Tuning Wiki
Applies to

Check the hardware and software limits stated in this article.

Scope and the useful question

This guide concerns the port-injected VAG ME7.1 and ME7.5 examples reported in NefMoto forum discussion, including the Bosch EV14 injector migration thread. The reports describe different engines, injectors, fuel pressures and software; a shared injector family or flow rating is not a complete calibration. Identify the ECU software, exact injector and any modifications, fuel, pressure regulator, MAF and sensor arrangement first.

After a swap, a good idle alone does not prove that part-load, cranking, warm-up and richer operation are correct. Before editing, record faults and the existing adaptations; investigate leaks, fuel supply, exhaust leaks and sensor faults. Establish differential fuel pressure and obtain injector data for the actual injector, pressure and electrical conditions. See ME7 fueling and MAF calibration.

Why low pulse widths need a separate check

KRKTE is part of the primary fuel conversion; TVUB is a voltage-dependent injector-time correction. The migration discussion notes that a given time offset has a greater proportional influence on a short pulse. For illustration only, a 0.2 ms timing difference is 10% of a 2 ms electrical command and 2% of a 10 ms command. Those fractions do not directly predict fuel error: the injector's opening and closing behaviour and short-pulse non-linearity also matter. Injector Dynamics explains pressure, voltage and short-pulse effects using a GM characterisation example; that example is evidence about injector behaviour, not ME7 map definitions.

Record battery voltage at each operating point, not just RPM. If idle and cruise occur at different voltage, changes in TVUB, charging voltage and operating state can be entangled. Do not assume that every injector with the same nominal flow has the same latency or minimum repeatable pulse.

A correction's input matters. A forum M-box injector discussion lists voltage and temperature-related corrections separately. The Bosch Audi 132 kW ME7.1 RKTI description identifies the TVTSPEV input as filtered, modelled intake-valve temperature, not a directly measured injector temperature; the same description includes neutral initial curve values. Establish whether the function is configured and what the installed file actually contains before using it to explain a hot/cold difference.

Reading a real injector data sheet

The Finjector EV14-790 data sheet provides a useful example, not settings for every EV14. It reports static flow of 790 cc/min at 3.0 bar differential and 930 cc/min at 4.0 bar, using ASNU test medium at 25 °C. At 14 V its dead-time table gives 0.929 ms at 3.0 bar and 0.992 ms at 4.0 bar. Both flow and offset must be matched to the actual injector and conditions; do not copy either pair into an unrelated calibration.

Bosch Motorsport's EV14 information gives a square-root pressure conversion for static flow and explains the effect of test-medium density. For this sheet, 790 × √(4/3) is approximately 912 cc/min, compared with the supplier's tabulated 930 cc/min. That comparison illustrates why the estimate is a starting point and why the actual characterisation and test conditions matter. Match differential pressure, voltage, electrical drive, medium, units and definition of the offset to the ECU and injector; a single point cannot establish a complete TVUB curve or short-pulse behaviour.

Equal static flow does not mean equal spray

Bosch Motorsport's 2006 catalogue, injection-valve EV 14 table, lists two standard-body conical-spray examples at 387.3 g/min of n-heptane: B 280 436 038_06 with a 25° spray angle and B 280 436 038_05 with a 70° spray angle. Both are listed as 12 Ω. The matching nominal flow and resistance do not establish identical spray targeting, dynamic behaviour or suitability for a particular inlet port. These are catalogue examples, not verified ME7 fitments or current ordering recommendations.

Alongside the required flow and voltage data, verify physical length and seating, connectors, electrical-drive compatibility, fuel compatibility and the spray pattern's suitability for the intake and valve arrangement. Consider cold-start and transient behaviour separately from warm steady-state lambda; a correct static flow figure cannot by itself validate fuel preparation or the completed calibration.

Compare feedback without making it a diagnosis

For many VAG controllers, Ross-Tech's fuel-trim reference describes measuring block 032 field 1 as additive/idle adaptation and field 2 as multiplicative adaptation. Confirm the label, units, bank and applicable measuring block on the actual vehicle. These are learned values, not continuously updating short-term feedback; a recent reset can leave them uninformative until the conditions for learning have been met.

Observation in comparable warm, closed-loop conditionsWhat to investigate
Idle differs substantially from part loadShort-pulse and voltage data, intake or exhaust leaks, fuel pressure and the active idle strategy. A difference does not by itself prove that TVUB is wrong.
Idle and part load show a similar proportional correctionInjector flow and differential pressure, MAF conversion, fuel and air leaks; more than one cause can give a similar pattern.
Steady operation looks plausible, but the fault appears on start, warm-up or tip-inOperating mode, temperature, wall-film/transient corrections and available pulse-width range. A warmed-up trim cannot validate all of these states.
Only one bank differs on a banked engineCompare the two banks' sensors, exhaust/intake leaks, injectors and wiring before applying a global correction.

Retain short-term feedback, long-term adaptations, actual lambda or sensor state, RPM, load, voltage, coolant and intake temperature, and injector-time signals with their software-specific definitions. Keep both pre-reset and relearned data. For rich operation, use an appropriate calibrated wideband where the fitted feedback cannot measure the mixture.

A reported pressure-reference mistake. In a forum-reported 1.8T injector-scaling case, rich idle and part-load trims were initially seen. The system pressure was then found to be 4 bar rather than the 3 bar assumed, and the learned trim was reported improved. The report does not establish a pressure for every 1.8T or independently verify the complete tune. Identify and measure the installed system's differential pressure before interpreting a nominal injector-flow figure or retuning corrections.

Effective time, voltage offset and short-pulse data

The Bosch Alfa 2.0 ME7.3.1H function sheet, RKTI 1.60, labels TEMIN as a minimum effective injection time. In that documented calculation the corrected effective time and voltage correction are combined to obtain bank injector actuation time; the sheet also names B_temin and B_temin2 as minimum-limit-active conditions. Therefore a logged total command must not automatically be compared directly with an effective-time limit. This is a separately documented Alfa example: confirm the signal, units, operating state and any active limit in the exact VAG software before interpreting its log.

Also check what an injector supplier's short-pulse table actually represents. Haltech's short-pulse documentation describes a correction added in milliseconds against effective pulse width. Bosch's cited RKTI example instead describes FKKVS as a fuel-system correction factor addressed by engine speed and effective injection time, with injector nonlinearity and system pulsation among its possible inputs. The tables have different units, axes and roles; a millisecond adder cannot simply be pasted into FKKVS. Supplier measurements may inform an appropriate calibration, but any conversion requires the actual function, electrical and pressure test conditions, and validation across the operating range.

Minimum time is not a universal number

TEMIN and TEMINVA are minimum-time constraints in the relevant application. Forum discussion questions whether a published injector minimum includes electrical offset or describes a linear/repeatable region; the answers and test conditions matter. Compare the injector data definition with the ECU's effective-time and total-command definitions before transferring a number. A minimum limit can conceal an inadequate short-pulse range; reducing it does not make an injector repeatable there.

The same migration thread contains conflicting suggestions, including opposite flow-ratio descriptions and different FKKVS settings. Treat its examples as observations to investigate, not a universal recipe. Validate data and map semantics for the fitted hardware; never select a value solely because an unrelated car ran with it.

Related: ME7 MAF and lambda log review · ME7.1 S4 fueling examples. Forum reports have not been independently vehicle-tested by this wiki.

For idle-speed control and changing ignition, see ME7 idle ignition and torque reserve. If steady fueling is acceptable but a cold start or throttle transition remains problematic, compare the phases described in ME7 cold start and throttle transition fueling.

For a worked explanation of why good lambda can hide compensating MAF and injector errors, and for comparisons across operating points, see ME7 identifying injector and MAF errors from logs.

Use ME7 fuel pressure and injector differential pressure to compare actual rail and manifold pressures before assuming an injector-flow error.

For selected 1.8T 20V injector sets and their manufacturer/supplier test conditions, see the 1.8T injector and fuel-pump guide. Product flow ratings do not replace ECU-specific offset and short-pulse data.

Continue the conversation

If this topic interests you, consider joining the following forums. They host material referenced in this article.