshaft position sensor (CMP) and crankshaft position sensor (CKP). The camshaft position sensor, a hall effect signal, is a sync signal generator. The distributor shaft is gear-driven by the camshaft, and the hall effect pickup (sync signal generator) is attached to the plate below the rotor. Here are FSM explanations for the roles that the crankshaft position sensor (CKP) and camshaft position sensor (CMP) each play: Above explains the location and signal from the camshaft position sensor... Above is the Jeep 4.0L (similar to 2.5L) distributor type ignition's camshaft position sensor (the plate below the rotor). Note that the curved pulse ring in the distributor housing is 180-degrees. It is attached to the distributor drive shaft, which is driven by the camshaft. Above explains the crankshaft position sensor (CKP) role...Shown at right is the 2.5L flywheel with two four-pulse groups. The flywheel or flexplate attaches to the crankshaft in a fixed position. Here is your 4.0L crankshaft position sensor (CKP) with three groups of four pulses (a hall effect like the camshaft sensor). The flywheel rotates and the CKP sensor sends a signal to the PCM. The crankshaft position sensor mounts at approximately 11 o'clock when viewed from the transmission. So, your camshaft sensor's scope pattern is the pulse signal generated to help time the fuel injectors. By contrast, the crankshaft position sensor (CKP) is a consistent signal to the PCM that pinpoints TDC for #1 piston. The square waves (shown on your scope overriding the spark firing lines) is the camshaft position sensor signal with its 180-degree pulses. (For correct spark and injector timing, it is necessary to install the ignition distributor exactly like described in the FSM.) The camshaft signal may be confusing. This signal is not intended for referencing the #1 cylinder's piston at TDC. The TDC signal comes from the crankshaft position sensor. The camshaft sensor feeds information to the PCM for sync'ing the injector firing points at the cylinders. The PCM determines spark timing with data from the CKP, CMP, O2, MAP and other sensors. Note: It's easier to get the #1 cylinder firing trigger from the #1 spark plug lead, using your capacitive/induction clamp attached to the #1 spark plug wire. We're using this trigger to identify the firing order. Clamping to #1 spark wire provides a rough TDC reference with #1 piston near its firing position. Base timing is not fixed or constant with electronic fuel-and-spark management. (There is negligible deviation in valve timing or the distributor shaft position with the engine running.) What you see as the #1 cylinder spark firing line represents the degrees of spark timing advance. The amount of advance is a PCM function. Spark timing reflects PCM algorithms based on the engine's sensor feedback. With EFI management, we usually see 12-14 degrees spark advance in an unloaded, idling emissions engine. That's close enough to get a firing order reference. Actual piston location would need a CKP or CMP pickup. CKP would be slightly more accurate than the CMP. If you probe the crankshaft position sensor for a #1 cylinder TDC reference point, the 4.0L (six-cylinder) flywheel generates three groups of pulses for each revolution of the flywheel. The crankshaft rotates twice for each 4-stroke cycle of the engine. (The #1 piston reaches TDC at the top of its compression and exhaust strokes. It takes 720 degrees of crankshaft rotation to complete the 4-stroke cycle.) The crankshaft is a steady rotational reference with the flywheel rigidly attached to the crankshaft. Despite the accuracy in finding "true" TDC, it's easier to get a #1 spark lead reference to determine the firing order. Footnote: For ignition diagnostics on a distributor engine, we have always used #1 spark lead for the firing order reference and probed the coil high tension (secondary) lead to parade firing lines for all cylinders. (Primary ignition patterns, dwell angle and breaker point condition are a separate signal on breaker point era scopes.) Traditionally scopes have an induction clamp for #1 spark wire and a clamp for the cap-to-coil high tension secondary lead. The secondary coil lead provides the full parade of cylinders while the #1 spark lead triggers the firing order. The two scope methods for running a "relative compression test" are 1) the starter motor cranking draw and 2) the exhaust pulse sensor readings with the engine running. Your pressure pulse and vacuum readings are sharp and clear. Yes, there is some irregularity at both the intake and exhaust pulses per cylinder, which could explain the rough idle. However, to once again separate mechanical from fuel/spark related causes, a cranking (starter motor) amp clamp test would be helpful. We're close but still not completely there... A cranking relative compression test is quick and easy to run. You can begin by either attaching a remote start switch at the starter motor (keeping the ignition/key switch in the OFF position) or removing the fuel pump fuse or relay. Either approach will prevent the engine from firing. However, I still crank the engine for a moment when using the key switch with the fuel pump fuse or relay removed. This will eliminate a hiccup or hot false start due to pressurized fuel still in the EFI/MPI fuel rail. Something like the Innova 3630 remote starter switch at AutoZone ($8) will do the job. (There are more robust remote start switches available if you intend to use the remote switch regularly.) The 4.0L Jeep starter relay at the starter motor is easy to access for attaching a remote start switch. The remote start switch closes the Jeep 4.0L starter motor's relay and engages the starter motor directly at the starter. Again, be sure to keep the ignition key switch in the OFF position while you attach and use the remote start switch for this test. You want to crank the engine without starting it. The amp clamp will still pick up the starter motor draw from the battery cable. Clamp around the battery cable running to the starter motor. (Since this is D.C., you can pick up this signal with the clamp around the negative battery ground cable if preferred.) By removing the fuel pump fuse or relay, you can turn ON the ignition switch with the scope's capacitive induction clamp hooked to the #1 spark cable. The induction probe clamp at #1 cylinder spark lead can serve as the trigger pickup for identifying #1 cylinder in the firing order. Caution: Automotive scope capacitive induction pickup probes are intended for secondary voltages. However, as a rule of thumb for any directly applied high voltages, use an attenuator if necessary to protect your scope! The cranking voltage changes from the amp clamp on the battery cable will indicate the amperage draw for each of the six cylinders. Any variance in a cylinder(s) is the relative compression difference. A higher amperage draw reflects higher compression (more piston resistance on the compression stroke) in that cylinder. Lower amperage draw is a weaker cylinder. Simultaneously, with your four-channel scope you should be able to get a read of intake or exhaust pulses during cranking. (The engine/flywheel is rotating at 180 or so rpm while cranking, which should be plenty.) There should be adequate intake air flow. Consider wedging open the throttle during this test to assure air flow volume. For more about this relative compression test and the amp clamp, check out the first three minutes and twenty seconds of the tune-up video I did for the magazine: https://4wdmechanix.com/jeep-4-0l-ignition-tune-up-and-injector-cleaning/ An accurate pattern occurs as the starter current stabilizes. With the battery fully charged, I usually run a cranking compression test cycle for 5-10 seconds (maximum) and record with the scope. You'll see from my video that there is initial voltage irregularity as the starter motor engages the starter ring and overcomes inertia. Simultaneously capture intake or exhaust pulse readings while cranking. The tailpipe pulses will not reflect combustion but instead the piston and valve movement that creates vacuum and pressure changes during the intake and exhaust strokes. With a #1 cylinder reference, you can follow the 1-5-3-6-2-4 firing order and align the parading cylinders with the pulse sensor readings. If you can get distinct intake vacuum and tailpipe pressure readings while using your pulse sensor at cranking speed, it would be valuable to align and compare the pulse sensor readings with the starter draw's "relative compression" pattern. We'll look at your findings. Below is an introduction to the Autel scope with the accessories kit. At 5:08 to 5:58 minutes, I show the amp clamps that are included with the Autel MP408 Oscilloscope Accessories Kit (OAK). You will recognize the clamps as similar to Hantek: Also, here's what I did with my pressure pulse sensor. I improvised a cone to somewhat "channel" the exhaust pulse inside the tailpipe. I'm not clear whether this improved my pulse readings. Your readings look good. Perhaps the cone is unnecessary. It's worth experimenting: From what I see in your patterns, there is intake idle vacuum and tailpipe pressure irregularity. The test of relative cranking compression simultaneous with vacuum or pulse tests would help pinpoint a mechanical problem (poor valve seating, compression blow-by, etc.). If relative compression is rock solid and uniform while these vacuum and pressure pulses act erratically, there could be a fuel/spark or vacuum leak issue. (Is the exhaust manifold/header leak free? The slightest leak can dilute the O2 reading and create a fuel trim issue. These manifolds are notorious for cracking and warping.) Of course, to eliminate any questions about mechanical issues, your in-cylinder pressure transducer kit will be the ultimate diagnostic tool. Excited to see your findings! For now, should you discover uneven relative compression that aligns with these vacuum and exhaust pulse fluctuations, use your pulse sensor at the dipstick tube. Engine idling, any pressure fluctuations in the crankcase would indicate a mechanical issue, most often compression loss from piston ring blow-by. This is how we separate valve seepage and valvetrain issues from ring blow-by when running these tests. The dipstick tube pressure pulse can also be a follow-up for a low compression reading with the in-cylinder pressure transducer. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 19, 2023 Author Share Posted March 19, 2023 Hey Moses, I finally got an OTC 600A AC/DC clamp from Amazon. I scoped the intake manifold using the _pulse_ sensor, along with the starter current and cyl #1 spark (to identify TDC) - while cranking (not running). I disconnected all 6 fuel injector harnesses to ensure the engine would not fire, and held the throttle wide-open. Attached are screenshots of the waveforms during cranking. I included 3 different "zoom" levels: the full capture containing a little over 2 complete firing orders, a close-up of a single firing order, and a close-up of just cyl #1 power stroke. Top waveform is cyl #1 spark (I cut this off to preserve screen space), middle is intake manifold, and the bottom is starting current. I am disappointed with the amount of noise on the amp-clamp - it's a very "fuzzy" trace. Is that level of noise to be expected or is something possibly wrong? Each cylinder's compression via starter current appears to be within about 10% when measuring _peak to trough_. I believe trough is the correct baseline for this measurement, as it (approximately) represents the current required to turn the engine over with no compression action. Cylinders 1, 2, and 4 appear to be slightly lower than 3, 5, 6. As for the intake manifold waveform: it occurred to me that with the throttle wide-open, substantially less pressure variation would be seen than if the throttle were closed. I could repeat this test with the throttle closed, if that would help. Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 19, 2023 Author Share Posted March 19, 2023 I also got the package in from Rotkee containing their in-cylinder transducer. To test the transducer, I measured the in-cylinder pressure for each cylinder, during cranking, with the throttle wide open. I did not include any other waveforms. Attached are pics - The cylinders look ok, with the highest-lowest variation being about 10%. The transducer measures absolute pressure, at about 50psi per volt output. Atmospheric pressure reads about 300mV. My cylinders all peaked near 3.2 volts, or roughly 150psi (+/-) after subtracting atmospheric pressure. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted March 20, 2023 Administrators Share Posted March 20, 2023 SomeBuckaroo...This is an exciting example of your oscilloscope being used to test the engine's mechanical condition. You did a stellar job and got useful results for diagnostics. The amp clamp starter current wave form is not abnormal or exceptionally noisy. This is a higher amperage draw and may have to do with our clamps. While a low pass filter could help, this is not critical. Peak voltage heights are more important, which you clearly captured. Below is a screen shot of the Autel MP408 displaying the 4.0L Jeep starter draw with my Autel (Hantek) 650A clamp. The pattern, especially if I were to stretch it out, looks much like yours: Your initial test of relative compression is revealing. Though relative, this is true cranking compression. Your earlier standard compression gauge and leakdown tests provided a baseline for using #1 cylinder as a reference. (You can do simple math to approximate the compression on the other cylinders.) The in-cylinder pressure transducer that you use later is also real cranking psi...You went from using a conventional compression tool to making the same observations with the transducer. This has made your oscilloscope that much more versatile. Note: It is possible to get a "normal" compression gauge reading with a cylinder that has weak or broken valve springs. The Schrader valve in the typical compression gauge allows the gauge to "pump up" to peak compression and hold it—even when several crankshaft rotations may have much lower compression. By contrast, your new in-cylinder pressure transducer should not have a Schrader valve. The transducer will show pressure changes with each rotation of the crankshaft. The in-cylinder transducer should only be used with a hollow hose or adapter tool that does not have a Schrader valve! The repeat pattern is similar, which rules out a fluke chance that the starter motor's condition is not up to par (bad sector, field coils, bushing drag or other issues). As for drooping compression, what stands out for me is #1 cylinder and #4 cylinder in the 1-5-3-6-2-4 firing order. There's nothing earthshaking yet a clear drop at #1 cylinder. #4 is slight; #2 is negligible. Compression variance does not account for a "rough idle" condition. Yes, you are correct that the intake vacuum reading should be different (higher negative pressure or vacuum) with the throttle closed. Let's look at the intake negative pressure with the throttle closed through a cranking test. Again, I purposely suggest the cranking test because the engine is not running. Intake pulses will be strictly a reflection of the engine's mechanical functions. What I do see is fluctuation in the intake pulsing, which also makes it worth testing with the throttle closed. Without any interference from ignition/combustion and a power stroke, the intake pulses do not show equal vacuum per cylinder. Look approximately 3/4ths time interval between cylinder compression peaks. This is the intake cycle when air/fuel is drawn into the cylinder, followed by the intake valve closing and the compression stroke. By cylinder firing order, the vacuum is not equal per cylinder. If this persists with the throttle closed, I'll share possible causes for vacuum loss per cylinder. Let's compare the closed throttle cranking results first. The in-cylinder transducer results look really good! Again, your in-cylinder transducer test does show #1 and #4 cylinders with a slight drop in compression. Let's see what this looks like with the engine running, testing one cylinder at a time front to rear. Cranking tests are different than readings with the engine running. Presumably, you ran the cranking in-cylinder transducer tests with all spark plugs removed. Since this is just isolating and testing individual cylinder pressures at a slower cranking rpm, you eliminate factors like reciprocating mass imbalance. (Mechanical imbalance would have to be extreme and obvious to affect cylinder pressure readings.) Unless the crankshaft, damper or flywheel/clutch are way out of balance, engine reciprocating parts imbalance is not likely with an inline six-cylinder engine. Footnote: An inline six is an optimal design for harmonic balance. Imbalance phasing issues would not be a factor unless the crankshaft, flywheel/clutch, crankshaft pilot bearing, damper, flexplate or torque converter are defective. In general, harmonic vibration (with the engine running) would raise questions like whether the engine has its original damper and flywheel/flexplate, clutch cover and disk imbalance or whether the flywheel has been resurfaced. Overall, the crankshaft and its reciprocating mass must be in balance. For the in-cylinder pressure transducer test with the engine running, the Jeep 4.0L engine will run on five cylinders as you move through the test at each cylinder. At each cylinder, test the engine running at idle speed and a snap throttle. This will show both idle operating pressures and the pressure changes under a brief compression load. Recall the snap throttle impact on your spark plug firing lines earlier? Watch for the corollary here. You must be excited about the diagnostics capability these tools provide... Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 24, 2023 Author Share Posted March 24, 2023 Real quick - here's some pics of the Rotkee hardware for those interested. Everything seems to be of decent quality, especially given the reasonable prices. I did have to run the compression adapter's threads through a die in order to get them to easily thread into my spark plug holes. The transducer itself does have standard spark-plug M14x1.25 threads; no adapters necessary. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted March 24, 2023 Administrators Share Posted March 24, 2023 SomeBuckaroo...Looks "industrial strength". Interesting that the transducer threads into the cylinder head/spark plug hole. Some questions: 1) Any instructions or comments about engine (cylinder head) temperature when testing with the transducer? Does the transducer look rugged enough to handle some cylinder head casting heat? Most transducers have a temperature operating range. Of course, there's no combustion in the cylinder being tested, so the concern would be the cylinder head temperature. 2) The transducer looks stout. The barrel's diameter should fit most spark plug reaches with a 14mm spark plug thread size. Is the "brake hose" with fittings an adapter for threading into the spark plug hole and remotely mounting the transducer? That would make sense for applications with narrow spark plug access like a "hemi" cylinder head. 3) Does the company offer adapters for other spark plug thread sizes? Many motorcycles and older Ford engines have odd thread sizes. This would not be a deal breaker. There are adapters available to mate 14mm thread size compression gauges to these odd size spark plug threads. The adapters could be used with these tools. 4) The test connectors do include standard BNC for hookup to common scopes? Electrical connectors look sturdy. They didn't skimp on product. You got a box full! Good to support companies trying to build affordable, quality tools and equipment. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 25, 2023 Author Share Posted March 25, 2023 1.) Rotkee's (English) spec sheet for their "PS16 in-cylinder transducer" is here:https://store.rotkee.com/en/ps16-in-cylinder-pressure-transducer-for-petrol-engine.html Surprisingly, they don't include a temperature range. However, all exposed area of the device is metal and it certainly looks like it could withstand 100C - 150C cylinder head temperature. 2.) The specs page above includes dimensions for the transducer, but I've included a screenshot here: Yes, the "brake hose" adapter is for remotely mounting the sensor. I used it on my initial test of my 4.0L - it made the job much easier, and I don't know if I could have measured Cyl #1 without it - the A/C compressor in my "early" TJ 4.0L makes that spark plug difficult. Since the sensor itself has male M14x1.25 spark-plug threads, the Rotkee brake hose adapter is essentially "universal" in that it would work with other devices designed to thread into M14x1.25 spark plug holes. 3.) Yes, Rotkee sells M12 and M14x1.25 Tapered Seat adapters. Maybe others too.https://store.rotkee.com/en/ad-m14-m12-adapter-for-in-cylinder-pressure-transducer.htmlhttps://store.rotkee.com/en/ad-m14-ts-adapter-for-in-cylinder-pressure-transducer.html They also offer rigid extenders:https://store.rotkee.com/en/ad-m14-rigid-compression-tube-for-in-cylinder-pressure-transducer.html 4.) Yes, all device-side connectors are BNC. The BNC connectors are a somewhat snug fit on my 'scope - which is better than a loose fit! But a little difficult. At the risk of sounding like I'm a salesman for Rotkee (I'm not), they have a really nice & affordable line of all kinds of automotive transducers, sensors, and cables. This is a link to their English product page - I recommend you check it out. The only downsides: 1.) Shipping to USA took 1 month (for me), and 2.) Payment was somewhat informal and clunky via PayPal. But communication from them was as good as any other vendor. https://store.rotkee.com/en/automotive-diagnostic-equipment/ Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted March 25, 2023 Author Share Posted March 25, 2023 Quick update: I test-fit the Rotkee flexible adapter in a spark plug hole and realized the issue with thread engagement: the adapter’s male threads are too short to properly engage the threads in the head. The threads on the transducer are the same length and would have the same issue. My other (standard gauge-type) compression tester is built the same way, and utilizes an extender of sorts which has threads of the appropriate length. I used that piece on the end of the Rotkee adapter and verified proper engagement in my 4.0L’s head. Picture attached. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted March 26, 2023 Administrators Share Posted March 26, 2023 SomeBuckaroo...Thanks much for all the details on this product line. Rotkee is trying! Your workaround with the compression gauge adapter is what I envisioned. Thanks for testing and validating. Rotkee may want to add a similar (common) adapter to its product line. Yes, the 3/4" thread lengths on Jeep and many other spark plug applications do need consideration. Rotkee and others try to "universalize" the tool. In the process, they use a shorter reach to fit cylinder heads/spark plug holes with shorter thread depth. If they made the threads 3/4" length, the adapter or transducer would interfere with the piston crowns! I would use Rotkee's adapter hose all of the time to get the transducer away from the warm-to-hot cylinder head. I am looking forward to supporting this vendor! They build quality products. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted April 17, 2023 Author Share Posted April 17, 2023 I finally found some time for Jeep work. The first project was to fix the vacuum leak around the throttle butterfly shaft. That was straightforward - and did appear to fix my intermittent high-idle issues - but my "slightly rough" idle issue persists! The second order was to install the Rotkee in-cylinder transducer and get some running measurements. I used their flexible adapter with my own 3/4" thread adapter at the end. The flexible adapter makes installation much easier and keeps the transducer away from the hot head. I first ran the engine to operating temp. For each cylinder, my test procedure was: - disconnect fuel injector - remove spark plug - use small alligator clip to ground the disconnected spark plug wire - install pressure transducer I captured intake manifold pulse sensor (top trace), in-cylinder transducer (middle), and crank position sensor (bottom) - to provide an absolute reference of crank position. Just over 30 seconds of scope video for each, including several throttle-snaps for each cylinder tested. I got through the first 3 cylinders and then decided I ought to validate my setup & procedure here before spending more time. My intake manifold and in-cylinder transducer waveforms jump off-screen on the throttle snaps, which is less than ideal. I think I also have a scope trigger setup issue which causes some events to be missed. Let me know what you think & any suggestions for improving my measurements: Cyl1: https://youtu.be/UiF4RiFAenI Cyl2: https://youtu.be/PUuuMAzdwmY Cyl3: https://youtu.be/hUi-LZ30t3g Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted April 20, 2023 Administrators Share Posted April 20, 2023 Footnote for our scope tests: From experimenting and reviewing scope results and other users' experiences, it is clear that the in-cylinder pressure transducer cranking tests (engine not running) should be done with the throttle closed. This is unlike traditional cranking compression tests where we hold the throttle open for maximum available air. The pressure transducer readings are far more stable with the throttle valve closed. SomeBuckaroo...For those following this testing and your original concern about a high and erratic idle, the throttle shaft air leak is a concern with any higher mileage EFI engine. For those wanting a quick test for throttle shaft leaks, try this: With the engine idling, a light mist of WD-40 at the throttle shaft ends will change the idle speed momentarily if a leak is present. (Keep WD-40 spray away from engine heat!) Good work here. As for the scope patterns, the CKP (your lower reading) looks like the pulse/phases from the flywheel's Hall effect. There is not a single TDC spike but rather multiple spikes for the same cylinder. As for this CKP reference, you really don't need it. True TDC with an in-cylinder pressure transducer is at the peak of the compression rise. The highest pressure point in a cylinder, the peak of your pressure waves, is the piston at TDC on the compression stroke. Instead of a channel devoted to the CKP sensor, I would do spark firing lines on that channel. See how your spark timing aligns with the pressure peak TDC. At an unloaded idle, I would expect 12-14 degrees of spark advance. At least 10. If your distributor is indexed properly with the crankshaft timing mark, ignition timing (advance degrees) is also a clue to whether the valve timing is correct. In-cylinder pressure testing can identify the valve opening and closing points. To do this, bring your scope pattern to view two of the highest compression points. These TDC points are at the top of the compression stroke, so between these two peaks is a distance of 720 crankshaft degrees. (This is a 4-stroke engine, two crank rotations per four stroke cycles.) If you have cursors, stage the curses at the two TDC pressure peaks and also the two BDC (bottom dead center) points between these pressure peaks. Start at the left with the first downslope from peak pressure. This is the Power Stroke. Near BDC, the exhaust valve begins to open for the Exhaust Stroke. The piston rises with the exhaust valve open, so pressure is slightly wavy and on a horizontal plane or plateau as the piston pushes exhaust from within the cylinder into the exhaust system. (This area of the pattern will reveal an exhaust restriction.) Hot gases are being pushed out of the cylinder. Then comes the valve overlap opening between the exhaust valve closing and the intake valve opening. The pressure goes negative (waveform drops) as the intake valve opens, beginning the intake stroke down near BDC. (The cylinder is filling with the intake air/fuel mixture.) Finally, you see the Compression Stroke as the waveform sweeps upward to the second pressure peak at TDC on the compression stroke. You have the cursors set at 180 crankshaft degrees apart. If you know the valve timing events (degrees) of your camshaft's profile, you can identify where the valve opening and closing events should take place. You can check the valve timing by plotting the intake and exhaust valve opening and closing degrees. Match up the crankshaft/camshaft degrees with cylinder pressure changes. Keep in mind that the camshaft rotates at 1/2 the speed of the crankshaft and 1/2 the number of degrees as well. The waveform will approximate the valve opening and closing points or degrees. (Flat tappet lifter ramp-up and other variables prevent a pinpoint reading of the camshaft timing and valve opening and closing points. A roller camshaft would be more accurate.) This will be close enough to indicate whether the valve timing marks are aligned on the sprockets or if there is slack in the timing chain (retarded valve timing). Your parts are new, and you were careful installing the valve timing set. There should be no issue here but verify for practice. If valve timing checks out, so far I don't see any abnormalities. Yes, it would be helpful to see full pressure spikes under snap throttle. The cylinder pressure rises considerably, and if telltale piston ring or valve leakage exists, it should show here. Compare the peaks of the snap throttle waveform. Cylinders should be uniform. When cylinders seal well, the waveforms will be equal if the ignition timing and fuel mixtures are correct. As for your cranking pressures versus these running engine pressures, they can be compared. You did the relative (cranking) compression check. That was without firing the engine. Running cylinder pressure is with fuel mixture in the cylinder on its compression stroke and firing for the power stroke. Cylinder pressure will spike with the snap throttle. See what pressure values the transducer reveals. There is nothing apparent yet that would contribute to a rough idle. The intake pulses seem to match per cylinder. To rule out a mechanical issue like valve leakage, weak valve springs or piston ring seepage under load, we'll need to see the entire snap throttle waveform. Then check and compare all six cylinders. When you get the next round of waveform patterns, I will help break down the cylinder patterns and pinpoint any signs of compression loss from either the rings, leaking valves (weak or broken valve springs, poorly seating valves, etc.) or even the head gasket. This will show up distinctly at either an unloaded idle or snap throttle loads and deceleration. Your in-cylinder pressure transducer can pay for itself here! We've been ruling out mechanical issues. The ignition has been spot on in other tests. Did you check the intake manifold-to-head gasket for a leak? (The WD-40 test works well here. Avoid heat with the spray mist.) Complete the cylinder pressure tests, keep using the intake pulse sensor. After that, among the few concerns left would be an exhaust manifold leak affecting the O2 readings or the fuel injectors. Have you checked the fuel rail pressure? Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted April 30, 2023 Author Share Posted April 30, 2023 Hey Moses - I completed another round of scope captures. This time, I used my scope's ability to capture 10 seconds of waveform (during which I snapped the throttle twice) and then "zoom in" to view & scroll. This capability is equivalent to more modern PC-based scopes, except I have to manually take a video. My previous scope videos filmed a succession of individual trigger events as they occurred live. The benefit of my new approach is clearly seen on throttle-snaps. For each cylinder, I captured intake pulse-sensor, in-cylinder pressure, and spark for the respective cylinder. The spark waveforms have considerable and apparently random level shifts. You'll see what I mean on the videos. Is this an indication of an issue? I have noticed this strange phenomenon on every spark waveform I've ever taken on the Jeep, regardless of the probe I used (Hantek & Rotkee). I used my scope's cursors to measure the spark timing for each cylinder at idle. The waveform periods were 163ms +/- 2ms; spark occurred 3.3ms +/- 3ms before TDC. My spark timing math is: (3.3ms / 163ms) * 720 degrees = 14.5 degrees BTDC (on average). Considering the +/- bounds of my measurements, spark timing across all cylinders is within approx 13 to 15.5 degrees BTDC. I think this seems fine. Links to 10 seconds of runtime-capture for each cylinder are below. I have not yet taken a look at the valve timing other than generally noting it appears to be consistent across all cylinders. One final note - I disconnected the fuel injector for each cylinder-under-test, and re-connected the harness to a spare injector to "simulate" a proper connection to the PCM. Interestingly, I did not get any CELs/codes during the entire test procedure. This is surprising as I would have expected the PCM to observe a non-firing cylinder via variations in the crank rotation rate (as indicated by the crank sensor's pulses). Maybe my "early TJ" (1998) computer does not employ this level of sophistication? Or maybe Chrysler didn't consider total failure of a single cylinder to be cause for concern!? 🙂 Cylinder 1: https://youtu.be/JOXoML21DGw (only 1 throttle snap) Cylinder 2: https://youtu.be/SY29Auky-no Cylinder 3: https://youtu.be/gxXaIXRLGTg Cylinder 4: https://youtu.be/UIGyj509Hn0 Cylinder 5: https://youtu.be/4oou6IcsztY Cylinder 6: https://youtu.be/22aqkqs6Me4 (The engine background noise in the videos is indeed my Jeep running with the respective cylinder disconnected, but does NOT correspond to the waveform being shown) Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted May 1, 2023 Author Share Posted May 1, 2023 Re-reading this, I noticed a significant typo: My spark timing was measured to be 3.3ms +/- 0.3ms before TDC. Apologies! Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted May 1, 2023 Administrators Share Posted May 1, 2023 SomeBuckaroo...I watched the six videos through with keen interest. Wave patterns for each channel need to be viewed individually when comparing cylinders. From what I can see, the valve opening events and related changes in pressure look normal as does the compression. There is no loss of compression toward the top of the in-cylinder readings, even when you accelerate or snap the throttle. In these tests, what is the actual compression peak in psi at an idle and at snap throttle? I have no concerns about the spark firing lines. When you accelerate, the cylinder pressures rise, and the voltage requirement to combust air/fuel is affected by increased resistance. The firing lines rise accordingly. This indicates that the ignition is meeting the firing demand. The rise is uniform, as expected, since the same coil is firing all six cylinder on a distributor ignition. (Compression psi and firing resistance are nearly equal per cylinder, too.) If this were a coil-on-plug engine, there would be some variation (likely minor) in the spark firing line heights for the cylinders. Deceleration causes the erratic spark lines as the ignition firing voltage varies here. The firing load and voltage demand are rapidly dropping, and there is a fluctuation in cylinder pressures with the decreased load, closed throttle plus the arbitrary cutting off of the air/fuel supply with an MPI/EFI engine. (There's no "venturi effect" like with a carburetor. With EFI, the fuel flow is electronically cut off during deceleration to reduce emissions and wasted fuel.) This doesn't look unusual. It would appear differently with my MP408 scope and software. Your "pure" lab oscilloscope can't leave anything alone...and you get to do more math. Regarding your hooking up a spare injector to simulate the firing of the injector during in-cylinder transducer testing, two things are going on. First, the PCM sees a completed injector ground activation signal and reads it as normal, regardless of whether fuel is flowing through the injector or not. (The system is not sophisticated enough to sense fuel pressure changes at individual injectors; it only monitors fuel pressure for the entire common rail.) Your extra injector is not hooked up to the rail fuel supply. If it were, there would be fuel spewing from the injector's nozzle with the engine running! You're running the spare injector dry, which raises the question of whether an injector needs fuel flow to lubricate the pintle, a topic for another thread. Secondly, the crankshaft position sensor (CKP) is oblivious to any "pulsing" of the crankshaft or injector firing. As long as the PCM gets a TDC signal from the CKP, all's good. Your spark timing varies as it should under different engine loads and throttle positions. I see nothing unusual here. Intake pressure pulses are stable at idle and vary as expected with the throttle snap and deceleration. Your scope is detailed here. As for injector pulse width and fuel flow, even though you remotely hooked up an injector to represent the dead hole/cylinder, that circuit is functioning. The spare injector simply did not receive fuel from the rail. You also did not have fuel flowing through the intake port at the tested cylinder since you disconnected that injector at the harness. That's a good thing, since you do not want fuel inside the cylinder while you're using the in-cylinder transducer! Always disconnect the injector for the cylinder being tested with an in-cylinder pressure transducer. As for no codes and no PCM sense that the false injector was in place: If you had left the cylinder's injector hooked up during the in-cylinder transducer test, the steady stream of unburnt fuel passing through the cylinder would have thrown off the O2 readings. Raw fuel passing out the exhaust would likely throw a code for a cylinder misfire. (Also, the system would try to compensate with a fuel trim adjustment, making the operating cylinders run lean.) Injector pulsing and fuel flow work independently of the ignition system. You grounded either the spark lead or spark plug for the test cylinder. That cylinder could not fire without fuel or spark...but the PCM did not "see" a problem since both the injector pulses and spark took place. The spark firing line for that cylinder would reflect the grounded lead or grounded spark plug "firing"—without a load. You would not get a "normal" spark firing height from that cylinder because there was no cylinder pressure or load affecting the spark plug. Summing it up, your compression and power stroke (expansion) patterns look good and uniform, no sign of compression loss. The valve opening and closing events and patterns are good. There is a consistent, slight "spike" near the exhaust/intake valve overlap point during deceleration after the throttle snap. This could be a function of the exhaust pulse change, maybe pressure changes in the exhaust system. There's no indication of a "clogged exhaust". Each cylinder shows this bump. Is the exhaust system (header/pre-cat, muffler, catalytic converter and pipes) "original"? Using your pulse sensor in the tailpipe, run the engine at an idle then a steady 2500 rpm for a moment. (Stay away from the engine driven cooling fan!) Compare the exhaust expulsion pressure waves and see whether the exhaust pulses look normal. Compare the pulse sensor readings at an idle, 2500 rpm and also a snap throttle test. You're looking for a significant restriction. A restriction affecting idle stability is unlikely. Restriction causes poor fuel mileage and a power decrease. In-cylinder transducer testing is for mechanical trouble. Spark firing lines, base and advanced spark timing (which are normal in this case) and injector flow are tune related. Your main concerns here were the in-cylinder wave form for pressure peaks and uniform rise and drop in pressures, the valve opening and closing events (reflecting cam lobe profiles, valve sealing and valve timing), the shift from pressure to vacuum and vacuum to pressure, and your intake and exhaust pulses. The intake pulse readings look normal, indicating that you whipped the air leak at the throttle body. I would run the simpler exhaust pulse test with a snap throttle. This will rule out any exhaust restriction. Then move on. If the MAP and O2 sensors are functioning correctly, any remaining idle instability (beyond what's "normal" for a Jeep 4.0L MPI engine) would have me looking for an upstream exhaust leak at the header/exhaust manifold. (The 4.0L exhaust manifolds are notorious for cracking between tubes.) Next I would flow test and compare the fuel injector flow rates. Given your mileage, and despite the new engine, either of these possibilities are worth checking. I invested in my Autool C200 6-cylinder fuel injector tester for this troubleshooting scenario. At high mileage, the OEM Siemens injectors needed ultrasonic cleaning and new filters but were otherwise fine. Even at nearly 200K miles, I trust these OE injectors for flow and reliability. They're more predictable than Brand-X offshore injectors that many owners buy on the cheap at Amazon or eBay. At the time, the C200 machine cost about the same as a full set of Bosch replacement injectors. The machine should last forever with the volume of gasoline injectors I will clean. Still very much worth the investment. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted May 1, 2023 Author Share Posted May 1, 2023 Thank you for the fast follow-up and analysis! I was planning on capturing exhaust pulses, which I think would be the most straightforward way to "quantitatively" describe the slight idle issues I'm experiencing. I could capture cyl 1 spark, O2 sensor, and injector pulse-width as well. Could I capture only 1 injector, and assume the pulse-width is the same for all? I have worked on the exhaust system extensively in the past: Over the years I replaced the header (several times), muffler, catalytic converter, all piping except for tailpipe. I am very familiar with the header-cracking issue, but I believe my current header is not cracked. I will investigate further. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted May 8, 2023 Administrators Share Posted May 8, 2023 SomeBuckaroo...Sounds like a good strategy for ruling out uneven or restricted exhaust pulses. As for pulse width, yes. The injectors all receive the same PCM pulse width signal. The duration of the pulse width is controlled by sensors and fuel trim. Rail fuel pressure is (should be) constant at each injector. As for the actual fuel volume flowing through each injector, that would have to do with the individual injectors and affected by nozzle clogging/restriction or a restricted injector filter/screen. If there is a restriction, the PCM adjusts the pulse width to compensate. The PCM does this based on receiving the O2 and input from several other sensors. This system is not intuitive. It is strictly mechanical. There are software PIDs and algorithms that control air/fuel ratios and injector pulse width. Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted May 9, 2023 Author Share Posted May 9, 2023 Thanks, Moses! I am also thinking about monitoring fuel rail pressure variations using the Rotkee in-cylinder transducer (along with a DIY adapter). I saw a write-up online describing how fuel-rail pressure pulses during each injector cycle can indicate & pinpoint clogged or malfunctioning injectors. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted May 11, 2023 Administrators Share Posted May 11, 2023 You're welcome, SomeBuckaroo...The fuel rail pressure test would be an interesting angle. Be aware that your single-rail system uses a damper on the rail. Many mistake this for the pressure regulator on pre-OBDII engines with the two-rail injection system. Your pressure regulator is at the fuel tank module. The wild cards would be whether the damper compensates for the pulses you describe and if the fuel pressure regulator is sensitive enough to offset these pulses. It's not likely that subtle variances in injector pulses would be picked up by either the damper or pressure regulator. So, it's worth experimenting. Give it a whirl and let us know if this is a valid way to detect fuel flow issues per injector! Moses Quote Link to comment Share on other sites More sharing options...
SomeBuckaroo Posted May 31, 2023 Author Share Posted May 31, 2023 Minor update: I got a cheap automotive smoke machine off Amazon (https://www.amazon.com/dp/B09R44F4JV) and used it to check for exhaust leaks starting at the tailpipe. I pulled the O2 sensor out for about a minute to verify smoke traveled up to the manifold. I found several exhaust leaks from the muffler back, but nothing near the exhaust manifold. I also checked for intake leaks, but found none. Somewhat surprisingly, when I pulled a vacuum hose off the check valve right near the intake, I could hear the hiss of air ("escaping vacuum"? 🙂 ) - so amazingly, no leaks from that point to the HVAC controls. The other day I revved the engine a couple times about 10 seconds after cold-start, and it almost sounded like it back-fired both times. Very strange. One warmed up, it seems to behave fine. I assume 10 seconds after cold-start the PCM is still running open-loop, which may (or may not) help narrow things down. One other point - my '98 4.0L fuel rail does not have a damper, but it does have a threaded port with schrader-style valve for pressure testing. Were fuel rail dampers introduced after '98? Or is mine missing that component? It is possible that a PO made ill-advised changes. Quote Link to comment Share on other sites More sharing options...
Administrators Moses Ludel Posted June 3, 2023 Administrators Share Posted June 3, 2023 SomeBuckaroo...Glad you confirmed no vacuum or exhaust leaks. Nice to have the smoke machine! Fuel rail dampers were used on the single-rail EFI/MPI systems. After your finding, I looked closely at the Mopar parts catalog, and the rail part number does change in 1998 (no illustration to substantiate whether the update includes the damper). Your part number was gradually phased out. The purpose of the damper is to act like an accumulator and eliminate fuel pressure pulsing. Likely with your engine, they discovered the issue. 1998-up corrected it. The two-rail system (Mopar MPI version from 1991-96) used a pressure regulator where the damper fits on the 1998-up engines. The distinction is that the regulator flows excess fuel back to the tank, maintaining rail pressure. The regulator also has a vacuum diaphragm that increases injector rail pressure when the engine cranks (zero or very low vacuum signal to this vacuum port). This slightly enrichens the fuel supply and is not the same as open loop, cold engine fuel trim. The diaphragm is a mechanical solution for boosting fuel flow/enrichment during cranking. Presumably, your engine without a rail damper may experience fuel surge pressure fluctuation from the fuel pump. Your system relies strictly on the pressure regulator at the fuel tank module and does not return fuel from the rail. Instead, excess fuel pressure bypasses at the pressure regulator and goes directly back into the fuel tank. There is no need for a second rail fuel return line to the tank. Would it pay to change your rail to a damper-type rail? This would have little bearing on the rpm fluctuations you describe. Of more concern to me would be the overall fuel pressure constant. The Schrader valve is where you hook up a diagnostic pressure gauge. It would be interesting to see whether the pressure is fluctuating (rising) when you experience the bump in rpm. If the pressure regulator were malfunctioning (not bypassing back into the tank or a restricted return), that could increase pressure at the rail and cause the injectors to flow more fuel, likely raising rpm. You should see fuel trim trying to compensate. Worth considering. Moses Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.