The MAP sensor: manifold pressure, and what a failed one does to fuelling

MAP stands for Manifold Absolute Pressure, and the middle word is the one that matters. It is not a vacuum sensor. It does not measure how much lower than atmospheric the manifold is — it measures the actual pressure in there, referenced against a sealed vacuum inside the sensor itself.

That distinction sounds academic until you drive up a mountain, at which point it is the entire reason the sensor exists.

What it is measuring

With the engine off, the manifold is open to the outside through the throttle body, so manifold pressure equals atmospheric pressure — around 101 kPa at sea level, less as you climb.

Start the engine and the pistons pull against a throttle plate that is nearly shut. That restriction is what creates manifold vacuum: the pistons are trying to draw more air than the throttle will pass, so pressure in the manifold falls, typically to 20–40 kPa at warm idle. Open the throttle fully and the restriction disappears; manifold pressure climbs back towards atmospheric and vacuum goes to almost nothing.

On a turbocharged or supercharged engine the top of that range is somewhere else entirely, and it is the half this article's original never had to consider. A compressor upstream of the throttle pushes the manifold above atmospheric, so the same sensor reads boost as well as vacuum — commonly to 250 kPa and beyond. That is why the corpus describes the part as measuring vacuum on a naturally aspirated engine and boost on a forced-induction one, and it has two practical consequences: a boosted engine's MAP sensor is a different part with a wider range, so the two are not interchangeable however well the connector fits, and every test below that expects the reading to stop at barometric needs reading with that in mind. Turbocharger failure covers what an underboost reading usually turns out to be.

So manifold pressure is a direct read on how hard the engine is working. On a speed-density system the ECU takes that pressure, the engine speed and the intake air temperature, calculates the mass of air entering each cylinder, and meters fuel to match. Get the pressure wrong and every injector pulse after it is wrong.

Why absolute, and why that matters at altitude

Because the reference is a sealed vacuum rather than the outside air, the sensor reports true pressure rather than a difference. That means it reads barometric pressure whenever the throttle is open enough for the manifold to equalise — most usefully with the key on and the engine not running.

This is what lets the ECU compensate for altitude. Air at 2,000 m is around 80 per cent as dense as air at sea level, so the same throttle opening admits substantially less oxygen. A system that measured only vacuum could not tell the difference and would run rich as you climbed.

What it is not

The Persian original of this article described the sensor as piezoelectric, and then as "in fact a simple potentiometer whose wiper slides with air pressure". Both are wrong, and they are wrong in a way that matters if you are trying to test one.

It has no moving contact. There is nothing to wear, nothing to go open-circuit at one spot in its travel, and no dead band — so the failure signatures of a throttle position sensor do not apply here.

It is piezoresistive, not piezoelectric. A piezoelectric element generates charge when it is stressed dynamically and produces nothing at all under a steady load, which makes it useless for measuring a pressure that sits still. What is actually inside is a micro-machined silicon diaphragm with strain gauges diffused into it, wired as a Wheatstone bridge, with a sealed vacuum on the far side. Pressure deflects the diaphragm, the strain changes the gauges' resistance, the bridge turns that into a small voltage, and on-board electronics scale it to the 0–5 V the ECU expects.

The practical consequence: a MAP sensor's output should be smooth and continuous across its whole range. Steps, flat spots and dropouts are electrical faults, not mechanical wear.

What it feels like when it fails

The sensor rarely dies outright. It drifts, or its vacuum supply leaks, and the fuelling follows it.

Reading too high — the ECU believes the engine is under more load than it is, and adds fuel. Rich running, poor economy, a smell of unburnt fuel, sometimes black smoke on a sudden throttle opening. The car often still feels strong, which is why this one gets tolerated for months.

Reading too low — the ECU believes there is less load than there is, and takes fuel away. Lean running, hesitation, stumbling on acceleration, and stalling as the engine comes back to idle.

Poor cold starting is common to both, because a cold start leans heavily on the ECU's model of how much air is going in and there is no oxygen-sensor feedback yet to correct it. A car that needs a lot of cranking when cold, then dies if you touch the throttle, is worth checking here.

Testing one

The barometric check — do this first

The single most useful test needs only a scan tool. Key on, engine not running, read the MAP value. The manifold is at atmospheric pressure, so the sensor should report today's barometric pressure at your altitude.

Compare it with the real figure from a local weather source. They should agree closely. A sensor that reads noticeably low with the engine off has drifted, and it will be wrong everywhere else too.

This check is important enough that the standard gives it its own fault code: P0069, Manifold Absolute Pressure — Barometric Pressure Correlation, is set when the ECU compares the two and finds they disagree.

Watch it run

Start the engine and watch the value fall to idle vacuum, then snap the throttle open. Pressure should rise almost to barometric — well past it on a boosted engine once it is actually making boost — and drop back cleanly. Sluggish response, a value that sticks, or one that never reaches barometric on a snap throttle all point at the sensor or its plumbing.

Calculated load is a useful second opinion, because on a speed-density system it is derived largely from this sensor. It should be low at idle and high under load. A load value that barely moves, or that sits implausibly high at idle, says the same thing the raw pressure does.

Voltage at the connector

A MAP sensor is normally a three-wire device: a 5 V reference, a ground, and a signal.

With the connector attached and the engine running, back-probe reference to ground: it should be a steady 5 V and should not move when you blip the throttle. If it does move, the fault is upstream of the sensor.

Then signal to ground. It should sit low at idle and rise towards the top of its range on a sudden throttle opening — roughly 0.2 V to 4.8 V across the full span on a typical sensor. A signal pinned at either extreme, or one that does not move at all, has failed.

Do not trust pin numbers from an article — this one, or any other. Which pin is which varies by manufacturer and connector, and probing a 5 V reference into a ground because a diagram was for a different car is an expensive way to learn that. Get the pinout for the actual vehicle.

The vacuum test

Where the sensor is mounted remotely and fed by a hose, apply vacuum with a hand pump. It should hold. A sensor that bleeds down has a split diaphragm and is finished.

Check the hose itself while you are there. These are usually the oldest, hardest bits of rubber in the engine bay, and a pinhole in one produces exactly the symptoms of a failed sensor at none of the cost.

Cleaning, and why they get dirty

Crankcase vapours are routed back into the intake through the PCV system, so everything downstream of that point is permanently exposed to oil mist. Over years it builds a film on the sensor element, which slows its response before it changes its readings — so the first symptom is often a flat spot or a hesitation rather than a fault code.

Sensors mounted directly on the manifold can be removed and cleaned carefully with a proper electronics or MAF-safe cleaner. Let it dry fully. Do not scrub the element, and do not use carburettor cleaner on the exposed silicon — it is aggressive enough to attack the coating.

Two details worth knowing:

• Mount it the way the manufacturer specifies, which for hose-fed sensors usually means the spigot pointing downwards. Pointing up lets condensation collect in the sensor, and water inside one produces wandering readings that come and go with the weather.
• Check the O-ring on manifold-mounted sensors when you refit. A leak there is an unmetered air leak straight into the manifold, which will chase you around the fuel trims for a long time.

On many later cars — the Peugeot 206, the Xantia, later Prides — the intake air temperature sensor is built into the same body. The two functions are independent, so it is entirely possible to have a good pressure signal and a failed temperature signal from one part, and worth reading both before condemning it.

Related codes

The circuit codes are P0105 through P0109 — circuit fault, range and performance, low input, high input, and intermittent respectively. P0069 is the correlation code described above.

Worth knowing which one you have: a low or high input code is usually wiring, because those are set when the signal leaves the range a working sensor can physically produce. A range/performance code is usually the sensor or its hose, because that one is set when the value is electrically plausible but disagrees with what the rest of the engine is doing.

Video guides

Video, for the parts of this that are easier watched than read. The English-language ones come first.