Advanced Live Data and Fuel Trim Analysis

Anyone can read a fuel trim of +18 percent off a scan tool. The harder questions are when it happens, on which bank, and whether the sensors behind that number are telling the truth. A lean code might mean a split hose, a weak fuel pump, a dirty airflow sensor or an exhaust leak, and those lead to very different repairs.

You already know what short-term and long-term fuel trim mean and how to compare idle with 2,500 rpm. This is the next level: reading data as a system, spotting a sensor that is lying, and using simple math to check what the computer reports.

What You Will Be Able to Do

• Set up a scan tool so the data is fast enough to be useful.
• Run plausibility checks that catch a bad sensor before you trust it.
• Read fuel trim across idle, cruise and heavy load to narrow the cause.
• Use bank comparison to split shared faults from one-sided faults.
• Calculate theoretical airflow and volumetric efficiency.

Make the Data Worth Reading

A scan tool asks each module for data, one value at a time. The more values you request, the slower each one updates. With 40 values on the screen, a short dropout can fall between samples and never show up. Pick the 4 to 6 values that matter for this complaint, and graph them on one screen.

A value on the screen is what the computer believes. It may be a measurement, a calculation or a substitute. If a sensor circuit fails, many computers replace the reading with a fixed default. A coolant temperature that never moves can be a default value and not a real one.

Plausibility Checks

Compare sensors that should agree, before you rely on any of them.

MAP means manifold absolute pressure. With the engine off, the manifold is at outside air pressure, so MAP and barometric pressure must match.

A coolant sensor that reads 40 degrees colder than the true temperature makes the computer add fuel. Long-term trim goes negative, and you could waste an hour hunting a leaking injector.

Reading Fuel Trim by Operating Range

The computer stores learned fuel trim in cells, which are small blocks of the engine's speed and load range. So there is no single long-term trim. The idle cell can be +18 percent while the cruise cell is +2 percent. That difference is a clue, so collect trim in three ranges.

1. Warm idle. Read trims in neutral or park, with accessories off.
2. Steady cruise. On a road test, hold a steady speed. A no-load 2,500 rpm reading is not the same as real load.
3. Heavy load. Climb a grade or accelerate firmly, but note that at wide-open throttle most engines go to open loop. Check the fuel system status value. In open loop, trim is not being applied, so that reading means nothing.

Add short-term and long-term trim for the total correction at each point.

Why does a vacuum leak fade? It lets in about the same extra air at any rpm. At idle that air is a big share of the total, and at cruise it is a small share. A weak pump is the opposite. It keeps up at idle, but it cannot meet demand at high flow.

Low fuel pressure and a low airflow reading look alike, so test them. Measure fuel pressure under real load with a gauge or a pressure sensor value. If pressure holds, look at airflow. On direct injection engines, compare desired and actual pressure on both the low-pressure and high-pressure sides.

Warning: Fuel systems hold pressure with the engine off, and direct injection runs at very high pressure. Follow the service procedure to relieve pressure, and use only approved fittings. Do not open a high-pressure line to test it.

Comparing Banks

On a V engine, each bank has its own oxygen sensor and its own trim. Comparing them is one of the fastest tests you have.

An engine with a single airflow sensor and one fuel rail feed has only a few parts that can affect both banks equally. If only one bank is off, you can rule those parts out.

Worked Example: Lean Codes on Both Banks

A V6 has P0171 and P0174, which are lean codes on bank 1 and bank 2. The driver says it feels flat on hills.

1. Both banks match, so the fault is shared.
2. The trim grows with load, so this is not a vacuum leak.
3. That leaves fuel delivery or the airflow reading.
4. A fuel pressure reading on the hill holds at spec, so the pump and regulator are good.
5. The airflow sensor is now the main suspect. Check for a dirty sensor, a leak after the sensor, and a dirty air filter, then compare airflow with the theoretical figure, as shown below.

Is the Oxygen Sensor Telling the Truth?

Fuel trim is only as good as the sensor that drives it. There are two kinds.

• Narrowband sensors switch between about 0.1 volt (lean) and 0.9 volt (rich). They tell the computer which side of the ideal mixture it is on, not by how much.
• Wideband sensors, also called air-fuel sensors, report how rich or lean the mixture is. Scan tools may show them as lambda, voltage or milliamps. At the ideal mixture, lambda is 1.0, and on a current display the reading sits at or near zero.

A quick check needs no special tools. On a warm engine, snap the throttle open and watch the sensor go rich. On a road test, lift off the throttle at speed and watch it go lean during fuel cut. If it barely responds, it may be slow or it may be contaminated.

The Exhaust Leak Trap

An exhaust leak ahead of the upstream sensor can draw in outside air between exhaust pulses, especially at idle. The sensor sees extra oxygen and reports lean. The computer adds fuel, so the trims go positive, even though the engine is really running rich. Adding fuel in response to a false lean signal can foul plugs and damage the catalyst.

If trims read lean but the spark plugs look sooty, or the exhaust smells rich, stop and check for a leak. Smoke test the exhaust with the engine off and listen for ticking on a cold start.

Using Airflow Math

You can check whether an airflow reading makes sense by comparing it with what the engine should breathe. Volumetric efficiency (VE) is how much air an engine takes in compared with how much it could hold. As a rule of thumb, a healthy non-turbocharged street engine reaches about 75 to 90 percent at wide-open throttle.

Theoretical airflow in liters a second is displacement in liters x rpm / 120. The 120 comes from two crank turns for each intake stroke, and 60 seconds in a minute. Air weighs about 1.2 grams a liter at normal conditions.

Worked Example: Checking the Airflow Sensor

A 3.0 liter engine pulls at wide-open throttle through 3,000 rpm on a safe road test. The scan tool shows 53 grams a second.

1. Theoretical airflow: 3.0 x 3,000 / 120 = 75 liters a second.
2. In grams: 75 x 1.2 = 90 grams a second.
3. VE: 53 / 90 = 0.59, or 59 percent.

That is well below the 75 to 90 percent you expect. This engine is either not breathing, or the sensor is reading low. Now split those two. A plugged exhaust or restricted intake makes the engine feel weak, and it may show low vacuum at raised rpm. A sensor that reads low with a healthy engine gives positive trims that grow with load, as in the worked example above. Clean or test the sensor, and repeat the math.

The figure is only a guide. Thin air at altitude lowers the result, and hot air does too. Turbocharged engines can go over 100 percent. Compare with a known-good vehicle where you can.

Warning: Do not read a screen while you drive, and do not run wide-open throttle tests in traffic. Use a helper to run the tool, or record the data and review it after you stop. Stay within the speed limit.

Common Mistakes

Check Your Understanding

1. A V8 with one airflow sensor reads +12 percent on both banks at idle and +13 percent on both at cruise. What kind of cause can you rule out? Causes that affect one bank only, such as a single intake gasket leak or one oxygen sensor. Both banks match, so look at shared parts.
2. A warm engine shows 120°F coolant on the scan tool, but an infrared thermometer reads 195°F at the thermostat housing. Long-term trim is negative. What is going on? The coolant reading is skewed cold, so the computer is adding fuel it does not need. Test the sensor and its circuit before you hunt for a leaking injector.
3. You lift off the throttle at speed, and the upstream sensor barely moves toward lean. What does that suggest? During fuel cut it should go lean quickly. A sensor that barely responds may be slow or contaminated.
4. You run the airflow math on a truck at 7,000 feet and get 70 percent. Is that a fault? Not by itself, because thin air lowers the result. Compare with a known-good vehicle at the same altitude before you decide.
5. A turbocharged engine works out to 115 percent volumetric efficiency at full boost. Is the airflow sensor reading high? Not necessarily. Turbocharged engines can go over 100 percent, so compare with a known-good vehicle.

Key Takeaways

• Fewer values means a faster refresh, so graph only what you need.
• Check that sensors agree before you trust them.
• Read trim at idle, cruise and load. The pattern across them narrows the cause.
• Trim that falls with airflow suggests a leak. Trim that rises with load suggests fuel delivery or airflow.
• Both banks point to shared parts. One bank points to that bank.
• Volumetric efficiency math shows whether an airflow reading makes sense.