Measuring wind

Jan. 1, 2020
In order for the engine to run, accurate calculations have to be made from reliable inputs, and we’re here to talk about the most important input on most of today’s cars.

We can’t see the ocean of air through which we move and breathe, but we can feel it and we can see what it’s doing. At 78 percent nitrogen and 21 percent oxygen, the atmosphere blankets our planet, about 350 miles deep if you count all four layers of it.

Like this article? Sign up for our enews blasts here.

That’s fairly thin when you consider the diameter of the planet is approximately 8,000 miles. At sea level, the air presses on everything from every direction at about 14.5 psi. Without atmospheric pressure, water couldn’t exist as a liquid and our bodies would explode. That pressure is why astronauts need heavy-duty space suits when they’re outside our air blanket, and that same pressure is why we can breathe, drink through a straw and build heavier-than-air vehicles that fly. And as our planet spins, this blanket of air we call our atmosphere spins with it.

Our vehicle engines breathe the same air we use in our lungs, and with modern emission controls, those engines breathe out carbon dioxide just like we do. That’s a good thing, actually. That means we’re going green; the plants love CO2. Why? Well, it’s because they breathe in that carbon dioxide, which is more than three times heavier than oxygen and nitrogen so that it hugs the ground, and that’s where the grass and trees live. And they breathe out oxygen, which is good for us and for the animals.

We gearheads all know that pistons and valves on one end of the manifold create low pressure in an intake manifold. And as we increase the flow of air by opening the throttle, the PCM modifies the flow of fuel proportionally so that, if things are working right, the engine picks up speed and power. With that power applied to the wheels through the flywheel, transmission, and final drive, we finally drive away. 

The now-almost-extinct carburetor of yesteryear takes advantage of the flow of that rapidly moving wind to draw fuel from a reservoir and mist it into the intake stream, where it finds its way to a place of combustion that slaps the pistons at the right time to produce much-needed power. When the air/fuel mix, the flame propagation, the spark timing are working together as they should, every flaming molecule of hydrocarbon fuel gets happily married to two molecules of oxygen, thus a molecule of CO2 is born. 

Of course, because of chemistry and heat, there’s some NOX, water vapor, oxygen and a couple of other gasses that leave the combustion chamber, too, and there are ways of dissolving the bad elements of combustion that prevent them from ever seeing the light of day on a properly operating system. The point is that, in order for this carefully executed dance to work as intended, accurate calculations have to be made from reliable inputs, and we’re here to talk about the most important input on most of today’s car engines.

Density vs. Mass The first fuel injection system I knew anything about used manifold
pressure (as measured by a Manifold Absolute Pressure (MAP) sensor) as the major modification of fuel injector pulse. And ever since the 1968 VW Type 3 squareback that used two sets of points under the breaker plate in the distributor to sense engine speed, electronic injectors have matched the frequency of their pulse to the speed of the crankshaft. The Bosch L-Jetronic fuel injection system appeared a few years later with its air-driven L shaped plate connected to a potentiometer.

Then there was the Bosch mechanical fuel injection known K-Jetronic on VWs, Audis and Deloreans and the movement of incoming air is important even on that mechanical fuel delivery system. These provide a steady spray through injectors that are more or less glorified atomizers.

The heart of the fuel distributor on those engines is a piston-type valve that moves inside a slotted sleeve that allows fuel to flow through tiny .02 mm vertical slits (one per fuel injector feed line). A lever that is hinged on one end and has a large round air-driven plate on the other end moving in a funnel shaped air passage controls the position of the piston. The more air that flows into the engine, the more the plate moves and the more of each slit is exposed, thus fuel volume is increased to those gloriously simple fuel atomizers.

As for more contemporary electronic fuel injection, before Mass Airflow Sensors (MAF), the old MAP sensor started out
remotely mounted with a rubber hose leading to the intake manifold, and sometimes it’s still connected that way. Toyota more or less favored the MAP sensor over MAF until the late 1990s.

The closer to the manifold that sensor is mounted, the better it works, and because Chrysler has never been comfy with MAF, they finally mounted the MAP sensor right on the manifold with its sensing port reading right out of the plenum. Interestingly, Chrysler never followed Ford and GM to the fixed orifice refrigerant systems either – they stayed with expansion valves in their A/C systems, but that’s a digression.

Higher manifold pressure (low vacuum) means the engine is under a heavier load, and for the PCM on vehicles that use it, that’s a dandy way to decide how much fuel the engine needs. The MAP sensor also provides barometric pressure information, and a MAP that doesn’t tell the truth about barometric pressure will cause fuel trims to go wacky. Besides the MAP, there are other modifiers, too, though. Engine Coolant Temperature, Intake Air Temperature, throttle plate movement and position, etc. But MAF or MAP most normally are the priority inputs for fuel delivery.

MAF – Order of the Day
MAF sensors are virtually always mounted between the air filter and the throttle plate and they don’t have any moving parts. They just measure the temperature, humidity and flow of the incoming air, and the ones we’re most familiar with are the hot wire MAF sensors. This type of MAF sensor has two wires; one that senses ambient temperature and the other one, called the hot wire, is kept at a fixed temperature above the measured ambient temperature. The Ford hot wire sensors produce a voltage from about 0.4 to 4.5 volts, varying with the airflow.

The GM sensor’s wires looks the same as the Ford sensor, but the output is a variable frequency square wave. A low frequency output signal from the MAF sensor indicates low air flow and high frequency indicates high air flow. Some Bosch hot-wire units have a self-cleaning cycle where the platinum wire is heated to 1,000 degrees Celsius for about a second after the engine is shut down to burn off contaminants. 
Most sensors don’t work that way and they sometimes have to be cleaned. Symptoms of sluggish acceleration or low power always lead me to have a look at those little hot and cold wires.

The earliest hot wire MAF sensors appeared on some imports as far back as 1979. That’s the most common type of MAF sensor used today. There are two tiny bridges, each with some platinum wire wrapped around something that looks like a piece of plastic.  The amount of current the sensor has to feed to the hot wire to keep it at the calibrated temperature is then converted to a voltage signal.

Rochester introduced hot-film MAFs on several 1984 model GM V6 engines. Like the later hot wire sensors used by GM, these hot film sensors produce a square-wave variable frequency output with a frequency range that varies from about 32 Hz at idle to 150 Hz at wide-open throttle.

In 1990, GM went back to MAP on most but not all of its engines. And if you buy a white box part to replace a GM hot film sensor, your MIL light might not go away even though the hertz reading is within specs, because the ECM sometimes just doesn’t like anything but a Delco sensor. Been there, done that.

Mitsubishi had oddball MAF sensors on their late 1980s and early 1990s cars that basically used a tiny speaker and a matching receiver with a honeycomb-like grid to direct the air so it passed straight on its journey through the chamber between them.  Since air deflects sound waves, this $700 sensor uses that principle to report airflow.

Ferreting Out a Problem
GM’s earliest MAF system came out in the mid-eighties, with sensors that turned out to be so unreliable that a retrofit was released to convert those early MAF equipped cars back to MAP sensor systems. The troubleshooting technique on the early GM MAF sensors was to gently rap the sensor with something and see if the engine stumbled. If it did, you replaced the MAF.

In the mid-90s when Ford’s platforms all went MAF, I personally became accustomed to a MAF voltage reading of about 0.82 volts at idle (it’d go up to about 4.6 volts at WOT), and more than a few times I’d see fuel trim numbers out of kilter only to find that the sensor was reading a smidge low. In these cases, the FT numbers would be at plus-27 percent with no vacuum leaks and normal BARO readings, but the MAF would be reading 0.78 instead of 0.82, meaning the MAF was slightly under-reporting airflow at idle – the HO2 sensors would detect a lean condition as a result, and the PCM would muster up some more fuel pulse to balance things out. A new sensor immediately would drive the Short FT readings back to the negative until the Long FT readings came back to single digits.

Every sensor I’ve seen on more contemporary vehicles reads grams per second (gm/s) and a V6 engine will typically read 3 to 4 grams per second (gm/s) at idle depending on temperature and engine size. The grams per second reading on your scan tool should pretty much mirror the Throttle Position Sensor voltage and react quickly with throttle angle changes, but it should remain stable at a given rpm and load.

Sometimes the MAF sensor can prevent the vehicle from starting – that’s a simple disconnect-and-try-it test. Then there are those times when the MAF simply can’t read past a certain flow – I’ve seen this a few times – the engine will run normally until the MAF tops out at less than an accurate reading and from that point on the PCM refuses to add the needed fuel.

It feels almost like a clogged fuel filter and you can see the MAF top out on the scan tool. Disconnect the MAF and drive it (the PCM makes it calculations a different way when the MAF is absent) and if it runs like a race car, you need a MAF.

Once a guy came to the Ford dealer complaining about an intermittent knock on his Explorer – sometimes the engine would run just fine and other times it would ping like crazy. I removed the air cleaner housing and found a roach’s wing lying on the air filter. Removing the roach’s wing fixed that one – sometimes the wing would get sucked up to the MAF inlet screen and block the air from passing through the small hot wire chamber. Sometimes it wouldn’t.

When the airflow was blocked by the wing, it would cause the engine to run very lean. Look for spider webs in that little hole, too, because that happens, and any kind of crack in the air inlet tube or an intake vacuum leak will allow unmetered air to enter the intake stream. That fouls up the PCM’s fuel delivery. Since MAF is the dominant input, that can cause nasty issues. If the intake air tube is cracked and the engine opens the crack wider with torque movement, it can cause a drive-away surge that will jerk a crick in your neck, so consider that as well.

Sometimes the PCM will internally short the MAF signal and act on the bad reading – this can cause rough idle and even low vacuum. In that case, I’ve disconnected the harness connector and wired the MAF directly to power and ground while reading the MAF signal with a meter. If I get normal readings that way and the signal wire isn’t shorted to ground between the MAF and the PCM, I know the PCM has gone stupid. I’ve seen that a couple of times, too.

The best way to handle any kind of scan tool reading is to read the stream on good cars so you’ll be able to recognize the bad ones. That’s the way I teach it and that’s the way I do it. Get used to the numbers. It’s surprising what a difference that makes in the long run.

Have articles like this sent to you weekly by signing up for our enews blasts here.