One of the systems that Subaru has perfected through improved reliability over the years has been its distributorless ignition system (DIS) for all of its models. Because of the horizontally opposed design of its engines, the ignition system can be centrally mounted and made simple.
In this article, we will touch on some common diagnostic tips, as well as the more technical tests of the components in the relationship between the ignition system and related sensors. We’ll also provide insight into some of the more frequent failures that we encounter, how we arrive at the diagnosis and make the repair. This article will cover the Legacy series, as well as the later Impreza, Outback and Forester models. The systems are pretty standardized for all of these models, except for pre-1995 models and the 2.5L DOHC engine.
The Subaru DIS is fully controlled through the Engine Control Module using inputs from several sensors. Ignition timing is controlled after receiving data from cam and crankshaft sensors, the engine coolant temperature sensor, air flow sensor, throttle position and knock sensors. The ECM analyzes this data and transmits a current off signal to the power transistor to control ignition and timing. On all models, the timing is set at 10° BTDC for starting.
The Legacy (typical) system consists of two ignition coils in a single housing and a power transistor assembly. The power transistor assembly contains two transistors that control the ground path for the coils. One transistor controls the coil for #1 and #2 cylinders; the other for #3 and #4. Each coil fires two spark plugs at the same time. Ignition takes place in only one cylinder though, preferably the one on compression stroke. The ECM signals the appropriate power transistor based on the inputs from the crankshaft and camshaft position sensors. The coils have power any time the ignition relay is energized.
Evolving Testing Techniques
In the days before electronic and distributorless ignition systems, we were able to isolate non-firing cylinders with various “Bubba” tests. By pulling the plug wires off their spark plugs one at a time, we could easily find a non-firing cylinder and, sometimes, get a recharge for our heart at the same time!
Note: With today’s higher-powered ignitions, this can cause damage to the system. A better way to isolate a weak or inoperative cylinder is with an inductive-type spark tester touched to individual wires to indicate a misfire. These testers work especially well on the Subaru system where the spark plug wires on some models are almost impossible to remove at the plug with the engine running (see Photos 1 and 2).
If you are working on an OBD II car, and you have a scanner that can access data, you also can check firing KV or injector pulse width to locate a “dead” cylinder.
We get quite a few Subarus in with DTCs for single- and multiple-cylinder misfires. They are generic codes such as P0300 to P0304. There can be a number of causes for these codes, but there are some basic tests that should be made before going too far down the wrong road. The spark plugs don’t usually go more than 30,000 miles without being worn enough to impact driveability. Even the platinum-tipped plugs in the 2.5L DOHC engines are questionable over 45,000 miles. With excessive resistance caused by worn plugs, the spark has a hard time determining the easiest route to ground, often going instead to the cylinder opposite the one on compression.
The other part of the resistance equation is the plug wires. Many times, the plug wires are so high in resistance after 50,000 miles that, again, the spark misses its intended path. Check the wires with your VOM. The inductive spark tester can help pinpoint weak spots. In both cases, with spark plugs and wires, use the correct type and don’t try to save a few bucks by buying less than the best quality.
Obviously, there are mechanical reasons for a misfire, as well, although usually more than a misfire code will be set if there is a loss of compression, cam timing change or a vacuum leak. If you get cam position sensor, fuel trim or other adaptive codes, or there is a noticeable misfire, a compression test is possibly warranted. The compression test can be tricky especially on 2.5L engines, so don’t waste much time here unless there is an indication of previous repairs (such as timing belt replacements, valve adjustments) that may have caused or didn’t correct an internal failure.
One other problem that we have found quite often is an injector failure that leads you to make numerous attempts at resolving an ignition problem, only to find that the spark was going to the wrong cylinder because there was no fuel to ignite in the correct one. Remember that the spark is going to take the easiest route to ground and that might turn out to be the opposing cylinder. The inductive spark tester comes in handy when it shows no spark activity to a cylinder with a good plug, wire and compression, but the same coil is sending plenty of spark twice as often to the cylinder that’s directly opposite. Using an injector probe is an easy way to determine that the injector is pulsing (see Photo 3). If there is no solenoid action, disconnect the injector and use a noid light to check for a signal. If a signal exists, pull the plug at that cylinder and check for unburned fuel. If there is no fuel, you probably need a replacement for the injector. This has been the source of many suspected ignition problems on the Subarus we work on.
We have found that the cam and crank sensors do fail from time to time, and they usually will set codes. We have also found that just by disconnecting and reconnecting them, in a no-start situation, it will often make the car start. Let your customer know if this happens. It usually is a safe bet that it will cost a tow bill if left uncorrected. The cam and crank sensors have come way down in price over the last few years, so replacement can be more easily sold as maintenance on high-mileage cars. Just recently, Subaru has released a set of sensors (cam and crank together) in the same box for $40. The part number for 2.5L engines is 22090AA000. They are easy to replace and carrying a set on a trip, just in case, makes good sense for some people. We probably replace at least three cam sensors to every crank sensor.
The ECM and transistor units have never been trouble spots. Unless there are some other causes such as shorted wiring, battery mistakes or crash damage, they are not failure prone. We couldn’t remember ever replacing a computer or transistor module on a Legacy series car in our shop.
Checking Connections and Sensors
If the “Bubba” test for the ignition coil above doesn’t show any faults, you’ll need to pull out your VOM and go to work. Primary circuits can be checked at the coil harness connector (see Photo 4). Ignition coil primary resistance should be checked with the coil connector disconnected. The resistance value should be between 0.62 and 0.80 ohms for all but the 2.5L engines. The 2.5L coil will have primary resistance of 0.40 to 1.00 ohms. Check the secondary circuits for resistance between two opposing secondary wire terminals. The secondary resistance will be somewhere between 10k and 24k/ohms, depending on the model.
The spark plug wires can and should be checked for resistance. I do this test any time that I have the plugs out, because wire resistance is so important to the operation of the ignition. Because the possibility of a failed cap or rotor is eliminated, there has to be something else that will fail over time. Plug wire resistance will vary by wire length but, for all models, will fall between 5k and 15k/ohms. Look for black burn-through spots and also inspect for rodent chews.
If you have a no-start condition and only get a DTC P0336, Crankshaft Position Sensor (range or performance problem), replace the sensor.
If there is a P0335, CKP Sensor Circuit (malfunction), you can test the sensor after inspecting the wiring harness for damage and checking to make sure that the sensor retaining bolt is tight. Check for shorts in the harness to the ECM by removing the connector from the sensor. Check resistance between terminal 1 (black wire) and ground. If there is less than 10 ohms resistance, find and repair the shorted harness. If the harness checks OK, check resistance between terminal 2 (white wire) and ground, or remove the sensor and test between both terminals. If resistance is between 1k and 4k/ohms, the sensor is probably OK. If the resistance is lower or higher, replace the sensor.
A DTC of P0340, Cam Position Sensor (malfunction), can be checked in the same way that CKP sensor was checked above, using terminal 2 (white wire) as well. The CMP should also have a resistance of 1k to 4 k/ohms. Replace the sensor if it is out of range or only DTC P0341 exists.
The Knock Sensor will have a DTC of P0325. If the harness appears intact, test the sensor at its disconnected harness connector. Check resistance between terminal 2 and ground. The Knock Sensor should have at least 400k/ohms of resistance.
If you suspect an injector is at fault with DTCs P0262, 265, 268 or 271, and testing with an inductive probe doesn’t convince you, disconnect the harness connector at the suspect injector. Check resistance between injector terminal 1 and 2. Injector resistance for all models should be between 5 and 20 ohms. If resistance is out of range, replace the injector. If injector resistance is OK, inspect and test the harness to ECM. There also is a TSB related to rough running caused by mixing up the connectors at the injectors. As the TSB indicates, this may seem too obvious, but it does happen! Use the wiring diagram if you suspect this may have happened.
There are other components that can be tested in this system, but to be honest, these systems have been so reliable that if you haven’t found the problem by now, the car you’re working on has a pretty odd problem. The other components involved with ignition, including the Throttle Position Sensor, Air flow Meter and Engine Coolant Temp Sensor, have not been problems and usually will set an obvious code if a problem does exist.