Scope and prerequisites
This guide concerns VAG EA113 2.0 FSI/TFSI MED9.1 research. Record the engine, ECU software, injector part number and revision, fuel, pump hardware and changes to the combustion chamber. Match definitions to the actual binary using the software matching guide. Establish the active lambda request and fuel-pressure behaviour before interpreting an injector correction.
Separate the injector functions
The table summarises Bosch RKTI 32.20.1, pages 1831–1837, in program 41W038_PQ35, 20 August 2004. It is a function reference for that implementation; verify selection logic, units and locations in the target software.
| Function | Role to identify |
|---|---|
| KRKATE and KLTIKRPR | Relative fuel mass to effective injection time, with pressure-dependent flow correction. |
| KLTVTSV | Pressure-dependent time offset added to the corrected effective time. |
| KLHDEV | Correction of injector nonlinearity at short effective times. |
| TIMINP | Minimum injection-time limit. |
| FKKVS | Fuel-system fine correction indexed by effective injection time and rpm. |
In this document the relevant pressure difference is across the injector, between rail and combustion chamber. Compression-stroke injection includes a chamber-pressure calculation; substituting rail pressure or manifold boost for the curve's specified input changes its meaning. The KRKATE derivation also depends on per-cylinder displacement, characterised flow and its test conditions; KRKATE and KLTIKRPR must share the appropriate nominal-pressure basis. The same section also describes starting branches where FKKVS is excluded, so a warm-running correction need not fix cranking.
Read the offset discussion through to its correction
The June 2015 KLTVTSV discussion began with confusing axis data and assumptions about opening delay. The author later reported a corrected axis definition. On 9 June, prj explained the signed offset in terms of opening and closing contributions; nyet distinguished duration offset from latency. The useful lesson is to identify what the calibrated time corrects before interpreting its sign. A negative entry alone does not prove a bad definition, and this explanation supplies no replacement curve for an unidentified injector.
The April 2018 fuelling discussion treats FKKVS as a multiplier and cautions that injection-angle control already has window handling. The follow-up distinguishes KLTVTSV's pressure dependence from the voltage dependence familiar from ME7 TVUB. Check the complete MED9.1 path instead of transferring a port-injector calibration recipe.
Injector swaps: preserve the later outcome
In the linked 079906036N case, ericpaulyoung described fitting newer RS4 injectors to a 2008 A4 2.0 FSI. An April 2015 positive report was followed by a May report of delayed cold starting, alongside concurrent throttle-body and airflow-model changes. In June the author reported poor starting after copying MED17 timing/offset data; another contributor reported no improvement from a suggested timing change. These observations support checking operating state and component identity; they do not establish a universal RS4-to-2.0 calibration.
Obtain flow-test conditions, spray-targeting and electrical-drive data for the fitted part. Bosch Motorsport's HDEV 5.2 data sheet, pages 1–2, illustrates these separate component specifications; it does not establish compatibility of the forum injector with an OEM MED9.1 installation.
Injection duration and the available window
Bosch AWEA 27.20.0, pages 1807–1808, distinguishes start/end angle by injection type and describes conditional advance of homogeneous injection when an end boundary would be exceeded, with an advance limit. Establish the active event, angle reference and limits before changing phasing. A simple time conversion is crank degrees = 0.006 × rpm × duration in milliseconds; this arithmetic does not identify the permitted injection window.
- For a broad running error, verify injector characterisation and pressure basis before using a local correction map to conceal the mismatch.
- For an error concentrated at small pulses, examine offset, short-pulse correction and minimum-time limiting together; use component data and independent checks to distinguish them.
- For cold-start-only trouble, record starting temperature, cranking voltage/rpm, rail pressure, injection type, duration, phasing and misfires. A warm fuel-trim result is insufficient evidence for the starting branch.
- For a high-rpm shortfall, compare final lambda request, actual mixture, rail request/actual and the verified event duration/angles. Determine whether delivery, representation or the available window limits the result before adding load.
These checks are editorial diagnostic reasoning from the cited evidence. No injector scalar, minimum pulse or timing value is prescribed. Forum attachments and the reported hardware were not independently tested. Sources reviewed 16 November 2026.