jwahaus

Anyone know what the expected output of the MAF sensor should
be for a given RPM?

For example, what should be the normal MAF reading in lb/min for
2500 RPM?
3000 RPM?

I'm wondering if you can determine if your MAF needs to be cleaned by
just looking at its output (with an OBDII reader) at a given RPM.

MTL_4runner

To answer your question, it will vary by sensor so, no, there is no precise reading for a given RPM.
It should be somewhere between 0-5 volts and porportional to the speed of the air movement.
There is no way to verify the linearity of the sensor except in operation.

jwahaus

Okay, what about a ballpark figure? They must be accurate within a certain
tolerance otherwise there would be no point in having them.

The truck's OBDII interface converts the voltage into "lbs/min". There has
got to be some guidelines somewhere that indicate reasonable values for
a given RPM.

Would you expect a supercharged engine to still give a linear result? I
would expect the FI to result in a non-linear output with the slope increasing
at higher RPM's. But I don't know, I just guessing here.

MTL_4runner

Here is a bit of info from another site which might help:

"The hot wire MAF used in 3rd generation Toyotas, measures air flow volume, indirectly by measuring the current flow that must be passed through a small heated wire exposed to the incoming air flow. As more air comes in, the wire which is heated to around 200 degrees Centigrade, is cooled by the passing air (and resistance lowered). As a result, the computer sends even more current through the wire to keep it heated such that the wire's resistance is kept equal to a reference resistance value. By knowing the amount of current required to keep the hot wire equal to the reference resistance value, the ECU indirectly knows the volume of incoming air. There is also a thermistor in the MAF sensor so the ECU knows the temperature of the incoming air and takes that into account.

As fine dust particles and other contaminates are "cooked" by the hot wire and accumulate on its surface, the wire slowly becomes increasingly "thermally insulated" to the passing air, and begins to under measure the incoming engine air flow at certain engine speeds. That is, more air is being drawn into the engine than is being reported to the ECU. With the MAF sensor under reporting the volume of incoming air, the ECU "base injection duration" of injector open time will be on the low side, resulting in a lean mixture.

This process of the wire becoming coated with contaminates is known to the industry and many hot wire MAF sensors now incorporate a "burn off" cycle activated every time the ignition switch is cut off. Sufficient current is passed through the hot wire to briefly raise it to 1000 degrees Centigrade to burn off any contaminants. I have not seen any information in Toyota documentation indicating that this feature is incorporated on the 3rd generation engines under discussion, so this might be the reason this problem occurs. Newer technology MAF sensors use a heated "thin film" versus "hot wire" that is reported to be more resistant to contamination problems."

And you are correct, I should not have used the term linear with the MAF because it is not in fact a linear voltage output like a TPS would be. The graph should look more like this (this is from another vehicle but it is a hotwire system):



Now your OBD system does correctly give a flow rate for the air because after all, that is what it is actually measuring with the MAF. If the sensor is dirty, it will show less voltage at the same flow rate and thus the most common error code is P0171 "System too Lean (Bank 1)". This is essentially the ECU taking into account all the other sensor readings as well as programmed fuel maps and saying that the system is getting too much oxygen and not enough fuel. The common cause of this is a dirty MAF.

Now the only way to have any actual correlation of the flow rate as measured by the MAF and engine RPM is to take a reading of both at several points and plot them on excel using a best fit curve. Obviously if the MAF is in fact dirty, the values will be different than a clean one.