Tuning Wiki
Applies to

Check the hardware and software limits stated in this article.

Why a good settled result is not the whole test

A boost controller has to handle both a changing request and conditions in which requested and actual pressure have settled. A calibration can track well once settled yet overshoot or respond slowly during spool, a pedal change or a gear change. Forum LDRPID discussion includes such reports; check the control strategy and channels for the exact ME7 software rather than assuming all similarly named maps behave identically.

First establish the turbo, wastegate and hoses, N75, charge system, sensor ranges and fuel/ignition condition. Record engine speed, pedal and throttle, gear or dyno load, intake and coolant temperature, ambient pressure, actual cam state, software and adaptation state. Match pressure location and absolute versus gauge units.

Cam state and operating conditions change the response

In a forum-reported Audi A4 B7 1.8T example, settled behaviour was acceptable but overshoot or undershoot occurred during spool. The earlier duty-characterisation logs had been made with variable cam timing off at about 20 °C; the later operation had cam timing on at about 0 °C. The different conditions were suspected, further logging was planned, and gathering data with the intended cam switching was advised. The report does not isolate which change caused the response, and the wiki has not reproduced the logs.

The practical lesson is to compare cam state, temperature, gear and load as well as RPM. A duty-to-pressure observation gathered in one state may not predict another. Do not change multiple controller maps just because two non-equivalent runs differ.

Fixed duty does not fix the operating conditions. In a forum-reported modified 1.8T case, back-to-back runs at a reported fixed duty showed different boost as temperatures changed after a stop. Different explanations involving intake and exhaust temperature were proposed; the cause was not isolated. Record the actual duty trace, heat-soak and prior operating history, intake and coolant temperatures, cam state, gear/load, ambient conditions and hardware before comparing runs. A repeatable pressure difference is evidence to investigate, not proof of one temperature mechanism.

Check which boost-controller mode is active

The translated Bosch Audi 132 kW 1.8T ME7.1 LDRPID function description distinguishes quasi-steady PI operation from a more strongly acting dynamic PID mode. In that documented function the derivative contribution is switched off in the quasi-steady mode; an error-dependent condition selects dynamic operation and its return has hysteresis. Therefore a D contribution that is absent once pressure settles does not by itself prove a failed derivative map. The behaviour and mode threshold must be established for the actual software.

When available, record B_lddy (dynamic-mode condition), B_ldr (boost-control active), pressure error and the individual P, I and D signals through the event. The reference says lde is set to zero when control is inactive, so a zero trace without the active flag cannot establish good tracking. Confirm logger definitions, sample rate and time alignment; a short overshoot or state change can fall between samples. The function description concerns the cited Audi application and its calibration procedure is not a general-purpose gain recipe.

Compare like with like

Log patternQuestion to investigate
Requested pressure is steady; actual settles below itIs the relevant pressure sensor in range? Are leaks, turbo/wastegate capability, duty limits, integral contribution or fault states limiting the response?
Pressure briefly exceeds request during spool but later settlesCompare the time history of request, measured pressure and controller contributions. Examine actuator and turbo response, operating conditions and P/D behaviour; one spike does not identify a single faulty map.
The request itself fallsInspect permitted load/torque, temperature, knock and diagnostic interventions before interpreting it as a controller tracking error.
Similar request but different duty between runsCheck cam state, ambient conditions, gear/load, adaptation, sensor scaling and hardware. Final duty combines several effects.

Useful channel groups include plsol_w and pvdks_w when correctly defined for upstream requested and measured pressure; pedal, throttle, RPM and temperatures; ldtvr_w and final duty where available; and the relevant P, I, D, limit and adaptation signals. Check names, units and sampling for the software. Include time before and after the event. ME7Logger alias reference; ME7 boost and load log review.

Review an event during rapid RPM rise or in only one gear with the torque intervention guide. For the reported possibility that pressure-range clipping can affect learned boost-control behavior, see 5120 pressure scaling and log interpretation.

For the separate effects of charge temperature, pressure request and timing, see ME7 intake temperature and boost interpretation.

Gear-change spikes and regulator signals. In a forum-reported 1.8T AJQ case, the posted log lacked individual PID variables, which limited diagnosis. The owner later reported improvement after more than one controller-related change, so the case does not isolate a universal correction. Capture the controller contributions, limits and final duty together with the comparable pressure request and measurement; compare the shift transient with the settled region. The wastegate hardware guide covers actuator and solenoid differences.

In the Bosch Audi ME7.1 LDRPID description, lde is zeroed outside active boost regulation, and the current integral limit can include adaptation and other corrections. Capture the control-active state and the I, P, D, limit and pre/post-linearisation signals when available; see ME7 boost adaptation and integral limits. A single final-duty value cannot identify the active contribution.

Community pre-control methods

A long community LDRPID thread describes a method that deliberately arranges KFLDIMX and KFLDRL to provide an approximate duty-based pre-control. The method rests on specific assumptions and applications; that interpretation is not a general definition of a factory table. Forced-duty observations, axis changes or copied map values can have serious effects and require appropriate equipment, a matched definition and independent validation. Keep diagnostics and protection active and follow the real response rather than treating a spreadsheet or final-duty average as proof.

For the standard request and limit chain see ME7 torque, load and boost control. Community reports are examples, not evidence that a particular adjustment will work on another vehicle.

Continue the conversation

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