ECU soft resets: Killing power to on-board modules does wonders in diagnosis and repair
Craig Truglia is an ASE A6 and A8 certified technician who presently works as a service writer for Patterson Auto Body, a repair facility in Patterson, N.Y. A former shop owner and editor of several automotive repair magazines, Truglia combines his Columbia University education with the real-world experience he sees daily in the automotive repair field. Technicians Truglia, Fred Byron and Cristian Cordova took part in diagnosing the different vehicles in this article.
When a smart phone is not working quite right or a PC gets frozen, what’s the easiest way to fix it? Turn it off, start it back up, and hope the problem is gone. Some PC users even wait a few seconds to turn the device back on, hoping it gives time for the computer chips to power down.
Without getting all scientific about why this works on computers, the principle is simple. When a computer acts screwy, sometimes it needs to be reset so it can work right (see Figure 1).
Being that today’s vehicle systems are computer-managed, it stands to reason that when they have weird sensor or computer-related driveability problems, resetting the vehicle’s computers may be the way to go.
Resetting on-board ECUs
Some technicians prefer removing the battery cables and jumping them for 30 minutes. Others like getting fancy and putting a one ohm 10 watt resistor in series and powering down the vehicle for a few minutes. The same is true of both strategies: A soft reset is known to work wonders.
Some high-end scan tools that enable reflashes, such as the Autologic, feature convenient help-lines. When calling the help desk after a failure reflashing, they will recommend that the technician perform a soft reset first. Further, many mobile diagnosticians will not even agree to take the time to look at a vehicle unless a soft reset was done first just to rule out a “computer glitch.”
Simply “deleting the DTCs” or “resetting adapts” does not solve the issue. Unlike computers, vehicles do not come with reset buttons. So, the soft reset remains the only way to get the job done.
It also does not hurt that doing a soft reset helps vehicles set their monitors quicker for inspection purposes, too. It should be noted that as more and more vehicles require idle, steering angle, and other relearns, we do not want to do soft resets haphazardly, especially if we do not have the proper scan tools to perform these relearn procedures.
2011 Buick Lucerne 3.9L speedometer moves erratically
A fleet vehicle was dropped off due to a speedometer which would erratically report different speeds. The vehicle was test driven and it did not act up for a long time. When it did, the scan tool picked up that the Vehicle Speed Sensor (VSS) parameter identification data (PID) was giving the same numbers that were being reflected on the instrument cluster.
A call was made to the dealer to see if they stocked a VSS. When they said that they had one the technician figured, “If the dealer stocks it, they go bad all the time.” The new VSS was plugged in, the problem did not occur, and the vehicle was considered fixed.
One day later, the customer went on a 160-mile drive and reported that it was “worse than before.”
The Lucerne was back and the first thing the technician tested for was to see if the VSS was defective. However, the sensor itself tested at 1.926 kohms, which is within specifications. Granted, checking resistance on an intermittently bad sensor does not prove anything, so it appeared wise to test out the circuit to see if we could narrow down things outside of presuming a sensor from the dealer was bad.
Using a wiring diagram, it was discerned that the transmission control module (TCM) was a separate module on this vehicle and that via serial data, it communicated VSS data to the powertrain control module (PCM) and instrument cluster. So, the best place to test first was the wiring to the TCM. The diagram showed that a purple wire was the VSS low signal and that the yellow/grey wire was the circuit high (see Figure 2). The wires tested out fine, with a resistance of less than an ohm each, generally settling at 0.4 ohms (see Figures 3, 4 and 5).
After setting up the vehicle to be test driven with a labscope, the labscope did not pick up any changes in the VSS signal when the instrument cluster was acting up. Even though the vehicle was a 2011, the only logical issue could be with the TCM giving incorrect data to the PCM and cluster (being that the PCM’s VSS PID would act up when the cluster did.
Before selling a module, and because interpreting serial data is close to impossible, a soft reset was done to the vehicle (see Figure 6). After the soft reset, the vehicle did not act up. The driver was informed if the vehicle were to act up again, that a TCM replacement would be recommended. However, it has been over three months and the vehicle continues to drive problem-free.
2005 Toyota Matrix P0171B
A first-time customer rolled by the shop with a check engine light. It was a Toyota four-cylinder with a P0171 DTC. This is a very common issue on these vehicles.
Here are the common problems: a TSB for bad PCMs (and a recall to replace them), bad intake manifold gaskets, and bad MAF sensors. Of course, major vacuum leaks and other causes are possible, but with the Toyota 1.8L engines these are extremely common (see Figure 7).
Because these Toyota lean codes are so common, they can be diagnosed very quickly. First, the telltale sign of a bad intake manifold gasket is to check freeze frame. It is wise to look for highly positive LTFT (long term fuel trim) and the ECT at ambient temperature, generally during cold weather conditions. If the freeze frame shows this, but the vehicle presently has good STFT and LTFT, even though the DTC is on, the condition is caused by the intake manifold gasket. Generally, the moment the engine gets warm, the gasket expands and seals up everything, and fuel trim returns to normal. A smoke test will not catch these.
Now, if freeze frame does not lead one to a faulty intake manifold gasket, it’s a good idea to check LTFT to see if it is highly positive in datastream. If it is, graph STFT and go WOT (wide open throttle). If STFT goes from highly positive to about zero, there is a massive vacuum leak. If there is not, the likely culprit is a bad MAF sensor.
Now, before replacing the MAF, there’s one more quick test to perform that only takes about 20 seconds. Go to the calculated load PID and when graphing it go WOT. If the PID does not go higher than 90%, as long as you are not at extreme altitudes such as in Colorado (in that case the technician will need to consult a volumetric efficiency calculator), you’ve just diagnosed a bad MAF sensor.
Suffice it to say, the first two tests passed on this Matrix. Now, before telling the customer she needs a PCM, it is wise to find any other evidence of a module issue.
One way to do this is to look at the PIDs and see if there is something that does not make sense. Because there was a system lean DTC, the O2 sensor PIDs were looked at first.
Since a reading of about 3.3V is normal for an air-fuel ratio sensor and the vehicle was in that ballpark, it was wise to move on to the rear oxygen sensor. The readings were frozen at 15 mV (see Figure 8).
This was obviously abnormal. There wasn’t an exhaust leak, so it was decided to backprobe the rear oxygen sensor. It gave us good readings. Why was the PCM not picking it up?
At this point, the connector to the rear oxygen sensor was disconnected and wiring to the PCM was checked. The wiring checked out, but interestingly enough after plugging in the sensor the PCM started picking up rear O2 data (see Figure 9).
Since there was not any corrosion or any debris on the connector, it did not seem wise to just consider the vehicle “fixed.” After all, the oxygen sensor connector did not disconnect itself or have any corrosion on it. However, the battery sure had corrosion! (See Figure 10.)
The battery terminals were cleaned, a soft reset performed, and idle was relearned. On Toyotas, to relearn idle the vehicle has to be test driven a few miles.
There have not been any problems with the vehicle since. ●