Tuning Wiki
Applies to

Check the hardware and software limits stated in this article.

Scope and starting point

This guide brings together tuning and diagnosis findings for EA888 Simos 18 from community discussion and cited tool documentation. The examples are mainly Mk7-platform cars and include Simos 18.1; they do not establish that every Simos 18 variant has the same strategies or maps. Forum material consists of individual vehicle reports and opinions, not independently verified calibration recipes. Before comparing cases, record the engine, gearbox, hardware, ECU variant, box code, software and calibration identifiers, existing patches, fuel and modifications. See Simos 18 for the family overview and tool and file scope for documented tool coverage.

Torque, air charge and the gearbox

Community discussion of the Mk7 GTI describes the selected air-mass setpoint and reference indicated-torque maps as related inverse relationships for torque control (NefMoto Mk7 GTI thread). The maps involved depend on operating configuration including valve lift, cams and flap position. This is a guide to understanding the model, not proof that similarly named maps are exact mathematical inverses in every operating state. Check the matching definition and function documentation, axes, units and active mode.

Indicated torque, clutch/output torque, requested torque and the ECU's estimated torque are different quantities. Community discussion of load and torque logging (NefMoto logging thread) stresses that the reverse torque estimate is modelled and may differ from a dyno measurement, and warns about underscaling the model on a DSG car. The transmission's torque information and behaviour need to be considered alongside engine calibration; an artificially low displayed value is not evidence that the actual load on the drivetrain is low.

Log the limit, and check the logger

Start with a healthy baseline and record faults before and after a controlled comparison. Where the software and logger expose them, capture engine speed, gear and pedal position alongside requested, limited and estimated torque; requested and actual or estimated cylinder air charge; upstream-throttle and manifold pressures; wastegate target and position; lambda target and actual; ignition and knock correction; fuel-pressure target and actual; and intake, coolant and oil temperatures. Add the relevant intervention and diagnostic states where available. Select a manageable group of channels and record the sampling rate so brief events are not hidden by slow updates. Channel names and availability depend on the application.

An air-charge display that stops at about 1,389 mg/stroke has been reported for a diagnostic channel (logging thread). The ceiling was attributed to that diagnostic representation, with RAM logging suggested, and converting an available airflow-in-g/s channel was described as an alternative. These are forum reports about particular channels, not a proven universal ECU air-charge limit. Check the channel definition, scaling, units, range and sample rate, and cross-check a suitable independently defined signal before deciding that the ECU or sensor is saturated. A converted airflow figure must use the correct channel meaning, engine configuration and units.

Pressure-channel matching and transient comparisons

Requested and actual boost can appear not to meet when channels are compared at different locations. In a 2019 Golf R log with an IE Stage 1 ECU tune and DQ381, one forum interpretation was that the charge-pipe upstream-throttle target and actual agreed while the manifold pair differed. This is an unconfirmed interpretation of one log, not independent confirmation of its channels or proof that a leak or fault is absent. The useful check is to match the location, units, pressure reference and requested stage of each signal before comparing traces; review throttle and the driving event as well. A mismatch should not be generalised as normal without diagnosis.

A 2015 Mk7 GTI IE Stage 1 report describes a large requested/actual gap around 3,000–4,000 rpm. The explanation offered was that the target rose before the turbo reached it, and that absolute pressure and wastegate polarity can be misread; the owner accepted this. No leak test or independent root cause was given, and the ECU software is not stated. Compare the traces over time, pedal and operating state, and measured actuator feedback where available; a transient lag does not prove a leak, and eventual target agreement alone does not rule one out.

For a pressure channel explicitly defined as absolute, gauge pressure equals absolute pressure minus the measured local ambient pressure in the same units. For example, 2,700 mbar absolute at an illustrative 1,013 mbar ambient is 1,687 mbar gauge, about 24.5 psi; use the actual ambient and channel definitions rather than automatically subtracting a fixed value. A logger's actuator percentage may describe a request, feedback, opening, closing or duty: establish its definition and polarity before interpreting a saturation. COBB's MQB monitor descriptions distinguish pressure, setpoint and actuator channels for its supported applications.

Lift-off and shift transients

A Simos 18.1 manual-transmission report describes the throttle remaining open on an in-gear lift after driver torque request fell to zero, while clutch-in shifting behaved differently. Throttle angle alone does not establish whether the engine is still making positive torque. The cause was not confirmed, and attempts to alter state transitions were reported to have side effects.

An in-gear lift should be separated from a clutch-in shift when recording a complaint. Where available, align pedal and clutch state, gear, torque and air-charge requests, throttle position, pressure, ignition, fuel-cut and operating state. A separate manual GTI report describes intermittent torque reduction after shifts despite normal dyno pulls; no cause or fix was established. A steady pull alone may miss a transient. Unexpected acceleration should be investigated as a fault rather than assumed to be normal rev hang.

Boost and wastegate diagnosis

Compare pressure target with the corresponding actual pressure at the same location and in the same absolute or gauge units, then examine actuator target and position, throttle behaviour, faults and fuel pressure through the event. A pressure mismatch can have mechanical, calibration or measurement causes; it does not on its own identify which one.

Forum discussion of wastegate strategy (wastegate strategy thread) covers turbine-model flow-factor axes and RAM logging when tuning pre-control. One hybrid-turbo case reported pressure-control and rail-pressure issues, with improvement after model and pre-control changes. This is one self-reported case, not a table to copy to another vehicle.

On a larger turbo, EanDem describes two scaling factors that affect where the ECU looks up the wastegate setpoint (bpa_sp); the post leaves the factors unnamed. With the original control strategy, plot the intake and exhaust flow factors against the corresponding table axes through spool and at high rpm. If either reaches the last breakpoint, record what happens to requested and actual wastegate position at that moment. EanDem's boost-control account. EcuTek's EA888 logging channels.

Whether the file contains a boost-control patch should be established before interpreting its axes. For example, VehiCAL's SIMOS/SDI boost patch describes changing the pre-control axes to engine speed and boost setpoint while retaining PID feedback; its instructions say that the axes and values need to be populated. Other patches may work differently. Verify the actual patch, file definition and retained protections rather than assuming a stock map or another product's behaviour.

Pressure sensor changes and high-rpm fall-off

A dual pressure-sensor swap report describes poor running and rich lambda after a manifold sensor change even though the pressure reading appeared plausible. The integrated temperature sensing was suspected, but no verified root cause was established. Check the exact sensor part number, its pressure range and transfer characteristic, wiring, ECU scaling and any integrated temperature element for the actual location. Compare engine-off pressure against a suitable barometric reference and check readings across the relevant operating range; an accurate-looking boost reading alone does not validate the whole sensor installation.

Separately, forum discussion (Simos 18 tips thread) suggests that modelled turbo-speed intervention or pumping-loss effects on torque control may contribute to a high-rpm boost decline. These are diagnostic hypotheses, not confirmed causes for a particular car. Where available, compare pressure request and actual with torque-intervention and modelled turbo-speed or limit states, and check physical system condition before considering a calibration change. A protection should not be removed merely to make the decline disappear.

Fuel system and ethanol

The intended hardware and fuel should be decided before selecting a fuelling upgrade. An MHH fuel-upgrade discussion compares high-pressure pump work with additional port injection and low-pressure supply changes, and reports idle calibration difficulty with large port injectors. That is an experience report, not a universal injector-size limit or maintenance interval. Identify the installed direct and port injection hardware, characterise the injectors for the actual calibration and monitor pressure, lambda and trims across the operating range.

An experimental MPI retrofit discussion describes coordinating injection modes, switching conditions and injector data; one earlier attempt showed a commanded port-injector pulse but poor idle and a lambda mismatch. A pulse-width value alone does not establish correct fuelling. Hardware and software support must be checked for the particular engine and application.

A Simos 18 flex-fuel thread distinguishes reading ethanol content and adjusting fuel from adding ethanol-dependent ignition timing; one contributor reported implementing the latter with a custom patch. The EcuTek EA888 flex-fuel documentation describes sensor and other ethanol-content options for its supported applications. A sensor, a particular calibration flag or an ignition blend should not be assumed present or enabled merely because another Simos 18 file has it.

Interpreting trims after fuel-system changes

EanDem reports that an engine with XDS1050 port injectors at 5.5 bar idled better with an 80% direct and 20% port fuel split. The proposed explanation for negative trims was a software-dependent minimum injection duration. At stable idle, check whether final pulse width stays at a floor as requested fuel falls. Record the active injection mode, pressure and lambda correction alongside both injector commands. The reported split describes that installation. EanDem's idle test, 26 October 2025. See also MPI injector calibration.

At wide-open throttle the same source reports a brief positive trim followed by sustained negative correction. Fuel collecting on the intake valve was suspected, and both port fuel pressure and injection phasing were changed; with two variables changed, the cause remains open. The trim reversal should be aligned with measured and requested lambda, the direct/port split and port injection timing relative to valve opening, after first confirming that lambda feedback was active during the event. EanDem's full-load observation. See also MPI injection timing and phasing.

A CaracalTech logged 2017 Golf Simos 18.1 case shows short-term fuel trim continuing to change during enriched full-load operation while measured lambda tracked the target. The original and modified pulls were made at different temperatures, and the reported zero long-term trim may reflect reset or incomplete learning. This is one reported calibration, not proof that all Simos 18 files use closed-loop control at full load. When diagnosing fuelling, log target and measured lambda, short- and long-term trims, the control state and fuel pressure together where available; a correct lambda trace alone does not show how much correction was required or establish injector headroom.

A related dual-injection discussion reports pressure loss between feed and port-injector rail in one plumbing arrangement under increased flow. The fuel circuit and the location of the pressure reading should be identified for the actual vehicle; a reading upstream of a restriction may not represent pressure at the rail. Verify under appropriate load; factory and retrofit layouts should not be assumed identical.

When port-injector pulse width approaches the available injection window under boost, compare fuel flow and commanded low-side pressure with measured port-rail and manifold pressure. The pump guide shows the flow-based ceiling in one Simos 18.1 definition and how to calculate pressure across the injector. Low-pressure fuel pump control and upgrades. MHH Auto low-pressure discussion.

Catalyst efficiency after an exhaust change

EanDem describes rich and lean phases of an oxygen-storage check after fitting a sport catalyst. Volkswagen's 2014 Golf and GTI reference gives an example of P0420 using measured oxygen storage relative to a borderline catalyst. For the car under test, identify the applicable monitor and record its completion status, result and associated faults. Check exhaust integrity and oxygen-sensor operation before interpreting a low result. Forum account. Volkswagen Golf and GTI reference, p. 89.

Retrofit hardware and displayed measurements

The same forum source also covers external or vacuum wastegates, flapless manifolds, alternative fuel controllers and alcohol fuels. For a conversion, establish which controller operates each actuator and which feedback reaches the ECU. EanDem mentions an exhaust-gas-temperature input on a car without a petrol particulate filter and describes a software change for continuous oil-pressure measurement. Verify the ECU hardware, wiring diagram and signal definition before treating a logged value as a sensor reading. EanDem's hardware observations. See exhaust temperature modelling and oil pressure control.

EanDem also reports displaying ECU data through Android Auto or Apple CarPlay with an ELM327-compatible adapter and a phone; the app is unnamed. For diagnosis, check which channels the combination can record and the time between samples before using it to interpret a short transient. Forum display example.

Worked example: a recurring high-load fault

A Simos 18.1 Audi A3 case involved a modified turbo and fuel system, power loss above 6,000 rpm and P060C00, with freeze-frame data supplied. Identifying the exact fault and its software-specific monitor before drawing conclusions was suggested. Improvement after changes was initially reported, but the fault later recurred at low load, so the case does not establish a general fix for P060C00.

For a comparable complaint, save the complete diagnostic report and freeze frame, correlate the event with time-aligned logs and identify the applicable manufacturer fault entry and conditions in the matching software documentation where available. Diagnose hardware and calibration together. Monitoring thresholds should not be disabled or broadly raised solely because a forum example mentions them; a similar symptom or fault code can have a different cause.

For map-switch selection, overrun crackles and DSG shift sounds, see Simos 18 map switching, crackles and DSG shift sound.

Detailed valve-train guides: camshaft timing · valve lift tuning.

Detailed air-path guides: air charge modelling · throttle control.

For compressor-speed modelling, temperature and protection-state interpretation, see Simos 18 turbocharger protection.

For cylinder roughness detection, misfire counters and calibration examples, see Simos 18 misfire detection and diagnosis. For manifold and upstream pressure conversion checks, see MAP and boost pressure sensors.

For oil-system fault diagnosis, see oil pressure control and piston cooling jet control.

For monitored torque deviations, process-monitoring map examples and safety intervention signals, see Simos 18 torque monitoring and safety plausibility.

Worked example: two faults on a tuned 2018 GTI

A 230 hp GTI with new plugs and coils showed vibration in sixth gear at 1,800–2,400 rpm and an EPC warning in third or fourth at 6,300–6,500 rpm. A separate event record should be kept for each symptom. For the high-rpm fault, save the full cylinder-specific fault codes and freeze frames, then align misfire counts, lambda, fuel pressures and torque interventions. Air-to-torque scaling or a torque limiter was proposed, but no confirmed repair was recorded. MHH Auto help desk, posts 12–13. See misfire detection and diagnosis.

In the same source, a stock Leon Cupra and Golf R both showed roughly 80% pedal and throttle at full pedal. Each channel, its scaling and the requested and measured throttle positions should be identified before diagnosing restricted opening. Opinions differ on GTI and R/S3 strategies, so the application software, gearbox and turbo calibration should be matched before transferring an interpretation between cars. Help desk, posts 10–11 and 14–15.

For a sound during a DSG upshift, first establish when the gearbox requested torque reduction. EanDem describes fuel cut and ignition retard during the shift. For a sound on lift-off in Sport, forum discussion instead considers impulse combustion and whether the ECU and gearbox software support the selected mode. Gear, shift state, drive mode, pedal position, fuel cut and ignition angle should be recorded to identify which event occurred. Help-desk shift discussion. Sport-mode follow-up.

Logging transport and event resolution

The Simos Wiki overview distinguishes broadcast CAN data, individual diagnostic memory reads and a dynamic identifier that groups multiple values in one response. It describes a Eurodyne high-speed logger as using a patched ECU routine; availability, permissions, addresses and scaling depend on the ECU software and logger. A value absent from a broadcast channel is not necessarily unavailable through another supported diagnostic method.

More channels can reduce the update rate when requests share limited throughput. For illustration, 30 separate responses per second divided evenly among 10 individual channels gives about three updates per second per channel; actual performance depends on grouping, bus load and implementation. Inspect timestamps and per-channel update rates, and use a smaller set of relevant signals for brief interventions. A grouped response can reduce request overhead, but its sampling synchronisation and real rate should be measured, not assumed.

Two community high-speed logger implementations

The SimosHighSpeedLogger project, linked by Simos Wiki, describes a UDS logger developed around Eurodyne Maestro on MQB with a compatible ASW. Its parameter file specifies address, data length, signedness, scaling and units; the project says addresses and scaling depend on the ECU software version. Match these to the actual ASW and verify values against known signals rather than reusing an arbitrary configuration.

The separate VW_Flash project documents its own simos_hsl.py backends using read-memory, dynamic identifiers and an aftermarket-specific patched service, and states that those backends require application-software patches. The two projects should not be assumed to have identical interfaces, patch requirements or definition files. Confirm the installed software and project documentation before connecting; measure effective channel rates in the resulting log.

Further reading

Not established

Forum reports and proposed fixes cited here have not been independently tested by this wiki. Several cited threads and attachments were only partly accessible (some require registration), so reports reflect available pages and indexed text reviewed on 17 November 2026.

Continue the conversation

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