Mass Air Flow Meter Testing

Using the Snap Throttle Test 

For the purposes of this article, I will be dealing only with mass air flow (MAF) sensors that use voltage variations as a signal to the PCM and not frequency. The manufacturers tests for a MAF sensor often deal with checking the minimum and maximum output voltages and a few resistance checks using an ohmmeter. As with any electronic component, an ohm test will verify a part is bad, but it will NOT verify a part is good. In other words, failing an ohm test indicates a defective part, passing an ohm test means absolutely nothing.

Occasionally a MAF will fail completely or be so far out of calibration that it is easy to diagnose, however, that is usually not the case. As MAF sensors age the hotwire used for sensing becomes dirty or sometimes coated in oil or other residue. This can cause many symptoms including pinging, spark knock, lean codes such as P0171 and P0174, sluggish acceleration etc.

On Ford vehicles there is a baro pid available in the scandata that can assist in diagnosing the fault. Since many Fords do not actually contain a baro sensor, this value is calculated using the MAF sensor at heavy or wide-open throttle conditions. Use the frequency to altitude chart to determine if the baro reading is correct. If a sensor is sluggish or its reading is skewed, it will usually cause a lower than normal frequency giving some of the above lean symptoms. Also note that a dirty MAF sensor on many vehicles can give negative fuel trim values at idle and positive fuel trim values at higher RPMs. This is another indication that the sensor needs cleaning or replaced.

The capture below is of a Ford Thunderbird with P0171 and P0174 codes.

The actual baro reading of 141 Hz is normal if the vehicle were at 6000 feet above sea level. The problem here is that the vehicle is in Atlanta, Georgia at approximately 1000 feet above sea level. The proper reading should be around 156 Hz as seen in the capture below.

The Snap Throttle Test

Now to verify the fault with a scope, use the snap throttle test. This test works on any voltage based MAF sensor, not just Fords. Adjust your scope to .5 volt and 200 ms divisions. This should give you about 2 seconds from one side of the screen to the other. Now connect to the MAF signal wire and monitor the output at idle. Set the trigger to an up slope at approximately .5 volts ABOVE the idle value.

Allow the engine to warm up and wait for idle to come down to its lowest point. Turn all loads off. Now, snap the throttle HARD and hold it for a second or so. Make sure you snap the throttle from low idle or the results will be skewed. The engine should reach around but not over 4000 rpm. You should get a pattern similar to the one below from a 1999 Toyota 4-Runner. This customer also had a complaint of a check engine light with P0171 and P0174 codes. He also stated that the vehicle seemed to be down on power.

The initial voltage spike occurs as the throttle is opened and a sudden burst of air rushes in to fill the intake plenum. The airflow then slows as the vacuum in the intake is filled and begins to rise again evenly as the engines RPM increases.

The plenum fill voltage spike is the telltale sign of a sluggish MAF sensor. It should reach around 3.5 volts or more, with most achieving closer to 4 volts or more. Looking at the above capture you can see that the initial plenum fill only reaches about 2.8 volts while the maximum voltage achieved on the snap is around 3.5 volts. While cleaning the hotwire sometimes fixes a MAF sensor, sometimes it does not as was the case with this one. A new sensor was installed and a new throttle snap was taken.

With the new sensor the initial plenum fill spike reaches about 3.75 volts with the overall voltage reaching around 4 volts at the end of the snap. The highest voltage reached at the end of the waveform is directly proportional to the RPM achieved during the throttle snap. Many vehicles will have a rev limiter that will not allow the engine to rev over a specified RPM while in neutral. For these, you should still be able to get a good initial plenum fill spike, but the waveform will be cut short as the PCM cuts out the fuel injectors.

Note that there is also a tremendous difference in response time to the throttle snap. In the first capture the throttle was held for about 800 ms and the MAF only achieved 3.5 volts. In the second capture the throttle was held for 700ms or less and the MAF achieved over 4 volts. The engine was much more responsive with the MAF and picked up a lot of lost power.

Cleaning Verses Replacing

While cleaning a sensor will often make it perform better, it usually does not bring it back to peak performance. The example below shows the difference between a cleaned sensor and a new sensor. The vehicle is a 1997 Mazda 626 with a complaint of very low power. No check engine light on and no trouble codes stored.

As you can see from the image above, the plenum fill spike only reaches about 2.6 volts. The overall waveform can only achieve about 3 volts. A quick inspection of the sensor shows that the sensing element is covered in fuzzies.

After cleaning, throttle response was much improved. This improvement is also reflected in the snap throttle waveform shown below. The initial spike will now reach 3.69 volts, but is that enough for this vehicle? Should it be higher?