This is the MAF sensor for an H25 (2.5L V6) Tracker engine.



The MAF sensor is located inside the throttle body.



The throttle body is located at the rear of the engine.



The MAF sensor is directly in front of the throttle position sensor (TPS) and throttle plate.


The accelerator pedal is often mislabeled as the "gas" pedal when in fact it has no direct connection to the fuel system. It could be more accurately described as the "air" pedal because that is what it does control. The computer's job is to match the correct amount of fuel to the amount of air entering the engine. In order to do that the computer must first measure how much air is entering the engine. That measurement is the job of the MAF sensor.



The heart of the MAF sensor is a type of hot-wire anemometer. Anemometers are typically used as part of a Wheatstone bridge configuration (shown above). The circuit is composed of two known fixed resistors R1 and R2 and a third variable resistor R3. The hot-wire probe is the fourth resistor Rw that completes the bridge. The bridge is balanced when the voltages at point 1 and point 2 are equal. It does not matter what the voltage is. It only matters that they are equal. Any difference in potential between points 1 and 2 is interpreted as an "error voltage." Depending on the polarity of the error voltage, the servo amplifier will increase or decrease the output of the current source driving the bridge. Unlike the other three resistors, the wire that resistor Rw is made from increases in resistance when it heats up. Conversely Rw's resistance decreases when it cools off.



With current flowing through the bridge circuit but no air flowing over the probe, R3 can be adjusted so the bridge is balanced.



When air starts to flow over Rw it cools down and its resistance decreases. Less voltage is developed over Rw and more is dropped across R1. This unbalances the bridge. The error voltage increases the current to the bridge. The extra current heats up Rw which increases its resistance. As a result, the voltage across Rw increases.



The feedback current continues to increase until the error voltage is driven back to zero. The amount of current it took to rebalance the bridge is proportional to the air flowing across Rw.



The MAF sensor converts the amount of current change into a proportional DC voltage which is fed back to the PCM as a mass-air-flow analog.



I wanted the see the MAF voltage change with a change in engine RPM. I had to first find a place to measure engine speed. From the schematic above, I assumed the PCM sent a DC voltage to the instrument cluster tachometer. It turned out that assumption was wrong.



The PCM sends a series of pulses (chanel 2) to the tachometer. I wasn't sure how the pulse frequency (35 Hz) was related to RPM. So I compared the tachometer pulses to the CMP sensor reference pulses at idle. From the "intro to the CMP sensor" I knew there were 360 REF pulses per camshaft revolution. 2,160 Hz divided by 360 yielded 6 camshaft revolutions per second. Since the crank turns twice as fast as the cam, that meant 12 engine revolutions per second. Multiplying by 60 gave an engine speed of 720 RPM. My Tracker generally idles at 750 RPM so 720 seemed acceptable.



When I compared the PCM tachometer pulses to the CMP sensor position pulses I got a one-to-one correlation in frequency. I knew there were six POS pulses per camshaft rotation so dividing 36 Hz by 6 gave 6 cam revolutions per second or 12 crank revolutions per second or again an engine RPM of 720.



Confident that a 36 Hz PCM tach signal represented 720 RPM, it was a simple step (720/36) to realize each Hz on the scope represented 20 RPM. As a test, I pressed the accelerator till I read 2,000 RPM on the tach. I recorded a PCM tach frequency of 99 Hz. 99 times 20 is 1,980 RPM. If I had used the CMP POS frequency, it would have come out to an even 2,000 RPM. This proved I could reliably use the PCM tach signal to measure engine RPM.



The factory service manual states the MAP signal should be between 1.5 and 1.8 volts DC at idle.



I connected channel 1 of the scope to the MAF signal coming into the computer. At idle (720 RPM) the MAF signal was 1.61 volts DC (well within tolerance).



A live data scan showed the mass air flow at idle was 3.78 grams/second.



With the engine running at 2,000 RPM the MAF voltage was 2.00 volts.



Live scanner data showed a MAF about 8.0 g/s near 2,000 RPM.

 

With the engine running at 3,060 RPM (153 Hz times 20) the MAF voltage was 2.40 volts.



Live scanner data showed a MAF of 11.89 g/s at 3,000 RPM.



An internet search revealed the output of the MAF sensor output should be linear over its operating range (red line). I graphed the live data MAF readings of my sensor (blue line) and it looked linear to me. If you suspect a bad MAF sensor in a V6 Tracker, you could plot your scanner data on this graph. If they match, your MAF sensor is probably working correctly.



I wondered if the raw MAF sensor signal voltage would also be linear and it was. So if you don't have a scanner with live data, you could back-probe the PCM or the MAF sensor connector with a digital voltmeter and plot the voltage here. The amount the MAF sensor wire is cooled is directly proportional to the temperature, density and humidity of the air passing through the sensor and as such, the current increase needed to heat the wire allows the computer to calculate the volume of air entering the engine.