Service Repair

Cold starts cause Mazda 3 to misfire

Feb. 13, 2025

A 2008 Mazda 3, with a 2.3L engine entered the shop with a complaint of rough-running and an MIL illuminated. The only DTC stored: random misfire detected.

Brandon Steckler

Figure 2- Suspect of a lean condition, additional hydrocarbon in the form of propane was used as an experiment for enrichment. The performance of the engine (via exhaust gas analysis) will be evaluated with the enrichment occurring. This will help determine the next step to take.

Figure 1- The results of the exhaust gas analysis reveal several clues that act as puzzle pieces, if you know how to interpret the gases. HCs (Hydrocarbons) are unburnt fuel. There are over 5,000 ppm (part per million) indicating that leftover fuel is in the exhaust stream. NOx (Oxides of nitrogen) are dangerous gases that occur in abundance when temperatures exceed 2500 degrees F. These appear to be in control. CO2 is an indicator of engine efficiency. A healthy engine typically yields about 15%. This engine is not performing well. Finally, normal engine combustion and a functioning catalyst should use almost all the oxygen that entered the engine. This is clearly not the case, here.

All too often a client brings his or her vehicle to us to rectify a driveability concern. Although we become very familiar with the symptoms of many driveability concerns over time, the root causes of these concerns can vary widely.

With the vast array of years/makes/models and engine configurations out there, it’s no wonder that so many technicians face “comebacks.” There is just so much to learn and unless you have a game plan, you’ll find yourself shooting from the hip, and the odds of long-term success are not in your favor.

You need to employ tried and true testing techniques. And to do that, you must become familiar and comfortable with those tests, as well as understand the results of the tests. I can’t think of a better tool or testing technique than employing exhaust gas analysis. The reason? The test technique applies to all internal combustion engines out there. This is due to the chemistry involved.

Today’s Subject Vehicle

A 2008 Mazda 3, with a 2.3L engine entered the shop with a complaint of rough-running and an MIL illuminated. The DTCs were scanned and stored was only a P0300 (“Random misfire detected.”) Upon questioning the customer, the driveability concern only seemed to surface upon cold-starts. The vehicle was in another shop recently, for replacement of the timing chain and related components. The vehicle indeed left the previous repair facility in good running condition. No driveability symptoms were present after that repair.

The decision to employ a five-gas analyzer was made. This test offers a tremendous amount of information simply by measuring and analyzing the exhaust gas content, right at the tailpipe. This test revolves around science (chemistry, to be precise) and we can’t change what science proves. This means we can rely on the test results no matter which internal combustion engine is being evaluated, or the fuel being used.

The Results Are in

Capturing the exhaust gas while running the vehicle through a cold-start, the driveability symptom surfaced. As expected, the fault only occurred when cold and in open-loop. As the engine warmed and closed loop occurred, the better the performance was exhibited. The gases were analyzed, and the results were reviewed (Figure 1).

Please see the results below:

Hydrocarbon (HC) = 5342 ppm

Carbon Monoxide (CO) = 0.24%

Carbon Dioxide (CO2) = 7.3%

Oxygen (O2) = 9.7%

Lambda = 1.30 (per calculator)

Figure 3- While evaluating the exhaust gases when supplemental propane enrichment occurs, the drastic improvements in the exhaust gas content are visible. Although additional hydrocarbon was introduced, the HC content in the exhaust stream dropped. Propane enrichment supported combustion and assisted in the events consuming hydrocarbon. As a result, better combustion used more oxygen and reduced what was left over. The better running engine yielded better efficiency, and this can be seen with the drastic rise in carbon dioxide levels.

A Logical Experiment

After evaluating the gas analysis, a decision was made to introduce propane, a vaporized/atomized hydrocarbon (Figure 2). The vehicle was allowed to cool, and the engine was restarted to exhibit the fault. The propane was bled into the induction system in a controlled fashion and the captured exhaust gases were analyzed again (Figure 3).

Hydrocarbon (HC) = 380 ppm

Carbon Monoxide (CO) = 1.13%

Carbon Dioxide (CO2) = 12.9%

Oxygen (O2) = 1.8%

Lambda = 1.03 (per calculator)

The Data Doesn’t Lie

With all the information in front of us, and the desired information not yet obtained, we are faced with deciding how to proceed. Here are some bullet points of what we know to be factual, and I will ask all of you, diligent readers, for your input on what they mean to you, collectively:  

MIL illuminated / DTC P0300
Engine runs rough at cold start-up (open-loop)
Engine runs normally when warm (closed-loop)
Exhaust gas analysis performed/ results improved with propane enrichment

Given this information, what would you do next?  

Replace spark plugs and/or COPs.
Inspect timing components.
Evaluate fuel delivery system.
Decarbonize induction system.

About the Author

Brandon Steckler | Technical Editor | Motor Age

Brandon began his career in Northampton County Community College in Bethlehem, Pennsylvania, where he was a student of GM’s Automotive Service Educational program. In 2001, he graduated top of his class and earned the GM Leadership award for his efforts. He later began working as a technician at a Saturn dealership in Reading, Pennsylvania, where he quickly attained Master Technician status. He later transitioned to working with Hondas, where he aggressively worked to attain another Master Technician status.

Always having a passion for a full understanding of system/component functionality, he rapidly earned a reputation for deciphering strange failures at an efficient pace and became known as an information specialist among the staff and peers at the dealership. In search of new challenges, he transitioned away from the dealership and to the independent world, where he specialized in diagnostics and driveability.

Today, he is an instructor with both Carquest Technical Institute and Worldpac Training Institute. Along with beta testing for Automotive Test Solutions, he develops curriculum/submits case studies for educational purposes. Through Steckler Automotive Technical Services, LLC., Brandon also provides telephone and live technical support, as well as private training, for technicians all across the world.

Brandon holds ASE certifications A1-A9 as well as C1 (Service Consultant). He is certified as an Advanced Level Specialist in L1 (Advanced Engine Performance), L2 (Advanced Diesel Engine Performance), L3 (Hybrid/EV Specialist), L4 (ADAS) and xEV-Level 2 (Technician electrical safety).

He contributes weekly to Facebook automotive chat groups, has authored several books and classes, and truly enjoys traveling across the globe to help other technicians attain a level of understanding that will serve them well throughout their careers.