Electronic Fuel Injection Diagnostics
By Mike Grage, Customer and Technical Support at Modern British Iron
“Understanding how the EFI sensor loop works is the foundation for finding a fault in the rainbow-colored spaghetti we call a wiring harness.”
The most common question I get when discussing EFI diagnostics is: How do you know this stuff? For most of my adult life my job title was ‘multi-line motorcycle service technician’. Motorcycle Mechanic’s Institute only gave EFI an honorable mention when I went there in 1990-1991, and that training was for the early EFI systems used by Japanese motorcycle manufacturers. Those EFI systems were not the most dependable, they were not very well understood by dealer service technicians, and they lacked on-board diagnostics. Additional fuel injection training over the years came from Triumph, Kawasaki, BMW, Suzuki, and Harley-Davidson. All that training was a great starting point, but when it comes to EFI training means very little without shop experience. My experience with EFI diagnosis was a regular thing for more than a decade.
“Then there are the dreaded intermittent MIL trouble codes.”
On-board diagnostics help, but you cannot depend on them to solve every problem. There are times when replacing a sensor will not solve the problem that is generating a trouble code. Other times the on-board diagnostic system continues to produce multiple trouble codes after replacing all the sensors that produced the trouble codes. Then there are the dreaded intermittent MIL trouble codes. The scary truth is that most EFI problems are electrical problems and hunting them down can be frustrating. Experience has taught me that sensor faults are usually caused by wiring problems, not the sensor itself.
The best piece of advice I can give is to use an analog multi-meter when testing components. The reason for this is because analog meters do not have a polling rate. The best example of why analog is better is the TPS tests. The throttle position sensor output for Keihin EFI should have 0V with throttle closed, 2.5V at half throttle, and 5V at full throttle. The needle on the analog meter should move smoothly as you open and close the throttle. A sudden drop to 0 off idle indicates a burnt spot in that potentiometer, something like that can be easily missed with a digital meter and trigger intermittent MIL codes.
This blog cannot solve every EFI problem out there. The idea is to give the readers some experience-based advice on where to start. The following are the three types of faults technicians deal with on a regular basis, and the methods I used start dealing with them. The three titles are the long form of what I would like to see a service writer put on a service ticket when a motorcycle comes into a dealership.
1. Troubleshooting No Start/Poor Running when MIL Codes are NOT present:
No start, poor running, poor engine acceleration, with no malfunction indicator light (MIL) codes after the engine starts tends to be a fuel pressure problem. Low fuel pressure is the most common EFI problem, and it will not trigger a malfunction indicator light code (MIL, check engine, FI, engine symbol light, etc.) in most cases. If the MIL light goes out after the engine is cranking, a fuel pressure test is needed. Low fuel pressure is usually caused by a weak fuel pump, bad pressure regulator, or bad fuel pump relay. A voltage check at the fuel pump connector is always a good idea before ordering a new pump, low voltage at that connector is usually caused by a bad fuel pump relay. Although it is a rare occurrence, a gremlin in the harness can cause low fuel pump voltage and pressure.
2. Troubleshooting Malfunction Indicator Light Codes:
If you have one sensor producing codes replacing that sensor should solve the problem. I suggest checking the connector and terminals before replacing that sensor. One habit I have is pushing wires into connectors after I get the connector to click to make sure the terminals have a solid connection. Missing that a tiny metal tab broke off a terminal is an intermittent MIL code nightmare. It takes 3 cycles to trigger a code and 40 cycles for the code to go away, so sometimes pushing a wire in and then starting the bike will clear the code and solve the problem. Terminal connections can come loose from vibrations. Just because you don’t feel it in your seat doesn’t mean your bike doesn’t have screw and connector loosening vibrations. Checking to make sure all fasteners are tight is part of a regular service on all motorcycles due to the high frequency vibrations.
3. Troubleshooting Multiple MIL codes:
I wrote a companion piece to this blog on the breakdown of OBD-II codes. Go read it first if you have codes caused by multiple sensors first. The sensor loop codes always start with P01xx or P11xx, multiple sensors producing codes with those designations are generally a single bad connection or a single bad sensor. Coolant and MAP sensors are notorious for causing multiple MIL codes. However, multiple system codes (ex. P01xx and P03xx or P03xx and C16xx) are usually wiring related. The wiring on motorcycles is not very protected from corrosion, and wires corrode. EFI is more prone to corrosion related malfunctions because it works with 5V DC circuitry. The part of the circuit that is most prone to corrosion are the case and frame grounding points. A bad ground for the sensor loop is the type of nightmare that will make wrenches fly across the shop and has the potential to cause alcoholism in motorcycle technicians. I remember the first time I dealt with this problem like it was yesterday. In 2001 a Triumph Daytona 955i rolled onto my lift for a warranty repair, the problem it had was so baffling that I was the third technician to get stuck with that demonic motorcycle. The idle speed control motor was bringing the idle up and down. The ECU produced multiple MIL codes, and then the bike would shut down when it reached operating temperature and would not restart until it cooled down completely. All the EFI sensors were replaced by the two previous techs and the Hinckley factory continued to throw parts at it to the point that they sent a new ECU for me to install, none of which solved the problem and all of that made me reconsider my career choices. Instead of giving up like the other two techs I decided to go old school. Got out the analog Micronta meter and started checking the individual EFI component readings at the ECU connector and tugging on wires and the wiring harness. A funny thing happened when I checked one of the black wire’s continuity to ground, the needle twitched every time I tugged on the harness. There was a loose engine ground wire, replacing the eye connector on that wire fixed the bike. It only took 6 months, 3 technicians, and 1 smashed computer monitor to find that loose (not even broken) wire. That bike taught me a lot about EFI diagnostics, and it taught me that throwing a wrench in the general direction of a computer is a bad idea.
Here’s the explanation of what was happening there. The EFI sensors operate on a 5V DC reference voltage which makes them more prone to failures from loose/bad connections. Understanding how the EFI sensor loop works is the foundation for finding a fault in the rainbow-colored spaghetti we call a wiring harness. The ECU has two modes, open loop and closed loop. The ECU takes readings from a group of sensors in a specific order in closed loop mode to determine fuel mixture/trim and it ignores the sensors in open loop mode. The specific order is why faulty readings from the coolant temperature or oxygen (lamda) sensor can trigger other MIL codes. The loose wire on the 955i was the ground for the sensor loop and ISC stepper motor circuits.
Those are the most common conditions and starting points for EFI faults. There are a lot of things that can go wrong in the rainbow-colored spaghetti called a wiring harness. There are over 5,000 OBD-II codes and I am not going write a blog on the causes and cures for each one of them anytime soon. Honestly, tugging on wiring harnesses and checking connectors is a good first step towards preventing heart attacks and strokes associated with EFI problems. Hope you enjoyed this, and I hope it helps.
Also, check out our blog on Explanation of EFI Diagnostic Trouble Codes for OBD-II for more information.