Buick Park Ave - 3800 Lack of Power, Noise Upon Accel
Getting to the root-cause of a fault is always of the utmost importance. Without doing so, could lead to repeat failures, further damaged systems or components and last but certainly not least, customer dissatisfaction. With all that being taken into consideration, I still always look for the most efficient path to take, to carry me all the way to the root-cause. Many times, it’s simple but other times, requires a bit more thought… Like the subject vehicle below. This one forced me to start at the end (the “symptom”) and work my way backward. Employing action/reaction testing is what led me to systematically eliminate possibilities. Knowing what is “good” on a vehicle is just as important as discovering what is “bad”.
I was recently faced with a 2002 Buick Park Avenue with the power-train listed above and 130k miles boasted on the odometer. The vehicle was brought to the shop with the complaint of a “noise” under acceleration as well as a lack of power. I drove the vehicle to qualify the complaint and the noise and performance-issue is quite evident. The vehicle exhibits a “ping” as well as the sensation of either a breathability-restriction or under-fueled powerplant. A scan for DTCs was carried out and not one was stored as current or in history. Some basic vehicle PIDs were monitored (during the fault) to determine the PCM's ability to deliver fuel. As can be seen by FIG A, both HO2 Sensors reflect over 900mV when the vehicle is placed under high load / RPM, indicating an adequate fuel supply. Keeping in mind that this vehicle utilizes a single-bank feedback fuel injection strategy, only one primary HO2 sensor reports for all 6 cylinder’s exhaust gas.
After initially seeing no lack of fuel delivery, I decided to monitor Knock sensor activity. I want to be able to see of the PCM detects the noise that I’m hearing audibly and if it will respond with a countermeasure. FIG B displays the PCMs response to the “ping”. It is attempting to quiet the noise by retarding ignition timing approximately 25 degrees (This explains why the vehicle is lacking in power output). I understand that a “Ping” is a noise due to collisions of multiple flame fronts within a combustion chamber. These can be caused by pre-ignition due to advanced ignition timing, simply discharging the spark at the wrong time. Hot-spots within the combustion chamber igniting fuel before combustion was intended can cause the same symptom. Another occurrence could be due to extreme combustion chambers due to either an insufficiently operating cooling system or improperly functioning EGR system. Both systems will lower combustion chamber temperatures when functioning correctly. Through visual inspection of the cooling system, temperature PID, inspection of the cooling fan and thermostat operation as well as coolant level, I’m convinced that the fault doesn’t lay within the cooling system, so I proceed to monitor the operation of the EGR system.
FIG C demonstrates the PCMs response to the “Ping” once more (viewed here just as a point of reference to indicate what I’m hearing with my ears) but also that the EGR is being commanded on with load / RPM changes. This signifies to me, the PCMs intent. By simply bringing the engine to an idle and commanding the EGR open with bi-directional control (through the scan tool), I’ve proven out the inert EGR gas is free to flow, as the engine began to stall. At this point, I’m done gathering preliminary data and its time to get back to the shop and start to dig-in.
This vehicle utilizes a waste-spark ignition system and the ignition cables are easily accessible and serve as “low-hanging fruit” to obtain preliminary data about the engine’s operation. By monitoring the secondary-ignition waveforms, I can get a quick look inside each cylinder during the combustion process. If you view the movie below you can see and hear the fault occurring.
FIG E represents a still-capture of the fault occurring in the movie. Comparing a proper firing cylinder (in RED) with the suspect firing cylinder (in YELLOW) its clear that they are very different from one another. The YELLOW trace is displaying an increase in secondary resistance as the spark continues to burn within the combustion chamber. This is crudely worded, but it very much explains the nature of the fault, exhibited in the waveform. A cylinder that is adequately fueled requires less energy from the ignition system because its easier to maintain the plasma channel (a very conductive tunnel for the electrons to flow through) occurring within the spark plug gap. A cylinder that lacks fuel requires more energy to maintain that plasma channel and the KV demand increases, causing the spark line to nose-up towards the end of the ignition event. So, the ignition waveforms are telling me that the rear bank is suffering from a lack of fuel supply, but the front bank is adequately fueled. The next step would be to see if the fuel injectors can deliver the same quantity of fuel given the same opportunity. This is where I perform an injector balance test. I also took the opportunity to sample the fuel to analyze for contamination. We’ve all seen contaminated fuel cause some strange symptoms. The results of the balance test reveal that all injectors deliver an identical volume of fuel per test. The results of these tests reveal that the issue is of a “controlled” nature and not a mechanical restriction. After analyzing the fuel, it appears it is not contaminated nor the reason for our symptom FIG F.
Eliminating mechanical-restriction / contamination from the list of possible causes was easy to do and made sense to perform those tests before getting too involved. The next logical step was to monitor for injector current flow at a common point (under hood fuse box) to verify the circuitry didn’t differ from injector to injector. This is another great test because its easy to perform, non-invasive and it checks all injector circuits for comparison, a good “go/no-go” test for circuit-health. (FIG G and FIG H). The results of this test don’t point to the health of an injector circuit to be the cause of the lean condition / ping. The only issue with how I the test was conducted was that the fault / ping wasn’t present throughout the test. This is crucial to proving out the cause of the fault. If the fault is present during the testing, it is more likely to exhibit in the test results. So, by placing the vehicle back into a state of operation when the fault presents, (under steady-state brake-torque conditions) I then analyzed the injector current, make sense?
Here, in the injector current waveform, its clear to see that the injector pulse width differs from injector to injector and there is an occasional drop-out. This further condemns this car to having a control issue.
The PCM simply provides the ground path for the individual fuel injectors as this vehicle utilizes a sequential injection strategy. Each injector has its own driver within the PCM. The injector current trace in FIG I demonstrates that at least on of them is not being controlled properly. However, we simply can’t blame the PCM as is only programmed to do what is it told. If the PCM receives bad inputs (CKP + CMP signals), it makes bad decisions (Ignition coil + Injector pulse width). The next logical step would be to monitor the action / reaction nature of the fuel injection system by analyzing the current (the work performed) and the inputs to the PCM responsible for calculating and carrying out the injector drive. I simply repeated the test under fault conditions and watched for the anomaly to occur in the injector current trace. I then looked for a correlating fault within the CMP or CKP signals. FIG K displays this perfectly. In GREEN is the CMP sensor signal and in BLUE is the fuel injector current trace. The anomaly that is occurring in the CMP trace seems to be perpetuating a fault in the current trace (the work performed). “THIS” is the cause of the lacking fuel supply and the “ping” the vehicle is suffering from. The significance of the capture a is a “bad” input to the PCM creating a bad output.
So, the CMP sensor signal is displaying a fault. After analyzing the signal, there are a few scenarios that can cause this:
- Sensor malfunction/ damaged sensor
- Poor connection on the reference voltage supply or between the sensor and the PCM
- Loaded reference voltage supply or sensor signal circuit
The next step would be to apply these ideas and reference the wiring diagram to develop a game plan that will yield answers to my questions efficiently and without getting too dirty. If you refer to FIG L, we can see that the ignition control module is supplied ignition voltage at terminal P. It’s clear to see that the Ignition control module then supplies the reference voltage for the three-wire, hall-effect CMP sensor, at terminal N. The diagram also indicates that the supplied reference voltage is shared with the CKP sensor, as well, via splice S136. Because of the wiring diagram’s indications, I then chose to monitor the operation of the CKP to see if it exhibits the same suspect-characteristics as the CMP. Doing so allows me to rule out the CMP as a “defective-sensor”. If the CKP signal is not affected, the fault is not common to both circuits. Therefore, it simply can’t be related to the ICM reference voltage output circuit in any way. Because I’m using a very capable multi-trace lab scope, I will capitalize on its operating characteristics. This scope allows me to gather lots of info and only view what I chose, by turning “off” the channels when I don’t wish to see them. I added, to the capture, ignition supply @ the Ignition control module. In one capture, I will be able to determine if the fault within the CMP signal is also present in the CKP signal. I will also be able to determine if the fault is coming from the ignition control module and if so, is the supply voltage to the Ignition Control Module, the root-cause of the fault. It’ like fishing with a net instead of a hook. FIG M displays that when the fault is present In the CMP signal (RED Trace), it is also present in the CKP signal (YELLOW Trace). This indicates the fault is related to both circuits. The fault is also present in the common reference supply voltage (BLUE trace, as indicated by the voltage drop). If you then reference the GREEN switched-ignition supply voltage, it remains intact. The results of the above test indicate the fault lays between the ignition control module and both the CKP as well as the CMP sensors. We are getting closer to pinpointing the fault, but we still must determine if the cause of the voltage-drop in the supplied reference voltage is due to a fault internal to the ICM, a faulty sensor pulling the voltage down or due to a loaded circuit (resistive-short to ground).
There are several ways to go about doing this, but I like to keep my hands as clean as possible and solve this riddle with a minimal amount of time and energy invested. The route I choose to go is with the implementation of a Micro-amp clamp FIG N. This device is very sensitive and will indicate the flow of minute amounts of electrical current. My thought process is as follows:
- If the fault is due to a short-to-ground, it will cause current flow in the supplied reference voltage circuit to increase.
- If the fault is due to faulty circuitry internal to the ICM, the current flow in the supplied reference voltage circuit will diminish.
FIG O demonstrates where I interfaced the Micro-amp probe to the vehicle, a point common to both the CMP and CKP sensor reference voltage feeds. I then place the vehicle back under brake-torque conditions (when the failure surfaces). If you view FIG P, you can see that the current increases as the voltage signal from the CMP sensor is skewed. This indicates that the ICM is not at fault and that the circuit is indeed loaded. From this point it’s a matter of finding which point of the engine harness the fault is located. A quick visual inspection of the wire harness was carried out. Assuming nothing suspect was found, Id must proceed with the micro-amp clamp to the individual legs of splice S136. I got lucky, though…a suspect-area was located. Directly above the crankshaft balancer is where the harness feeds the CKP sensor. It’ s supposed to be secured to the timing cover however, the harness is flapping in the breeze. Upon closer inspection the source of the fault has been pinpointed: FIG Q. The reference supply voltage wire insulation is rubbed-through and the copper is finding a ground path through the crankshaft balancer intermittently.
So, to recap on the chain of events…the customer feels a lack of power and a correlating “ping” from the engine. The scan data reveals no lack of fueling however, the HO2 sensor proving the input is referencing all 6 cylinders and the fault is erratic, not consistent. The data is not supportive of the fault. Secondary ignition revealed the entire back bank to be lean. Initially, I believed “Erratic fuel injector on-time” was the cause of the lean condition but was the due to faulty CMP /CKP sensor inputs. The sensor signals were skewed due to a compromised common-reference voltage input to the sensors and causing the PCM to resync. The thought crossed my mind that the same inputs were utilized for ignition coil firing strategy, so the ignition system was scoped as well. FIG R shows the reference supply voltage in BLUE and the coil current ramps in white. Clearly a relationship between mis-triggering of the coils and the voltage drop in the reference supply voltage. My thought about the “ping” is leaning more towards spark being discharged too early (due to the mis-triggering of the coils) In FIG S I spanned cursors to even space the ignition firing events. You can see that one is discharging relatively early. This may be enough to generate the “ping”, but I can only speculate and didn’t prove it any further than that.
The end-result was that an understanding of your available tools capabilities, understanding of system design and familiarizing yourself with the design of the related circuits/ strategies it is simply a matter of time until the root cause of a fault is pinpointed. I can think of several components that would’ve been thrown at this vehicle as an attempted repair, had a technician not been properly equipped to analyze this fault. It seems it would likely have been a costly error.
It was great seeing this the other night on the live stream! Great find and better is the way you got there. One test leading to evidence for the next test. So glad I joined up with you guys!
Brandon I loved hearing about this case study the other night that I had to read it on here again. haha. Are you teaching any classes?
I really appreciate the positive feedback, Thank you. I have recently begun producing my own classes. They will all be based around my own experiences and practices. I am willing to hold a class where ever there is interest. Feel free to message me privately if necessary. Thanks again.
Brandon, I loved listening to this case study live and hearing your thought process on the issue on the "Trained By Techs" YouTube Live stream. But I think it was even better to be able to read your explanation and see the captures here to analyse them myself!
R it’s always great to get feedback from “ MY DUDEs”! I do appreciate your input, Keith as I very much admire your diagnostic abilities. Thank you
Good reading your Diagnostic procedure for the people that are too old to stay up to the weeeeee hours of the evening (Like 9:00 pm), I like this diagnostic stuff, I might get into it myself someday. ;-) Good Work, Grasshopper
When you can snatch the current probe from my hand, you will be ready to diagnose.
I saw this on the stream the other night Brandon.
I kept thinking, "This would have been a two minute video if DeFazio had the car." ;-)
I have no doubts...proud to call Keith DeFazio a “Friend”
btw, The idea of using the micro-amp-clamp how you did is quite clever.
Really nice. I like the thought process and methodical approach and of course, the really cool scope captures. ;-)
I was lucky to have caught this on the live stream/chat, to go through your diagnostic approach with you, and it was amazing then. I come to find it again, here, with the same screen captures, but the way you express the process in written form is just as impressive. You have an amazing way of articulating and engaging an audience. Well done!
I believe your thought process of going "Backwards " is a very clear and thought out way of streamlining your approach. I attempt to explain to students " Put the known good in a Box" That way your direction will not be skewed, if I may say...........Very good ...no! excellent Process Mr. Steckler, you covered every probable disruptor with confidence and expertise and ultimately your evaluation/diagnosis was .....on the mark.
CR…. AMAM, CMAT L-1
Loved the live stream version as well, we all got a free class so thank you Brandon!
Makes you wonder how many parts would be loaded in the parts cannon if a technician had not used your logical approach and practiced methods. Your thorough understanding of voltage compared to current in a faulted circuit didn't hurt either ;). Thanks for the share.
Much-appreciates, my friend. Thank you!
I like your style of teaching , I asisted to your honda class here by San Rafael couple months ago organized by world pack and I wss very satisfied with the learning , last Saturday I was in other world pack class in Burlingame for sprinter which I considered over all the class was a waste of time and money , most of the class was if you have this dtc repace this part if if does not fix it replaced this part .
I wish we had had dtc , setting criteria scan data than scope to confirm how ever there is not a single scan data or scope scren capture at the entire material , besides the teacher lebby almost wanted to kick me out the class for asking specific questions, I asked when we were on data lines communications from what ranges of voltages does can b , is switching 0 to 12 or what, his unswer was you do not need to remember that , so I said I agree how ever i need to know if I am trouble shooting this system to know if this system , if I am flat 12 volt id this a module hanging the voltage or shorted to 12 volts system or shorted to 5 volts system, his unswer was I am electronic engineers so if you want we can even talk of binaries numbers , So I do not need binaries number because I am not software developers or programmers I need simple ranges of low and high voltages . later he ask specific question about can network and pointed to me and said mr can wi unswer that question , come on I am here to learn and by the way I knew the unswer because I am familiar wih with can network voltage ranges and resistance values but not with can b where my question was ,
I would like to know if there is a world pack or place to specific hear this comments for student asisting to this training , or they wilm automatically kick me out , I think You will remember me in your class I ask a lot of questions , also let me know what I can inprove how to ask and keep flowing you class too.
thanks apreciate your hard work .
I’m very sorry and disappointed to hear of your experience with that particular class. I urged you to reach out to the folks at WTI and tell them what you told us here on DN.
I do want to thank you for your feedback, both on my class and the sprinter class you attend it I do want to thank you for your feedback, both on my class and the sprinter class you attended. It is very important that we get this feedback from technicians like you… Weather it is positive or negative feedback, it needs to be known so changes can be made. Thank you