Tuning Wiki

This is a historical revision. Read the current page or return to its history.

Page organisation in this revision
Page type
Technical article
Subject
Tuning
Parent
Bosch MG1
Applies to
Bosch MG1
Tool groups
None
Manufacturers
BMW

Changes from revision 2 to revision 4

Removed text is struck through; added text is highlighted.

Applicability and evidence

This guide concerns BMW B58 petrol engines discussed as MG1 applications. Hardware evidence comes from BMW's B58TU training; calibration evidence comes from bootmod3 and EcuTek documentation and public tuning discussions. These sources answer different questions. No binary, definition or vehicle has been validated by this wiki for this article.

Source/applicationWhat it establishesBoundary BMW B58TU training, July 2018B58B30M1 and B58B30O1 hardware context, revised direct injection and DME 8.6T0 designation.Factory training, not a software-specific map definition or current repair instruction. BMW source. bootmod3 BMW MG1 guideThe vendor groups B58 Gen 1 under PPC and Gen 2 under Aurix; its worked example is a Z4 M40i B58D.Check the actual ECU, ROM and tool version; those group names do not identify every vehicle. Vendor source. NefMoto B58 resource discussion, September 2023Discussion of definitions, strategy knowledge and logging.Contributor observations, not a verified Bosch function description. Discussion.

The fuel-system generation matters

BMW describes a revision from 200 to 350 bar, with a revised high-pressure pump and injectors. Its comparison distinguishes HDEV5, HDEV5 EVO and the Bosch HDEV6 used on Bx8TU. The DME generation has a common external appearance while internal hardware varies. See printed pages 28 and 31–33 of the training document.

The 350 bar figure describes that factory system's design. It is not a request to enter into every B58 calibration. Identify the fitted pump, injectors, pressure sensor and matching calibration before considering pressure changes. An upgraded pump alone does not establish compatibility of the complete earlier system with a later pressure specification. This is the practical inference from BMW's documented hardware changes.

Read the BMW strategy before following another MG1 platform

In the bootmod3 example, requested torque is related to load through reference-torque calibration. The guide documents separate filling constraints associated with operating conditions, plus Aurix logging of filling-limit reasons and a torque-limit status channel. ROM-dependent differences matter. This supports diagnosing the active path before increasing more limits; it does not establish common addresses or status-bit meanings across BMW software.

Some vendor example settings involve very high limits or monitor changes. They are not verified component ratings or a general prescription for this wiki. Retain a documented reason for each change and investigate an intervention before deciding whether its calibration is wrong.

What the discussions add

The 2023 NefMoto thread contains disagreement about learning from other platforms. One contributor specifically argues that BMW petrol application software differs from other manufacturers' MG1 implementations and has continuity with earlier BMW petrol software. Treat that as a contributor's architectural claim, not proof that any two binaries match. The useful research rule is to establish the function in the BMW file rather than transfer a Porsche or Audi map name.

The public BMW-XDFs repository separates B58gen1 and B58gen2 directories. That organisation illustrates the need for matching definitions. Only the repository overview was reviewed here; individual definitions were not audited for addresses, completeness or suitability.

A separate 2025 torque/boost-limiter question reports a plateau and receives a request for better logging and examination of torque demand. The reviewed indexed discussion establishes no verified remedy. Its reported numbers should not become B58-wide torque or boost ceilings. A complete software identity was not established from the reviewed posts.

A useful investigation record

The following is a proposed evidence workflow, not a tested calibration sequence:

Identify the combination. Record engine variant, ECU hardware and software/calibration identifiers, transmission software, installed fuel and turbo hardware, fuel composition and any custom code. Keep the original file identifiable. Match the definitions. Document each signal's units, reference and update rate. Decode active-limit flags using documentation for that ROM and logger; a familiar numeric flag is insufficient. Separate demand from delivery. Where exposed, compare pedal/torque demand, permitted torque, filling request and actual filling, throttle, pressure request/actual, wastegate control and active-limit status on a common time base. Check combustion and fuel delivery. Compare lambda request/measurement, fuel trims, requested/actual rail pressure, injection timing or duration where available, ignition corrections and temperatures. A falling pressure trace needs diagnosis before further demand. Retest the explanation. Change one justified calibration relationship, document the expected trace change, and compare repeatable conditions. Include part-load response and shifts alongside full-load output.

For the general concepts see MG1 torque, air and fuel basics. This page supplies no universal pressure, lambda, ignition, torque or protection settings.

Injection monitoring can become the active load limit

EcuTek's B58g1-onwards injection-monitoring note describes protection of both pump delivery and the available injection window. A pump-control angle reaching its limit, an excessive rail-pressure shortfall, or the calculated pump-flow ceiling can each trigger intervention. The flow calculation includes pump capacity, efficiency and fuel-temperature correction. Consequently, an angle below its maximum does not rule out this path.

The source describes a fast reduction and slower restoration of permitted filling. It also warns that raising the minimum filling permitted by this monitor can conceal the restriction while allowing substantial pressure loss. For a changed pump, establish the correct capacity model for that hardware rather than treating the minimum-load setting as pump calibration.

Diagnostic inference: Capture the first pressure/flow or window restriction and the subsequent filling-limit change. A later log may show pressure recovered because demand has already been reduced. Recovered pressure alone does not establish that the pump was never restrictive. Check low-pressure supply, fuel and operating conditions alongside the high-pressure system.

Use a limit-reason channel only where it is defined

EcuTek's limit-state reference defines the cylinder-filling reason monitor for B58g2/Supra and S58, and explicitly excludes B58g1, N55 and S55 from that monitor's availability. The following is a selection from that reference, not a Bosch-wide enumeration:

ValueMeaning in this EcuTek reference 0Normal torque-demand-derived filling without another restriction. 2Turbocharger component protection involving pressure ratio. 3High-pressure fuel-system control limit. 4Electric lift-pump capacity. 5Intake-temperature, knock or driving-behaviour retard contribution. 6Turbocharger exhaust-temperature protection. 8Available injection window cannot accommodate the requested fuel.

Confirm the ROM and logger definition before decoding a value. An unavailable monitor does not establish that its underlying function is absent. The reason code is a starting point for investigating a restriction, not a reason to disable that restriction.

Separate the boost model from the final actuator command

EcuTek's B58 guide describes a path from required airflow and compressor pressure ratio to shaft-power demand, then turbine/exhaust relationships and wastegate position. Its logging definitions distinguish base, corrective and final turbo-power quantities from the wastegate command. It also separates pre-throttle pressure from manifold pressure and identifies positive boost error as being below target in that tool.

The practical interpretation is to locate the discrepancy. If the filling request has already fallen, more wastegate command does not restore the missing request. If the pressure request remains high while measured pressure falls short, examine delivery, actuator behaviour and model accuracy. If the base estimate is wrong and feedback is compensating heavily, changing feedback gains alone can leave the underlying mismatch. These are diagnostic hypotheses to test with the matched signals.

Put injection duration and angle on the same basis

For an arithmetic check, crank-angle travel is duration in milliseconds × rpm × 0.006. At 6,000 rpm, one millisecond spans 36 crank degrees. This does not define the usable injection window: the software's start/end references, pulse splitting and relevant limits still determine which part of the cycle is available. A displayed duration, an end angle and a pump-control angle are different quantities.

For example, if a log indicates a window restriction, compare all relevant pulses and the applicable angle limits before concluding that rail pressure alone is the problem. Avoid transferring start/end-angle sign conventions from another engine or logger.

Public log case: rail-pressure deficit and recoveryThe October 2023 B58 log-review discussion links this bootmod3 recording, displayed as MultiMap Stage 2+.csv. Selected public-table samples were inspected on 17 November 2026. Full ECU hardware/ROM and map-revision identifiers were not established.Viewer elapsed timerpmPedal %Rail targetRail actual 7.713 s31429928832971 10.624 s45519928832449 12.033 s51339928832894Pressure units are shown as psig by the viewer. The middle sample is 434 psi below demand, while the later sample has recovered. This supports checking the sequence rather than declaring the pump adequate from a later point. It does not identify a failed part or a universal pump-capacity threshold. These are table readings, not a downloaded or independently validated CSV.Discussion case: a limiter flag is not a universal reason numberIn the 26 October discussion, an MHD Stage1 RON98 participant reports values 131072, 2048 and 4, associating them with the load ceiling, rev-limit reduction and speed control respectively. The contributor also relays a response attributed to MHD/PureBoost about load shaping and throttle closure. That response was not obtained directly from the vendor in this review.Those reported bit values must not be decoded using the 0–8 filling-reason enumeration above. The linked Datazap testrun1 now returned a landing page in this review, so its raw samples were not checked. Treat the post as an example of questions to resolve with the correct logger/ROM definition, not a verified bit table for all MHD or BMW software.Select the channels needed to test the explanationMG Flasher's parameter documentation distinguishes default selections from optional channels. Its “None” selection status does not mean a function is unsupported. Injection-monitoring status, fuel-related filling limits, pressure demand/actual and injection timing or mass-flow channels can require explicit selection. Verify availability and units on the actual engine/software; the catalogue of parameter names is not a guarantee that every BMW ROM exposes every channel.A separate Bimmerprofs B58 fuel-system capture compares requested/actual rail pressure and pump-valve phase during acceleration through the first three gears. The author observes changing phase with an apparently constant PWM indication. The useful distinction is between the displayed signal and the physical control being investigated. This is an independent specialist's case with no complete ROM identity established here; its quoted pressure range is not a replacement specification for B58TU hardware.Published software fixes with identifiable map familiesbootmod3's release notes record the following corrections. The family labels are the vendor's, not complete BMW hardware/ROM identifiers:Software scopeReported correction B58-T0-C OTS v2.2, 18 March 2020Reworked load/boost control and addressed throttle closure after gear changes. B58-G OTS v2.1, 10 May 2020Addressed low-rpm overboost associated with a midrange output dip. Gen2 B58 update, 22 August 2023Addressed boost-plausibility behaviour on mild-hybrid applications; the entry does not give a complete ROM list or map revision. B58-T0-C v2.7, 25 March 2024; Stage2 E50 OTS/multimapsChanged end-of-injection behaviour to prevent a potential torque restriction.These are documented vendor-reported fixes, not independent before/after tests by this wiki. Check the exact installed map revision and supported ROM first. An issue addressed in one map family does not justify copying its tables into another file. The earlier public Stage2+ log lacks the identifiers needed to match it to these fixes.A factory correction for a matched hesitation faultBMW SIB 12 13 18, revision 4 describes rough running or hesitation from rest with fault 21A02D. It lists B58M among its covered engines. Its combined model list is F22/F23/F30/F31/F32/F33/F34/F36 and G01/G12/G30/G32; confirm the actual engine/chassis combination rather than treating this as support for every B58 derivative.BMW identifies a DME software error and specifies ISTA 4.27.10 or later, integration level 20-11-500 or later, followed by both prescribed Valvetronic reset/relearning test plans. It warns that omitting either plan can allow the fault to return. Other faults must be diagnosed first; parts replacement solely for this matched error pattern is not the prescribed correction.For tuning research, this is a baseline software/service-state check. It does not prove that a tuned vehicle with a different fault pattern needs the same remedy, or that raising a torque limit will fix hesitation.An owner-reported resolution supported by cam-position checksIn B58 Bucking, a 2018 F30 340xi owner reports intake-cam actual position oscillating roughly 10–15 degrees around demand, while exhaust position tracked closely. A repeated ISTA test eventually failed. On 25 February 2022 the owner reports that replacing the intake VANOS actuator resolved the surging. The intake/downpipe had been returned to stock; the full ROM was not identified.This is a reported repair outcome, not a universal B58 parts recommendation. Read the later follow-up before treating earlier suggestions in the same thread as successful fixes.

Sources and review limits

BMW Service Training: B58TU Engine, information status July 2018; printed pages 27–33. Manufacturer-authored document on an independent mirror. bootmod3: Tuning Guide — MG1 PPC / Aurix, introduction and Load and Torque sections. First-party tuning-tool documentation; its worked example and recommendations are application dependent. NefMoto: B58 engine on MG1, tuning resources, September 2023. NefMoto: B58 MG1 torque/boost limiters?, February–March 2025. Unresolved discussion in the reviewed material. dmacpro91: BMW-XDFs, public repository overview. EcuTek: Injection Monitoring (B58g1 onwards), updated 17 June 2022. Pump/window monitoring and filling-limit behaviour. EcuTek: BMW Limit Flags & States. Tool-specific reason values and monitor availability. EcuTek: B58 Tuning Guide, updated 13 September 2023. Model-based boost control and logging definitions. EcuTek sources reviewed 17 November 2026.Bimmerpost: B58 Log Review Thread, page 177, posts dated 25–31 October 2023. Owner context, software claims and linked recordings. bootmod3 public log 6540144a1ed066592c33fe7c, viewer title MultiMap Stage 2+.csv. Selected table rows inspected; CSV export required sign-in and was not obtained. MG Flasher: Logging Parameters. Channel-selection meanings and parameter descriptions. Bimmerprofs: B58 Fuel system. Author's diagnostic capture and interpretation. Additional cases reviewed 17 November 2026.bootmod3: Release Notes, entries dated 18 March 2020, 10 May 2020, 22 August 2023 and 25 March 2024. Historical, family-specific vendor corrections. BMW SIB 12 13 18, revision 4, 2021; published by NHTSA. Matched 21A02D fault, engine/model scope and software/service requirements. Bimmerpost: B58 Bucking, page 3, 24–25 February 2022 follow-up. Owner-reported cam-actuator repair; full software identity unavailable. Sources reviewed 17 November 2026.BMW Service Training: B58TU Engine, information status July 2018; printed pages 27–33. Manufacturer-authored document on an independent mirror. bootmod3: Tuning Guide — MG1 PPC / Aurix, introduction and Load and Torque sections. First-party tuning-tool documentation; its worked example and recommendations are application dependent. NefMoto: B58 engine on MG1, tuning resources, September 2023. NefMoto: B58 MG1 torque/boost limiters?, February–March 2025. Unresolved discussion in the reviewed material. dmacpro91: BMW-XDFs, public repository overview. EcuTek: Injection Monitoring (B58g1 onwards), updated 17 June 2022. Pump/window monitoring and filling-limit behaviour. EcuTek: BMW Limit Flags & States. Tool-specific reason values and monitor availability. EcuTek: B58 Tuning Guide, updated 13 September 2023. Model-based boost control and logging definitions. EcuTek sources reviewed 17 November 2026.

Sources reviewed 17 November 2026. NefMoto material was available through public search indexing; direct forum access was blocked. No forum attachments or confidential function documents were retrieved.

Rendered revisions

Open each version to inspect formatting, tables and photographs.

Revision 2: BMW B58 and B58TU: fuel-system and MG1 tuning checks

Applicability and evidence

This guide concerns BMW B58 petrol engines discussed as MG1 applications. Hardware evidence comes from BMW's B58TU training; calibration evidence comes from bootmod3 and EcuTek documentation and public tuning discussions. These sources answer different questions. No binary, definition or vehicle has been validated by this wiki for this article.

Source/applicationWhat it establishesBoundary
BMW B58TU training, July 2018B58B30M1 and B58B30O1 hardware context, revised direct injection and DME 8.6T0 designation.Factory training, not a software-specific map definition or current repair instruction. BMW source.
bootmod3 BMW MG1 guideThe vendor groups B58 Gen 1 under PPC and Gen 2 under Aurix; its worked example is a Z4 M40i B58D.Check the actual ECU, ROM and tool version; those group names do not identify every vehicle. Vendor source.
NefMoto B58 resource discussion, September 2023Discussion of definitions, strategy knowledge and logging.Contributor observations, not a verified Bosch function description. Discussion.

The fuel-system generation matters

BMW describes a revision from 200 to 350 bar, with a revised high-pressure pump and injectors. Its comparison distinguishes HDEV5, HDEV5 EVO and the Bosch HDEV6 used on Bx8TU. The DME generation has a common external appearance while internal hardware varies. See printed pages 28 and 31–33 of the training document.

The 350 bar figure describes that factory system's design. It is not a request to enter into every B58 calibration. Identify the fitted pump, injectors, pressure sensor and matching calibration before considering pressure changes. An upgraded pump alone does not establish compatibility of the complete earlier system with a later pressure specification. This is the practical inference from BMW's documented hardware changes.

Read the BMW strategy before following another MG1 platform

In the bootmod3 example, requested torque is related to load through reference-torque calibration. The guide documents separate filling constraints associated with operating conditions, plus Aurix logging of filling-limit reasons and a torque-limit status channel. ROM-dependent differences matter. This supports diagnosing the active path before increasing more limits; it does not establish common addresses or status-bit meanings across BMW software.

Some vendor example settings involve very high limits or monitor changes. They are not verified component ratings or a general prescription for this wiki. Retain a documented reason for each change and investigate an intervention before deciding whether its calibration is wrong.

What the discussions add

The 2023 NefMoto thread contains disagreement about learning from other platforms. One contributor specifically argues that BMW petrol application software differs from other manufacturers' MG1 implementations and has continuity with earlier BMW petrol software. Treat that as a contributor's architectural claim, not proof that any two binaries match. The useful research rule is to establish the function in the BMW file rather than transfer a Porsche or Audi map name.

The public BMW-XDFs repository separates B58gen1 and B58gen2 directories. That organisation illustrates the need for matching definitions. Only the repository overview was reviewed here; individual definitions were not audited for addresses, completeness or suitability.

A separate 2025 torque/boost-limiter question reports a plateau and receives a request for better logging and examination of torque demand. The reviewed indexed discussion establishes no verified remedy. Its reported numbers should not become B58-wide torque or boost ceilings. A complete software identity was not established from the reviewed posts.

A useful investigation record

The following is a proposed evidence workflow, not a tested calibration sequence:

  1. Identify the combination. Record engine variant, ECU hardware and software/calibration identifiers, transmission software, installed fuel and turbo hardware, fuel composition and any custom code. Keep the original file identifiable.
  2. Match the definitions. Document each signal's units, reference and update rate. Decode active-limit flags using documentation for that ROM and logger; a familiar numeric flag is insufficient.
  3. Separate demand from delivery. Where exposed, compare pedal/torque demand, permitted torque, filling request and actual filling, throttle, pressure request/actual, wastegate control and active-limit status on a common time base.
  4. Check combustion and fuel delivery. Compare lambda request/measurement, fuel trims, requested/actual rail pressure, injection timing or duration where available, ignition corrections and temperatures. A falling pressure trace needs diagnosis before further demand.
  5. Retest the explanation. Change one justified calibration relationship, document the expected trace change, and compare repeatable conditions. Include part-load response and shifts alongside full-load output.

For the general concepts see MG1 torque, air and fuel basics. This page supplies no universal pressure, lambda, ignition, torque or protection settings.

Injection monitoring can become the active load limit

EcuTek's B58g1-onwards injection-monitoring note describes protection of both pump delivery and the available injection window. A pump-control angle reaching its limit, an excessive rail-pressure shortfall, or the calculated pump-flow ceiling can each trigger intervention. The flow calculation includes pump capacity, efficiency and fuel-temperature correction. Consequently, an angle below its maximum does not rule out this path.

The source describes a fast reduction and slower restoration of permitted filling. It also warns that raising the minimum filling permitted by this monitor can conceal the restriction while allowing substantial pressure loss. For a changed pump, establish the correct capacity model for that hardware rather than treating the minimum-load setting as pump calibration.

Diagnostic inference: Capture the first pressure/flow or window restriction and the subsequent filling-limit change. A later log may show pressure recovered because demand has already been reduced. Recovered pressure alone does not establish that the pump was never restrictive. Check low-pressure supply, fuel and operating conditions alongside the high-pressure system.

Use a limit-reason channel only where it is defined

EcuTek's limit-state reference defines the cylinder-filling reason monitor for B58g2/Supra and S58, and explicitly excludes B58g1, N55 and S55 from that monitor's availability. The following is a selection from that reference, not a Bosch-wide enumeration:

ValueMeaning in this EcuTek reference
0Normal torque-demand-derived filling without another restriction.
2Turbocharger component protection involving pressure ratio.
3High-pressure fuel-system control limit.
4Electric lift-pump capacity.
5Intake-temperature, knock or driving-behaviour retard contribution.
6Turbocharger exhaust-temperature protection.
8Available injection window cannot accommodate the requested fuel.

Confirm the ROM and logger definition before decoding a value. An unavailable monitor does not establish that its underlying function is absent. The reason code is a starting point for investigating a restriction, not a reason to disable that restriction.

Separate the boost model from the final actuator command

EcuTek's B58 guide describes a path from required airflow and compressor pressure ratio to shaft-power demand, then turbine/exhaust relationships and wastegate position. Its logging definitions distinguish base, corrective and final turbo-power quantities from the wastegate command. It also separates pre-throttle pressure from manifold pressure and identifies positive boost error as being below target in that tool.

The practical interpretation is to locate the discrepancy. If the filling request has already fallen, more wastegate command does not restore the missing request. If the pressure request remains high while measured pressure falls short, examine delivery, actuator behaviour and model accuracy. If the base estimate is wrong and feedback is compensating heavily, changing feedback gains alone can leave the underlying mismatch. These are diagnostic hypotheses to test with the matched signals.

Put injection duration and angle on the same basis

For an arithmetic check, crank-angle travel is duration in milliseconds × rpm × 0.006. At 6,000 rpm, one millisecond spans 36 crank degrees. This does not define the usable injection window: the software's start/end references, pulse splitting and relevant limits still determine which part of the cycle is available. A displayed duration, an end angle and a pump-control angle are different quantities.

For example, if a log indicates a window restriction, compare all relevant pulses and the applicable angle limits before concluding that rail pressure alone is the problem. Avoid transferring start/end-angle sign conventions from another engine or logger.

Sources and review limits

Sources reviewed 17 November 2026. NefMoto material was available through public search indexing; direct forum access was blocked. No forum attachments or confidential function documents were retrieved.

Revision 4: BMW B58 and B58TU: fuel-system and MG1 tuning checks

Applicability and evidence

This guide concerns BMW B58 petrol engines discussed as MG1 applications. Hardware evidence comes from BMW's B58TU training; calibration evidence comes from bootmod3 and EcuTek documentation and public tuning discussions. These sources answer different questions. No binary, definition or vehicle has been validated by this wiki for this article.

Source/applicationWhat it establishesBoundary
BMW B58TU training, July 2018B58B30M1 and B58B30O1 hardware context, revised direct injection and DME 8.6T0 designation.Factory training, not a software-specific map definition or current repair instruction. BMW source.
bootmod3 BMW MG1 guideThe vendor groups B58 Gen 1 under PPC and Gen 2 under Aurix; its worked example is a Z4 M40i B58D.Check the actual ECU, ROM and tool version; those group names do not identify every vehicle. Vendor source.
NefMoto B58 resource discussion, September 2023Discussion of definitions, strategy knowledge and logging.Contributor observations, not a verified Bosch function description. Discussion.

The fuel-system generation matters

BMW describes a revision from 200 to 350 bar, with a revised high-pressure pump and injectors. Its comparison distinguishes HDEV5, HDEV5 EVO and the Bosch HDEV6 used on Bx8TU. The DME generation has a common external appearance while internal hardware varies. See printed pages 28 and 31–33 of the training document.

The 350 bar figure describes that factory system's design. It is not a request to enter into every B58 calibration. Identify the fitted pump, injectors, pressure sensor and matching calibration before considering pressure changes. An upgraded pump alone does not establish compatibility of the complete earlier system with a later pressure specification. This is the practical inference from BMW's documented hardware changes.

Read the BMW strategy before following another MG1 platform

In the bootmod3 example, requested torque is related to load through reference-torque calibration. The guide documents separate filling constraints associated with operating conditions, plus Aurix logging of filling-limit reasons and a torque-limit status channel. ROM-dependent differences matter. This supports diagnosing the active path before increasing more limits; it does not establish common addresses or status-bit meanings across BMW software.

Some vendor example settings involve very high limits or monitor changes. They are not verified component ratings or a general prescription for this wiki. Retain a documented reason for each change and investigate an intervention before deciding whether its calibration is wrong.

What the discussions add

The 2023 NefMoto thread contains disagreement about learning from other platforms. One contributor specifically argues that BMW petrol application software differs from other manufacturers' MG1 implementations and has continuity with earlier BMW petrol software. Treat that as a contributor's architectural claim, not proof that any two binaries match. The useful research rule is to establish the function in the BMW file rather than transfer a Porsche or Audi map name.

The public BMW-XDFs repository separates B58gen1 and B58gen2 directories. That organisation illustrates the need for matching definitions. Only the repository overview was reviewed here; individual definitions were not audited for addresses, completeness or suitability.

A separate 2025 torque/boost-limiter question reports a plateau and receives a request for better logging and examination of torque demand. The reviewed indexed discussion establishes no verified remedy. Its reported numbers should not become B58-wide torque or boost ceilings. A complete software identity was not established from the reviewed posts.

A useful investigation record

The following is a proposed evidence workflow, not a tested calibration sequence:

  1. Identify the combination. Record engine variant, ECU hardware and software/calibration identifiers, transmission software, installed fuel and turbo hardware, fuel composition and any custom code. Keep the original file identifiable.
  2. Match the definitions. Document each signal's units, reference and update rate. Decode active-limit flags using documentation for that ROM and logger; a familiar numeric flag is insufficient.
  3. Separate demand from delivery. Where exposed, compare pedal/torque demand, permitted torque, filling request and actual filling, throttle, pressure request/actual, wastegate control and active-limit status on a common time base.
  4. Check combustion and fuel delivery. Compare lambda request/measurement, fuel trims, requested/actual rail pressure, injection timing or duration where available, ignition corrections and temperatures. A falling pressure trace needs diagnosis before further demand.
  5. Retest the explanation. Change one justified calibration relationship, document the expected trace change, and compare repeatable conditions. Include part-load response and shifts alongside full-load output.

For the general concepts see MG1 torque, air and fuel basics. This page supplies no universal pressure, lambda, ignition, torque or protection settings.

Injection monitoring can become the active load limit

EcuTek's B58g1-onwards injection-monitoring note describes protection of both pump delivery and the available injection window. A pump-control angle reaching its limit, an excessive rail-pressure shortfall, or the calculated pump-flow ceiling can each trigger intervention. The flow calculation includes pump capacity, efficiency and fuel-temperature correction. Consequently, an angle below its maximum does not rule out this path.

The source describes a fast reduction and slower restoration of permitted filling. It also warns that raising the minimum filling permitted by this monitor can conceal the restriction while allowing substantial pressure loss. For a changed pump, establish the correct capacity model for that hardware rather than treating the minimum-load setting as pump calibration.

Diagnostic inference: Capture the first pressure/flow or window restriction and the subsequent filling-limit change. A later log may show pressure recovered because demand has already been reduced. Recovered pressure alone does not establish that the pump was never restrictive. Check low-pressure supply, fuel and operating conditions alongside the high-pressure system.

Use a limit-reason channel only where it is defined

EcuTek's limit-state reference defines the cylinder-filling reason monitor for B58g2/Supra and S58, and explicitly excludes B58g1, N55 and S55 from that monitor's availability. The following is a selection from that reference, not a Bosch-wide enumeration:

ValueMeaning in this EcuTek reference
0Normal torque-demand-derived filling without another restriction.
2Turbocharger component protection involving pressure ratio.
3High-pressure fuel-system control limit.
4Electric lift-pump capacity.
5Intake-temperature, knock or driving-behaviour retard contribution.
6Turbocharger exhaust-temperature protection.
8Available injection window cannot accommodate the requested fuel.

Confirm the ROM and logger definition before decoding a value. An unavailable monitor does not establish that its underlying function is absent. The reason code is a starting point for investigating a restriction, not a reason to disable that restriction.

Separate the boost model from the final actuator command

EcuTek's B58 guide describes a path from required airflow and compressor pressure ratio to shaft-power demand, then turbine/exhaust relationships and wastegate position. Its logging definitions distinguish base, corrective and final turbo-power quantities from the wastegate command. It also separates pre-throttle pressure from manifold pressure and identifies positive boost error as being below target in that tool.

The practical interpretation is to locate the discrepancy. If the filling request has already fallen, more wastegate command does not restore the missing request. If the pressure request remains high while measured pressure falls short, examine delivery, actuator behaviour and model accuracy. If the base estimate is wrong and feedback is compensating heavily, changing feedback gains alone can leave the underlying mismatch. These are diagnostic hypotheses to test with the matched signals.

Put injection duration and angle on the same basis

For an arithmetic check, crank-angle travel is duration in milliseconds × rpm × 0.006. At 6,000 rpm, one millisecond spans 36 crank degrees. This does not define the usable injection window: the software's start/end references, pulse splitting and relevant limits still determine which part of the cycle is available. A displayed duration, an end angle and a pump-control angle are different quantities.

For example, if a log indicates a window restriction, compare all relevant pulses and the applicable angle limits before concluding that rail pressure alone is the problem. Avoid transferring start/end-angle sign conventions from another engine or logger.

Public log case: rail-pressure deficit and recovery

The October 2023 B58 log-review discussion links this bootmod3 recording, displayed as MultiMap Stage 2+.csv. Selected public-table samples were inspected on 17 November 2026. Full ECU hardware/ROM and map-revision identifiers were not established.

Viewer elapsed timerpmPedal %Rail targetRail actual
7.713 s31429928832971
10.624 s45519928832449
12.033 s51339928832894

Pressure units are shown as psig by the viewer. The middle sample is 434 psi below demand, while the later sample has recovered. This supports checking the sequence rather than declaring the pump adequate from a later point. It does not identify a failed part or a universal pump-capacity threshold. These are table readings, not a downloaded or independently validated CSV.

Discussion case: a limiter flag is not a universal reason number

In the 26 October discussion, an MHD Stage1 RON98 participant reports values 131072, 2048 and 4, associating them with the load ceiling, rev-limit reduction and speed control respectively. The contributor also relays a response attributed to MHD/PureBoost about load shaping and throttle closure. That response was not obtained directly from the vendor in this review.

Those reported bit values must not be decoded using the 0–8 filling-reason enumeration above. The linked Datazap testrun1 now returned a landing page in this review, so its raw samples were not checked. Treat the post as an example of questions to resolve with the correct logger/ROM definition, not a verified bit table for all MHD or BMW software.

Select the channels needed to test the explanation

MG Flasher's parameter documentation distinguishes default selections from optional channels. Its “None” selection status does not mean a function is unsupported. Injection-monitoring status, fuel-related filling limits, pressure demand/actual and injection timing or mass-flow channels can require explicit selection. Verify availability and units on the actual engine/software; the catalogue of parameter names is not a guarantee that every BMW ROM exposes every channel.

A separate Bimmerprofs B58 fuel-system capture compares requested/actual rail pressure and pump-valve phase during acceleration through the first three gears. The author observes changing phase with an apparently constant PWM indication. The useful distinction is between the displayed signal and the physical control being investigated. This is an independent specialist's case with no complete ROM identity established here; its quoted pressure range is not a replacement specification for B58TU hardware.

Published software fixes with identifiable map families

bootmod3's release notes record the following corrections. The family labels are the vendor's, not complete BMW hardware/ROM identifiers:

Software scopeReported correction
B58-T0-C OTS v2.2, 18 March 2020Reworked load/boost control and addressed throttle closure after gear changes.
B58-G OTS v2.1, 10 May 2020Addressed low-rpm overboost associated with a midrange output dip.
Gen2 B58 update, 22 August 2023Addressed boost-plausibility behaviour on mild-hybrid applications; the entry does not give a complete ROM list or map revision.
B58-T0-C v2.7, 25 March 2024; Stage2 E50 OTS/multimapsChanged end-of-injection behaviour to prevent a potential torque restriction.

These are documented vendor-reported fixes, not independent before/after tests by this wiki. Check the exact installed map revision and supported ROM first. An issue addressed in one map family does not justify copying its tables into another file. The earlier public Stage2+ log lacks the identifiers needed to match it to these fixes.

A factory correction for a matched hesitation fault

BMW SIB 12 13 18, revision 4 describes rough running or hesitation from rest with fault 21A02D. It lists B58M among its covered engines. Its combined model list is F22/F23/F30/F31/F32/F33/F34/F36 and G01/G12/G30/G32; confirm the actual engine/chassis combination rather than treating this as support for every B58 derivative.

BMW identifies a DME software error and specifies ISTA 4.27.10 or later, integration level 20-11-500 or later, followed by both prescribed Valvetronic reset/relearning test plans. It warns that omitting either plan can allow the fault to return. Other faults must be diagnosed first; parts replacement solely for this matched error pattern is not the prescribed correction.

For tuning research, this is a baseline software/service-state check. It does not prove that a tuned vehicle with a different fault pattern needs the same remedy, or that raising a torque limit will fix hesitation.

An owner-reported resolution supported by cam-position checks

In B58 Bucking, a 2018 F30 340xi owner reports intake-cam actual position oscillating roughly 10–15 degrees around demand, while exhaust position tracked closely. A repeated ISTA test eventually failed. On 25 February 2022 the owner reports that replacing the intake VANOS actuator resolved the surging. The intake/downpipe had been returned to stock; the full ROM was not identified.

This is a reported repair outcome, not a universal B58 parts recommendation. Read the later follow-up before treating earlier suggestions in the same thread as successful fixes.

Sources and review limits

Sources reviewed 17 November 2026. NefMoto material was available through public search indexing; direct forum access was blocked. No forum attachments or confidential function documents were retrieved.