Shop by Price
Join our Newsletter
Subtotal: $0.00
Blog Tags


The Cranky Outputs of P0172 on a 2002 Chrysler PT Cruiser

by Adam Robertson, instructor for the CARQUEST Technical Institute, Western United States


The throttle must be wide open to start this 2002 Chrysler or the spark plugs will foul from excessive fuel. Once finally running, the vehicle will stumble, yet can be driven if kept above about 1200 rpm. If you let it idle, it will blubber down and try to stall. Black smoke puffs from the tailpipe at times. Check engine light is on. Continually sets a P0172 (rich condition bank one). There are no other codes.

Some Basics
This vehicle uses a speed density system, which is a MAP sensor (not a MAF). In this type of fuel control system, proper vacuum is critical. We are working at about 4000-ft above sea level and the idle vacuum is about 13 inches. It will achieve about 17 inches when free revving at about 2500 rpm. The O2 sensors are soot fouled but when warmed enough are pegged near 1v. The fuel trims show a rich condition but this vehicle is hesitant to go into closed loop. A tremendous amount work by various shops has been performed on this vehicle to no avail.

Initial Thoughts
Vehicle obviously running rich:
   • Where is the excessive fuel coming from?
   • What equipment could I use to determine this?
   • What systems do I test?

My Starting Strategy
In this situation, initially there are no clues specific enough to provide a specific diagnostic direction. Therefore, I have decided to do a quick test to look at three key dynamics: 1- Mechanical condition. 2 - Spark. 3 - Fuel. To accomplish this, I will connect my 4-channel Pico scope to ignition and injector; and via a pressure transducer to do an in-cylinder running compression test. These tests will give me a quick overview of the system including the general mechanical integrity via the compression waveform.

The following waveforms were captured from cylinder #4 (it was the easiest to access).

Test Condition: Cranking but did not start.
Red: Injector synch for cylinder #4
Green: Ignition noise (synch) from all cylinders.
Blue: In cylinder compression waveform.

The waveform below is of the same conditions but I have zoomed in for detail and added some cursors. Notice the peaks of the blue compression waveform. Each peak represents TDC Compression. From first peak to the second peak is a full 720degrees of rotation.

First, notice I have removed the green ignition sync signal, as it was fine. Looking at the blue compression waveform, we see about 140 psi and the waveform overall looks normal (cylinder #1). Next, look at the second injector pulse in the red waveform. I have the time measurement cursors (vertical cursors) aligned to measure the on-time of this injector event and it is approximately 16.5ms. All of the pulses look about the same. But, I do see an issue! This is a port fuel injection system. Meaning, each injector is pulsed independently via its own driver. The red waveform is only from injector #1 and the expectation is that it should only fire once for each combustion cycle. Instead we are seeing 4 pulses for a total of 66ms (16.5ms  x 4) of fuel delivery in 720 degrees of rotation!

The waveform below was captured at about 1000 rpm.

Although the compression waveform is showing some variances, my focus is the sporadic firing events of the injector. In this waveform capture there is an obvious issue with the injectors causing the extreme rich conditions, low vacuum and the P0172 code.

The Analysis
I know that outputs, such as injectors, need to have proper “QQT” (Quality, Quantity and Timing) in order to function properly. So, what is the QQT for this injector? 

Does it have proper Quality (electrical values)?

Observation:  the electrical values (power, ground and inductive kick) for each firing of the injector are within normal range

Conclusion: Yes, the electrical Quality is good.

Does it have proper Quantity (amount of fuel delivered)? 

Observation: each individual injector firings have an acceptable pulse width and individually deliver the correct amount of fuel and the ECM seems to be able to control it properly.

Conclusion:  Yes, although the overall quantity is far too high due to the extra pulses, the Quantity per pulse (amount of fuel delivered) appears to be good.

Does it have proper Timing (correct number of events at the correct time)?

Observation: in any 720-degree event or complete 4-stoke cycle, only one injector pulse width is usually required (some newer vehicles can have two). This vehicle has far too many and in some cases skips the event completely.

Conclusion:  No.

QQT Summary:  the injector firing events have proper Quality and Quantity but they do not have proper Timing.

At this point I ask myself:  What input(s) has control or some control over this output? More specifically, what tells the injector when and how often to fire?

It is time to look at the CMP and CKP sensors. Here the CMP is red and the CKP brown:

Below, I am zooming in on the waveform. CMP sensor (red), CKP sensor (brown).

Zoomed in, it is easier to see the signal integrity and synchronization. My experience from looking at many good known waveforms tells me that there is something wrong with the brown CKP signal. It looks upside down. To be sure I pulled out the reference waveform below from the ACE CKP/CMP Signal Reference, by Thompson Automotive Labs. It provides known good examples of 60 different CKP/CMP signals from domestically produced engines.

The CMP signals from the two waveforms are similar (red vs red), however, I see two differences in the CKP signals (brown vs blue): The vehicle’s CKP is upside down and the duty cycles do not match. To confirm I am not making a mistake I double check my scope settings and connections. They were all correct based off the vehicle’s wiring schematic.

The Solution
Upon further inspection of the CKP sensor wiring circuit and reluctor wheel, I determined that the CKP sensor itself must be at fault. After replacing the CKP sensor, the oil, and spark plugs, the vehicle ran good and the vacuum came back up to normal.

The following waveform is the final result!
Green: CKP sensor
Brown: CMP sensor
Red: Cylinder #4 injector pulse width
Blue: In-cylinder compression with pressure transducer on cyl #4

The original problem of rough/rich running with only a P0172 code really did not provide a clear diagnostic direction and several shops wasted a tremendous amount of time on this vehicle chasing that code.

This case study is a simple example of using “outputs to develop a diagnostic direction”.

In regular day-to day-diagnostics, some problems leave behind substantial clues on where to begin. Take for example a vehicle that cranks but won’t start, most technicians will test for things like fuel pressure, ignition, injector pulse and mechanical integrity. These are relatively simple and in many cases don’t require too much in the way of advanced diagnostic routines or tooling.

In the case of “diagnostic trouble codes”, some can point exactly to the problem whereas some codes, such as our P0172, are more like a “postal zip code,” in that they give you an area of failure but don’t point to the exact “address”.

As problems become more elusive/intermittent and complex it is becoming more difficult to determine where to start. Without a diagnostic direction which tool do you use and which circuit/component do you test?
If we broke down module logic to basic operation levels it could be easily divided into five steps. A breakdown in any of these five steps could be the root cause of the failure.

1-Base or default programming - These operations could be non-reliant on some direct sensor inputs, a “default” or “starting point”.

2-Inputs - These will provide the changing information for which the module(s) must adapt. CKP, CMP, VSS, MAF, TPS, AFR, switches, etc. Obviously the list changes dependent on the system that the module controls.

3-Processing - The gathered input data must be calculated/processed out logically to determine response or output actions.

4-Communications - The inputs or calculated data may need to be shared with other modules prior to performing output actions. This is typically performed via the on board data network systems. Keep in mind that even though one module may have performed input gathering/processing, another module(s) may be required to perform actual outputs.

5-Outputs - These are the actual output actions, injector pulse width, ignition coil firing, all types of solenoid, motor operations, etc.

Outputs require proper “QQT” (Quality, Quantity and Timing). Modern test equipment and procedures have given us the ability to properly measure many of these very quickly in many cases. Measuring the output or “end result” can give us the clues or direction necessary to continue our testing. That is exactly what happened with 2002 Chrysler PT Cruiser. Another advantage of using outputs to develop a diagnostic direction is that even though a problem may not be evident at the time of testing, some output testing could still reveal a problem.

Equipment used in this case study
PICO 4-channel lab scope
PICO WPS500X Pressure Transducer
Ckp/Cmp Signal Reference (CD) by Thompson Automotive Labs


Share your thoughts: Log in to post a comment.


Phil Fournier emailed an extended reply, read now.

  1. Brandon Steckler
    11/25/2015 at 06:43:39 PM
    Interesting...thank you for sharing! I couldn't agree with you more regarding the decision to monitor to gain a diagnostic-direction. I find most of the vehicles I've encountered that have been through the ringer; I've pinpointed the issue after gaining a diagnostic-direction via the monitoring of outputs.
  2. rpalom1
    11/26/2015 at 09:52:54 AM
    Good information
  3. Dave Howell
    11/26/2015 at 12:05:46 PM
    Great job! I'll be sending this to my automotive engine diagnostics students!
  4. Dave Comstock
    03/12/2016 at 10:27:02 PM
    Good job Adam! Learning to change my thought process has been the best for my diagnosis! Thanks Adam for all you do, helping the automotive industry to all in can be. Shade tree to car Doctors!