Tuning Wiki

This guide describes a practical approach to fuel calibration on programmable port-injection ECUs such as Speeduino and MegaSquirt. Menus, algorithms and available features differ by firmware and hardware. Start with the documentation for the exact ECU and use a calibration intended for the engine and firmware as a starting point, not a finished tune.

Before adjusting fuel

  1. Identify the ECU, board revision and firmware. Save the current tune and record the firmware, injector part numbers, fuel, fuel pressure and hardware configuration.
  2. Check for fuel leaks, correct fuel pressure and sound mechanical condition. Verify the injector electrical compatibility with the drivers and use the correct flow and dead-time data for the actual injectors, driver and pressure.
  3. Calibrate coolant, intake-air, throttle and MAP sensors as applicable. Check readings against plausible conditions. Set up a correctly calibrated wideband oxygen controller; an oxygen sensor alone does not supply a usable wideband signal.
  4. Confirm stable crank synchronisation and check commanded ignition timing against a timing light before attempting to correct running faults with fuel. Confirm that the selected load algorithm and units are understood.

Understand the tables

On a speed-density setup, the VE table generally uses engine speed and manifold absolute pressure as axes; the ECU combines the table with other configuration and corrections to calculate fuel. An AFR or lambda target table is a target, not a guarantee of measured mixture. Other algorithms, such as throttle-based Alpha-N or MAF, use different load references. Fuel and ignition are separate calibrations.

VE and direct injection-time tables

A table labelled VE does not necessarily use the same definition or reference conditions on every ECU. In Haltech's documented VE method, the ECU combines its efficiency value with inputs including displacement, injector data and target lambda. In its injection-time method, the base table instead specifies injection time. These are different strategies, not interchangeable tables; consult the specific firmware's definitions. Haltech: base fuel tuning.

Even when a measured mixture matches its target, the VE value can be compensating for an incorrect injector flow, dead time, pressure or sensor calibration. A seemingly smooth table therefore does not prove the air model is accurate. Haltech notes that errors in engine volume and injector data compromise the table, and that mixture should be checked after changes even with a well-characterised model. Haltech: tuning base tables.

On systems where the target table participates in the base fuel calculation, changing the target can change commanded fuel even without closed-loop feedback; the exact behaviour depends on the selected ECU strategy. A target remains a request, so validate the measured result and fuel system capacity. Haltech Nexus target lambda.

Breakpoints and interpolation

The values along a table's axes are breakpoints, not separate engine operating modes. Many controllers interpolate between neighbouring cells when the current speed or load lies between them. For a one-axis illustration, halfway between two cells containing 40 and 60, linear interpolation gives 50 before other corrections. A two-axis lookup can involve neighbouring values in both directions. The exact calculation and behaviour outside the axis range depend on the ECU. Speeduino: interpolated VE table; Haltech: interpolation during idle tuning.

Check axis units and placement before editing. Changing a breakpoint can affect a region of operation, not just its labelled point; table compare tools may themselves interpolate when axes differ. Review neighbouring values and log the actual conditions after a change. Haltech: comparing maps with different axes.

Warm running and VE

  1. Use a valid firmware-matched base tune appropriate to the engine and review every critical setting. Record a baseline log. A base tune is not proof that full-load operation is safe.
  2. Once the engine starts, bring it to normal temperature while monitoring oil pressure, coolant temperature, fuel pressure, lambda and engine behaviour. Resolve leaks, misfires and sensor faults first.
  3. At stable, repeatable operating points, compare measured lambda with the intended target and adjust the relevant fuel cells gradually. Record the change and repeat the measurement. Interpolation, other corrections and sensor delays mean that the highlighted cell is not always the only influence.
  4. Work through the accessible load and speed range progressively under controlled conditions. A load-bearing dynamometer and an experienced operator are appropriate for sustained high-load or boosted tuning. Do not use a universal AFR or ignition number as a substitute for the engine-specific calibration.

Interpreting a correction

A measured-to-target lambda difference can suggest a fuel adjustment only when the reading and operating point are representative. Check sensor readiness and exhaust integrity, fuel pressure, injector data, active feedback and transient corrections first. A misfire or a delayed exhaust reading can make a single sample misleading; an observed error at one speed and load does not establish the error everywhere else.

Interpolation means nearby cells can influence the delivered value. An unvisited region still needs plausible values and later validation; extrapolating a trend is an estimate, especially across a cam changeover or into boost. Review the affected area and remeasure progressively. Haltech notes that its Quick-Tune instant method adjusts the selected cell without considering neighbouring values, and that active oxygen trims affect its correction. Those details apply to that implementation, not every tuning tool.

Example: feedback during VE calibration

Check how the particular tuning function treats closed-loop fuel correction before accepting its suggested table change. If feedback is already adding fuel to meet the target, the measured lambda can look correct while the underlying table still needs attention; blindly adjusting the base table from the displayed lambda alone may miss that correction. Conversely, simply turning feedback off is not a universal requirement: follow the procedure for the exact controller and software.

For example, MoTeC's M1 package tuning notes describe an Engine Efficiency Quick Lambda function that accounts for closed-loop adjustment using its normalised exhaust-lambda channel. Haltech's NSP Quick-Tune documentation describes a different tool and cautions about current oxygen trims and neighbouring cells. These are product-specific examples; identify the software version, active correction, target and measurement channels, then review the result and recheck the engine under the same conditions.

On the Haltech base-fuel method, VE, injection time, fuel flow and MAF correction give the base table different meanings. A correction or scaling approach suitable for one method is therefore not automatically applicable to another. Record the chosen method and changes with the calibration and logs.

Cold start and transients

Once the warm VE calibration is sound, review priming, cranking pulse, after-start and warm-up enrichment during repeatable cold starts. Tune acceleration enrichment after the steady-state fuel table is established; distinguish a brief lean transient from a steady-state error using a time-aligned log. Make one change at a time and compare runs.

Distinguish the stages

Priming, cranking, after-start and warm-up are separate functions on many controllers, but their names, units and transitions differ. A prime pulse provides fuel for starting; cranking settings apply while the ECU regards the engine as cranking. Once the engine starts, an after-start correction can taper away, while a temperature-dependent warm-up correction can continue. For example, Haltech describes its prime-pulse strategy and Speeduino documents its after-start enrichment. Do not transfer percentages or pulse widths between ECUs without checking their definitions.

Before altering these corrections, verify the coolant reading against the actual cold engine, fuel pressure, injector data, battery voltage and the ignition and trigger setup. Capture the first start and subsequent warm-up in a log, noting coolant temperature, engine speed, corrections, pulse width and lambda when the sensor is ready. A cold exhaust sensor or an exhaust leak can make an early lambda reading misleading. Compare starts from similar conditions, and keep the previous tune so that a change can be evaluated.

Cold surfaces can retain some of the injected fuel as a film; Haltech describes post-start correction as compensating for fuel lost to wall and valve wetting. This does not imply a universal cold-start AFR or enrichment number. Haltech: post-start correction.

Why a throttle transient differs from a steady point

During a rapid throttle change, sensor and manifold response, injection timing and fuel deposited on intake surfaces can make the fuel reaching the cylinder lag the changing air demand. A calibrated steady-state table is therefore necessary but may not fully describe the transition. On throttle opening a transient correction can add fuel; on closing, some controllers can reduce fuel. The amount, triggering and duration are ECU-specific. Haltech’s transient-throttle documentation gives one implementation.

As another controller-specific example, Haltech describes a fuel-film model that estimates pooling and evaporation as demand changes. This does not imply that Speeduino, MegaSquirt or every Haltech ECU uses the same model or settings. Check the actual firmware and avoid transferring values. Log throttle, pressure, pulse width, correction and lambda with suitable time alignment; investigate sensor delay, injector timing and mechanical faults before using extra enrichment to mask a problem.

Using automatic analysis

TunerStudio VE Analyze Live and MegaLogViewer VE analysis can suggest or apply fuel-table adjustments where supported and licensed. They depend on valid sensor calibration, appropriate target tables and filtering. They do not verify ignition timing, engine safety or the suitability of an AFR target; review changes and keep a known-good tune.

Related reading

For the relationship between engine output and fuel use, see understanding torque, power and fuel consumption.