Tuning Wiki
Applies to

Check the hardware and software limits stated in this article.

Scope and map selection

MED9.1 ignition calibration must be matched to the exact engine, hardware and software. Bosch's function description for 41W038_PQ35, 20 August 2004, identifies KFZW and KFZW2 as ignition-angle maps indexed by engine speed and relative charge. Its ZWGRU function description, page 3093 describes linear interpolation to account for camshaft adjustment and charge-motion flap operation, with additional map variants depending on the application. This describes that program, not a complete map list for every MED9.1 version.

A question in the long-running NefMoto MED9.1 Tuning/FAQ/Help thread asked, for an Edition 30 car, which of several KFZW and KFZW2 maps was used. A reply described coding-dependent variants and blending with inlet cam position; a later follow-up does not establish a tested outcome. Bosch supports checking interpolation in its documented program, but the exact selection, coding, axes and available maps must be established for the target software. One table alone should not be assumed to supply the full-load value, and a setting should not be copied between variants without that evidence.

Bosch MDBAS, page 736, places the KFZWOP optimum ignition reference in the torque model. It is a model reference, not a direct reading of actual spark advance. The torque, boost and ignition overview provides the wider context.

A definition mismatch to investigate

In a 2017 BWA Golf GTI report, knock correction appeared after working from AXX/BPJ experience. A reply suggested the KFZW definition was displaced by one byte. The owner acknowledged the suggestion, but the cited follow-up provides no verified corrected file or post-fix log. This is a reason to check the definition, not proof of the cause or a transferable address.

The ECU identification, original read and software version should be compared against a verified definition for that exact file. Map start, data width, byte order, scaling, units, dimensions and both axes should be inspected in context; plausible-looking numbers alone do not establish correct identification. The software and definition matching guide explains the identification process.

Interpreting knock and ignition logs

  1. Record the software and coding, engine, intake and exhaust hardware, fuel, fault codes and service condition. Compare with a known-good baseline for the same configuration where available. Diagnose mechanical, sensor and wiring faults before attributing a change to calibration.
  2. Compare time-aligned engine speed, load, requested and actual boost, intake and coolant temperatures, requested and actual lambda, rail pressure, actual ignition and per-cylinder knock correction. Include requested and actual cam position and charge-motion flap state when available. Compare similar operating points, temperatures, gear and fuel; verify what each logged channel actually represents.
  3. Look for repeatable cylinder-specific behaviour and changes that track operating conditions. Knock correction is one piece of evidence, not a direct measurement of knock intensity; a short log without reported correction does not establish a safe tune. Keep protective functions active and investigate persistent correction or misfires before increasing load or advance.

Ignition retard can affect exhaust heat. A MED 9.1 basics follow-up reported changes in an exhaust-temperature value alongside changes in ignition and fuel quality; because conditions changed together and the measurement basis is not established there, it cannot quantify the effect. Whether a temperature channel is measured or modelled should be identified before conclusions are drawn. For intake-related load and correction changes, see intake comparison and fuel-trim diagnosis.

Not established

  • The cause of the reported knock correction on the BWA Golf GTI, and whether a one-byte definition offset was the fault.
  • A tested outcome for map selection on Edition 30 software.
  • A quantified link between ignition retard and exhaust temperature.

Sources are scoped to the cited examples and have not been independently vehicle-tested by this wiki. Sources reviewed 17 November 2026.

Continue the conversation

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