Much like the automotive repair industry, emission regulations and On-Board Diagnostic (OBDII) technologies used to monitor and diagnose the components that control them advance at an incredible pace. Sometimes, as technicians, we lose sight of the difficulties and challenges faced by the manufacturers when they design a system. The SAE On-Board Diagnostics Symposium provides technicians with the opportunity to interact with the individuals responsible for making the regulations and the engineers who design the systems with which we interact on a daily basis.
In September of each year, the symposium attracts engineers and industry professionals to network and exchange their experiences in OBD with the California Air Resources Board (CARB) and other industry experts. In turn, CARB provides updates on regulations and standards responsible for establishing emissions requirements.
Last year’s event in Anaheim, Calif., featured three days of technical sessions, networking breaks and luncheons, where attendees discussed topics and asked questions of the symposium organizers and guest speakers. These included engineers from General Motors, Ford, Chrysler, Toyota, Volkswagen and Cummins. Because of the event’s proximity to CARB’s headquarters, there were several team members from that organization present at the event. Attendees shared experiences and discussed new or impending regulations and standards with the group.
The first SAE OBD Technical Workshop was organized by General Motors in August 1991 in Romulus, Mich. The event was formed to allow CARB to assess where all the manufacturers were with regard to OBD development. The annual event grew to more than 700 attendees by April 1994. The symposium eventually expanded to include heavy-duty content and moved to Indianapolis in August 1999. Since then, it has since rotated among major cities featuring both automotive and heavy-duty industries.
The evolution of OBD standards and regulations
The original OBDI regulations totaled three pages in length. Today’s regulations have become much more substantial given the technological advancements and complexities that have occurred over time. Additions to the regulations and associated standards include oxygen sensor monitoring, diagnostic trouble codes, isolation of fault on inputs, in-use rates, Evaporative Emission Control System (EVAP) leak detection, comprehensive component fault isolation, six-pattern testing, rationality vs. functionality diagnostics, smart devices and more.
In the late 1980s, most industry experts thought that the key to emissions control was the O2 sensor. In fact, it was proposed by the SAE Vehicle E/E System Diagnostics Committee that all wires for the O2 sensor be extended to the vehicle’s diagnostic connector (J1962 Connector). The members from that group actually developed Mode $05-Oxygen Sensor Monitoring so that the same data/test results could be read from a generic scan tool.
The SAE J2012 Diagnostic Trouble Code (DTC) Definitions technical report was written in 1990. It defines the standardized DTCs that the OBDI system is required to report when malfunctions are detected. The original document had a total of 152 DTCs. To put that number in perspective, today we have more than 6,000 DTCs.
In 2003, a CARB regulation implemented isolation of faults on inputs, resulting in the circuit open, circuit shorted, and etc. codes.
In-use rates is a familiar term often used with OBDII development. It came about when CARB determined that some manufacturers designed component self-tests to detect faults when performing the Federal Test Procedure (FTP) drive cycle. Unfortunately, they were unable to detect faults during normal customer driving. As a result, CARB set a regulation in which it specified the enabling criteria for a monitor to run. Another group proposed the in-use rate calculation, which uses software to calculate the parameters and also required the monitor to run at specific times.
EVAP leak detection came about when emissions data research showed that any leak greater than 0.008” was considered uncontrolled. This equated to approximately 20 grams of hydrocarbons or more entering the atmosphere each day from the vehicle. Several methods to detect the small leak (less than 0.020”) have been tested such as ultrasonic testing — in which air is injected into the system — HC detectors and smoke. Even today, many experts contend that 0.020” is the smallest size leak that technicians will be able to detect in the field.
Another term that technicians may hear, but are unfamiliar with is six-pattern testing. This basically refers to the procedure used to test O2s, as they can fail in six different ways:
· Symmetric Slow Response Rich
· Symmetric Slow Response Lean
· Asymmetric Slow Response Rich
· Asymmetric Slow Response Lean
· Delayed Response Rich
· Delayed Response Lean
Mode $06 has been standardized with CAN, which includes the tests from Mode $05 and is used primarily by CARB for testing and enforcement. It is used for non-continuous monitors. Mode $07 is used for continuous monitors. Another area mentioned in Mode $06 is the Fast Pass, which means that when the threshold for the test is reached, it stops running. An area of confusion for technicians is when they look at the test results and see that the passing result was very close to the threshold. Many take this to mean that the test was close to failing, which is incorrect; it means that once the minimum value was achieved, the test stopped.
Another area of concern to OBDII designers was Input Comprehensive Component Fault Isolation Requirements. This dealt with the difference between rationality fault codes and circuit faults codes. Questions posed for discussion by previous symposium attendees were how much fault isolation is required and also if technicians actually use the added fault detection. The example given was for O2 DTCs; there are 11 DTCs for the front O2 sensor, of which five are circuit codes. The effectiveness of this method was debated at length.
The OBDII regulations define rationality diagnostics being for input sensors and functionality diagnostics for actuators. Technicians often become frustrated following the diagnostic trouble code chart and cannot come to a conclusive determination of the fault code. However, it has been discovered by industry experts that a no-fault found during diagnostics typically averages an astonishing 80 percent among most manufacturers.
OBD discussions at the annual symposium
At the last SAE OBD Symposium, Mike Regenfuss from CARB gave an informative presentation on the OBDII Program Update. Items of interest include:
1. LEV III thresholds for an emission malfunction
2. Changes required for hybrid vehicles
· Monitoring the hybrid battery, including being able to narrow down the fault. It is still undetermined how precise the fault detection must be (cell, module or pack)
· Monitoring of the hybrid cooling system, such as the inverter and actively controlled outputs that maintain the temperature of the cooling system
· Monitoring the performance of the generator, electric motor and regenerative braking system.
· Misfire Monitor for Plug-in Hybrids
3. Smart devices
· These are micro-controlled input/output devices that do not have enough impact to be considered an emission critical module, but are subject to OBDII regulation when the input or output are an emission monitor, comprehensive component or another device that controls an emission component
4. Non MIL guidelines
· Only available to Comprehensive Components
· Requirement is that the systems use is continued in such a way as to have no increase in emissions and the diagnostic monitoring remains functional
· Default strategy must remain in effect until the monitor passes on a successful drive cycle or the fault is cleared by a scan tool
Standardizing Fault Codes
· OBDII currently uses a 2-byte structure for fault codes, but this has a limited number of available combinations
· Due to the advances in pinpointing faults, about half of the available codes have been used up.
· Options to address this are using a 3 or 4-byte fault code structure or even developing a completely new fault code structure
Also discussed were additional Data PIDs to evaluate real-world fuel economy, driving habits and performance of off-cycle technologies (cumulative all electric miles, cumulative fuel consumed, cumulative system active time).
Monitors that are problematic will be reviewed, such as EVAP and PCV as these will present major challenges for LEV III emission requirements. Purge Under Boost is hard to accurately control, so it may be able to test out. The PCV system monitor tests for breaks in lines, but does not guarantee the effectiveness of the system.
Paul Baltusis from Ford shared his experiences over the last year with CARB. One area where Ford systems had an issue was with Dual Path Purge Monitoring. Basically this means there is one purge line for low-load operation and a second high-pressure line that is used during boost conditions. The problem with the system was that the high-pressure line was not monitored for leaks. Baltusis then described how the system was redesigned to comply with CARB’s requirement. A similar situation with their PCV system also existed where the section of the PCV line that was under vacuum during normal operation was monitored for leaks; however, the pressure line used during turbocharging was not monitored for leaks, so there would be no way for the system to know if the line was broken or not connected, allowing crankcase vapors to escape to the atmosphere.
Bob Gruszczynski from Volkswagen presented some background on how an ECU is determined to be an OBDII ECU or not and surprisingly a total of seven different departments are involved in the process. There are only eight CAN IDs available for a vehicles network, meaning that the choice of which ECUs are a primary OBDII ECU must be chosen carefully.
OBDII communication encompasses the entire vehicle; it does not pertain to a certain ECU. Something the insurance and telematics groups do not understand is that when they query the vehicle, they get more than one response to the same question. This differs from a generic scan tool, since the scan tool realizes it will get more than one response, but only displays one answer. Generic scan tools only communicate with Primary OBDII ECUs.
Gruszczynski also provided some insight into various possibilities to OBD Architecture, such as Smart Code Clear vs. Functional Code Clears. A benefit of the Smart Code Clear is that it only deletes the data and resets the monitor that had the fault, and also any other monitor that is dependent on the item that had the fault. This way other adaptive values and monitors do not have to be relearned or rerun.
Toyota’s Mort Smith presented some of his company’s OBDII experiences. He stated the industry currently had LEV (Low Emission Vehicle) II standards, which have been in place since the 2004 model year. Vehicles manufactured from 2015-2025 will be under a new category called LEV III.
Smith also discussed the Air/Fuel Ratio Imbalancing Monitor Requirements, which started to be incorporated in the 2011 model year vehicles. All current model year (2015) vehicles should now be compliant with the final cut point thresholds. Also warmup catalysts will need to be able to light off faster, which will require them to be closer to the engine on most vehicles. Because a turbocharged engine has a large heat mass, it will not heat up a catalyst as fast as a naturally aspirated engine. As a result, this requires the catalyst to be moved even closer to the exhaust ports on a turbo vehicle. However this does create conflict due to the fact that as the warm-up catalyst is moved closer, increased use of over temperature protection must be used, which is basically enrichment of the fuel mixture. On the downside, this will increase CO emissions. Other ways to cool the catalyst were also discussed, such as liquid exhaust cooling.
Something rarely considered when evaluating a certain design is that one change has a domino effect. For example, by moving the catalyst closer to the manifold, the design of the exhaust manifold must be changed. When the exhaust manifold runners become shortened, the placement of the air/fuel (A/F) ratio sensors becomes more important because they need to be located where there is uniformity of exhaust flow from all cylinders on the bank. This could impact the accuracy of the sensor if not placed in an exact location and could also impact the A/F ratio monitor itself.
Jaguar Land Rover (JLR) has also had its share of experiences with OBDII. Martin Haggett’s example was that they have had to adjust the MAF Sensor Rationality Monitor to cover a wide range since it is a comprehensive component monitor. Standard MAF sensors use an additive threshold at lower air flows and a multiplicative threshold at higher speeds.
Unfortunately, when more of the JLR vehicles started using the twin MAF setup, some changes needed to be made to fit them into the new 2010 vehicles. This change resulted in false detection of MAF sensor rationality faults. The airflow difference between the banks varied significantly at times, resulting in codes being set. After some research, JLR engineers determined that the miscalculation was caused by side winds, which had a dramatic effect due to the positioning of the air intakes on the vehicle. The side winds created enough of a pressure difference between the two sides to cause the fault.
SAE recently shared updates on its latest efforts on J1979. The standard outlines the communication requirements between OBDII and test equipment with regard to emissions. Some interesting notes are the time to EVAP Decision PID. Basically, this means that if a vehicle is brought in for testing and the OBDII system has completed the complete evaporative system monitor in the last 750 miles and there are no pending or present leak codes, the system is considered to be working correctly. If the monitor has not run in the last 750 miles and there are no pending or present codes, then the manufacturer will have the option to run the test in the lab after meeting the enabling conditions.
Some focus was put on generic OBDII starting to use Unified Diagnostic Services (UDS), which is what several manufacturers are already using in their systems. UDS specifies the data link requirements in vehicles and is covered in ISO 14229. It allows the diagnostics to control the functions of a module. This covers data transmission, communications, input and output controls, in-bay services tests, etc. In addition, UDS would also support the 3-byte DTCs mentioned earlier.
It is predicted by many industry analysts that before 2020, we will run out of DTCs, one of the big issues solved by using UDS. It is independent of protocol.
The SAE J1979, which defines the standards for the diagnostic test modes, was invented in 1988 and many in the industry believe it is time to move away from that technology. Many technicians hold the opinion that enhanced data is needed to correctly diagnose a vehicle and even for system maintenance procedures.
OBDII: The service perspective
Not all presenters at the annual symposium were OEM engineers. Jason Smith from Snap-on Diagnostics and Steve Caruso from OBD2Training.com presented a class from the service perspective by demonstrating anomalies found when diagnosing vehicle failures.
Smith presented a case study where a variable cooling fan set throttle body DTCs. This caused the vehicle to revert to a reduced power mode. All the possible causes for the codes stored and tests that they were instructed to perform due to those codes caused the technician to replace the electronic throttle control.
Smith also showed what a Google search found when the problem for the codes had been entered for that vehicle. It demonstrated the impact of what happens when diagnostics of the stored trouble codes are misleading. The wiring diagrams listed in the service information systems show no connection between the cooling fan and the throttle actuator, so the technician has no idea the systems are related. According to Smith, some of these issues can be addressed through an industry-wide effort to improve training and education.
Caruso presented a case study where a vehicle passed the OBDII portion of California’s emissions test, but failed the tailpipe portion. The tailpipe was emitting black smoke and the CO reading was 7.2 percent. Parts replaced at other shops before the vehicle was inspected were: catalytic converter, PCM and front O2 sensors. All signs pointed to a defective PCM, since there were no codes, the vehicle passed all the OBDII Monitors and was running very rich. However, after further analysis, Caruso made an interesting discovery. The front O2 sensor, which was already replaced a total of three times by other repair facilities, was reading 800-900mV; however the rear O2 sensor was stuck at only 100mV. The manufacturer of that vehicle did not list any service information about the rear O2 sensor having influence or control on fuel trim. It only tested for activity that had occurred when the vehicle was started, but then tapered off until it sat at low voltage.
As a result of these discussions, Smith recommended creating a database collection of anomalies; and finding resources to get technicians and educators to better prepare themselves to deal with the more complicated systems. According to Smith, this can only be achieved by manufacturers and the aftermarket working together: “Collaboration can be more effective than competition.”
The last day of the annual symposium focused on the heavy-duty industry applications. Jeffrey Potts, from Cummins Inc., gave the update on the heavy-duty Industry and some of the specific challenges they face. Of interesting note was the way the heavy-duty industry approaches remote fault isolation, which we will see occurring more often in the automotive industry in the future. Vehicle down time is much more crucial to the heavy-duty industry since the vehicles are a source of income, not just convenience.
The Next OBD Symposium
The SAE 2015 On-Board Diagnostics Symposium will be held Sep. 15-17 at the Indianapolis Marriott Downtown in Indianapolis, Ind. The Monday before the event begins, John Van Gilder from General Motors will conduct a one-day course, “Emissions-Related OBDII Systems: A Design Overview.” The course covers topics including fundamental design objectives for OBDII systems, basic design features, defining "good" vs. “bad” systems and anatomy of an OBDII. I had attended the class two years ago and gained a new and valuable perspective on OBDII systems from it.
In addition, symposium organizers such as Gruszczynski will be looking to continue to add focus on the serviceability aspects of OBDII to future events. As noted, the increased system complexity added via electrification and other new technologies required to meet ever-more stringent emissions regulations will place a larger burden on the service community. The knowledge exchange provided by the SAE OBDII Symposium is a great starting point to ensure that engineers and technicians share the same understanding of those systems and technologies