Removed text is struck through; added text is highlighted.
What MAF calibration does
The mass airflow (MAF) calibration translates the sensor signal into an estimate of air mass for the installed sensor and intake. Depending on the ECU, this estimate contributes to cylinder charge, load, fuel, ignition and torque calculations. It is not simply a fuel adjustment. A replacement sensor, housing, airbox or pipework can change the relationship between signal and flow. Identify the exact ECU software, sensor and housing before selecting a curve. See airflow metering for sensor types.
A transfer curve may express voltage or frequency against grams per second, kg/h or a scaled value. Confirm axes, units, offsets, limits and any later corrections in the actual definition. Some sensors report frequency rather than voltage, and a frequency signal is not necessarily a mass measurement: a Karman vortex meter measures volume and needs the relevant density calculation. A different sensor or housing cannot be assumed to share a curve. The Haltech MAF documentation describes voltage and frequency inputs and sensor curves.
Examples across ECU strategies
SystemCalibration approach and qualification
Bosch ME7 petrolIn the documented ME7 examples MLHFM converts the MAF signal; MLOFS and additional correction functions affect the result. The effective air mass also affects load. Verify the exact software and calculation order before scaling; see ME7 fueling and MAF calibration.
GM example using frequencySome GM applications use airflow versus MAF frequency. EFILive's Calc.VE tutorial shows a frequency-indexed MAF calibration. Other airflow models may participate depending on the operating state and ECU; confirm the active strategy rather than assuming MAF is the sole input.
Subaru BRZ / Toyota 86 exampleEcuTek documents a voltage-to-g/s scaling map for the specified BRZ/FRS/GT86 family. Its tuning guide discusses changes for the housing or intake. This is an application example, not a table definition for every Subaru or Toyota ECU.
Nissan 350Z VQ35 exampleEcuTek describes a multi-stage conversion on the covered ROMs: voltage to a percentage, then a maximum-flow conversion and further airflow corrections. Check bank, ROM and the actual calculation rather than assuming a single direct voltage-to-g/s table. See EcuTek's VQ35 strategy guide.
Diesel air managementAir mass can be used for air-path and EGR monitoring and fuel-quantity limiting, depending on the ECU. Diesel operation is generally lean and petrol-style stoichiometric fuel-trim scaling is not a general diesel method. Compare specified and actual air mass under the manufacturer's test conditions and diagnose the air path before changing a curve. The Ross-Tech TDI procedure provides an application-specific logging example; see also Volkswagen TDI air-mass measurement.
Standalone ECUSome support a configured voltage or frequency input and a loaded sensor curve. Haltech documents these choices and a MAF fuel-load mode; other standalone systems may instead use speed density or another load strategy. Follow the specific controller's documentation and use data for the installed sensor and housing. Haltech MAF setup.
Worked examples: calculating and applying a correction
These figures are invented to illustrate the arithmetic; they are not recommended calibration values. First establish the fuel and sensor health, ECU strategy, exact table axes, units and correction sign as described below. A mixture error does not prove a MAF error. Retain the original file, make changes only in the evidenced region, and re-log the result.
1. Subaru BRZ / Toyota 86: voltage and closed-loop trims
The EcuTek guide describes this application's voltage-to-g/s table. Suppose a stable 2.00 V bin has a table value of 30.0 g/s and an established combined fuel correction of +6%: in this example positive means the ECU adds fuel. If diagnosis supports an underestimated MAF and the applicable correction acts proportionally, the starting estimate is:
factor = 1 + 6 / 100 = 1.06
new airflow at 2.00 V = 30.0 × 1.06 = 31.8 g/s
Apply 31.8 g/s to the corresponding voltage point, review neighbouring bins using their own adequate stable samples, and check interpolation and curve shape. Do not copy 6% across the entire curve. Log again at similar conditions and check the bank and learning state. This example assumes the displayed +6% is already the effective combined correction; do not blindly add STFT and LTFT. Their sign, representation, learning regions and combination depend on the ECU and logger. See EcuTek's BRZ MAF guide.
2. GM frequency-output example: measured versus commanded lambda
For a GM application with a frequency-to-airflow table, as shown in the EFILive Calc.VE tutorial, suppose the table gives 200 g/s at 8,000 Hz. During a stable point the final commanded lambda is 0.80 and a calibrated, time-aligned wideband reads 0.84. Assume the MAF is the active load input, injector and pressure data are correct, feedback and learned fuel corrections are accounted for, and the error is attributable to the MAF:
approximate factor = measured lambda / commanded lambda
= 0.84 / 0.80 = 1.05
candidate airflow = 200 × 1.05 = 210 g/s at 8,000 Hz
The mixture is 5% leaner in this simplified ratio; the candidate adds 5% to reported air and consequently calculated fuel under the stated model. Apply only after confirming the table and active airflow model for that operating system. Inspect neighbouring frequency bins and re-test with a wideband, fuel pressure and load monitoring. If lambda becomes dangerously lean, stop the test and diagnose; do not continue a loaded pull to populate a histogram. The same calculation can suggest a reduction: 0.76 / 0.80 = 0.95, or 5% less, under the same assumptions.
3. Bosch ME7: include the MLOFS offset
The ME7 guide describes a documented relationship in which effective air mass equals MLHFM minus MLOFS. Suppose a particular point has MLHFM 120 units, MLOFS 10 units, and a justified factor of 1.05. Use the actual definition's units and confirm that this relationship applies to the software:
old effective air mass = 120 − 10 = 110
new effective air mass = 110 × 1.05 = 115.5
new MLHFM value = 115.5 + 10 = 125.5
Write the calculated 125.5, rounded according to the editor's precision, into the applicable MLHFM point; retain MLOFS in this example. Directly multiplying 120 by 1.05 gives 126 and changes the offset's contribution as well. This arithmetic illustrates the offset, not a claim that every ME7 has a value of 10 or that every correction is applied at this stage. Verify calculation order, corrections and load effects, then re-log. The MxT ME7 guide documents one software-specific way to group and analyse trim data.
4. Nissan 350Z VQ35: a percentage stage
EcuTek's VQ35 guide describes voltage-to-percentage and percentage-to-g/s stages on covered ROMs, followed by further corrections. For arithmetic only, suppose a particular point reads 40% and a separately established 1.05 factor is appropriate at that stage. The candidate is 40% × 1.05 = 42%. This does not by itself establish the final g/s: trace the subsequent conversion, bank and corrections, and confirm the editing tool's representation and recommended adjustment location. Do not assume a percentage table is a direct airflow table or change a shared maximum-flow scalar to fix one local bin.
5. Standalone ECU: choosing the correction location
Haltech documents a configurable sensor curve and, in MAF tuning mode, a Fuel Base correction factor that starts at 1.000. Suppose a valid MAF curve point reads 50.0 g/s and an independent flow reference for the same installed sensor and housing establishes 52.5 g/s: 52.5 / 50.0 = 1.05, so the candidate sensor-curve point is 52.5 g/s. Alternatively, a base-fuel factor of 1.050 would apply a 5% correction in the documented fuel calculation, but the resulting effects and conditions are not necessarily equivalent throughout the ECU. Do not apply both 1.05 changes for the same error: 1.05 × 1.05 = 1.1025, or 10.25% combined. Use the appropriate correction location and verify the resulting fuel and load behaviour. See Haltech MAF setup and Fuel Base documentation.
6. Diesel: use a known flow reference
Suppose an independent, appropriate flow-bench reference measures 440 kg/h for the complete sensor and housing at a stable signal, while the ECU conversion at that signal is 400 kg/h. If the definition is a direct, proportional flow value at that signal, 440 / 400 = 1.10, giving a candidate of 440 kg/h. Repeat at multiple flow points and verify the sensor, units, setup and reproducibility before changing the curve. In a vehicle, a requested 440 and measured 400 kg/h are not an independent 1.10 calibration factor: EGR, boost, leaks, restrictions and control conditions can explain the difference. Follow the vehicle-specific test, such as the Ross-Tech TDI logging procedure, and retain emissions and diagnostic functions.
Which maps to find: ME7 and EDC16
The MAF transfer curve is the primary calibration for a changed sensor or housing. Other maps may need checking because they consume the resulting air mass, but they are not automatically changed. Exact names, units, axes, availability and addresses depend on the ECU variant, software and definition. Identify the software and confirm the function against its matching documentation before editing.
Bosch ME7: example map roles
Map or parameterAction and reason
MLHFMPrimary MAF voltage-to-air-mass characteristic in the documented ME7 implementations. For a verified sensor/housing change, revise the points for which a valid new transfer curve is available. Confirm the definition's axis, output units and interpolation.
MLOFSAir-mass offset in the documented relationship. Check it when calculating new MLHFM values; it does not necessarily need changing. The worked ME7 example above shows how to scale the effective value while retaining the offset.
KFKHFMAdditional air-mass correction in the documented ME7 strategy. Inspect its axes, values and role so its contribution is not overlooked or applied twice. Change it only where the application-specific evidence supports that correction; do not automatically use it to hide injector or pressure errors.
KRKTE, TVUB and FKKVSFuel and injector calibration functions, not MAF transfer curves. Check the injector flow, voltage offset and fuel-system correction when diagnosing mixture error; modify them only for a separately established fuel-system change or error.
For example, if the identified MLHFM point is 120, MLOFS is 10 in the same units and an independently justified effective-air factor is 1.05, enter (120 − 10) × 1.05 + 10 = 125.5 at that point, subject to the actual map's scaling and precision. Review adjacent points and downstream calculated load; re-log. This is an illustrative calculation, not a stock map value. The map roles and offset caveats are explained in ME7 fueling and MAF calibration; the MxT ME7 guide describes a logging-based MLHFM workflow for its supported applications.
Bosch EDC16: identify the specific sensor conversion
For an analogue EDC16 application, look in the matching definition for the MAF sensor characteristic / air-mass linearisation: a voltage input and an airflow output. This is the candidate curve to modify when a documented replacement sensor and housing have a different transfer function. The published Trifonov and Hadjiivanov sensor study experimentally reports an EDC16U1 diagnostic voltage-to-flow characteristic. It does not identify the calibration map's Bosch name, address or raw scaling, so those cannot be inferred from the diagnostic values. A separate VAG EDC16 map list calls one MAF linearisation curve AFSCD_dmLin_CUR; it does not identify the software or provide its axis scaling. Do not assume that all EDC16 files use this name, call the curve MLHFM, or share ME7's offset.
For example, that study reports 43.91 g/s at 3.0 V for its tested EDC16U1 setup. If an independently calibrated reference for a different, installed sensor and housing instead measures 46.11 g/s at the same signal, the illustrative ratio is 46.11 / 43.91 ≈ 1.0501. A direct g/s table point would therefore be set to approximately 46.11 g/s, provided the correct table and its scaling have been verified. If the definition uses kg/h, multiply by 3.6: 43.91 g/s is 158.08 kg/h and 46.11 g/s is 166.00 kg/h. These are a published baseline and a hypothetical replacement measurement, not a transferable curve. Build and verify the whole curve from multiple known points rather than extrapolating a single ratio.
Also inspect, but do not automatically alter, the MAF-based fuel-quantity or smoke-limiting maps, EGR air-mass targets and plausibility ranges in the exact file. They use or compare air mass; they do not convert the sensor signal into air mass. Correcting the transfer curve can change the values seen by those functions, so verify their behaviour in logs. Requested versus actual airflow is not an independent calibration reference. Some EDC16 variants use frequency-output sensors, so a voltage curve cannot be assumed for the whole family. If the definition does not reliably identify the conversion curve or its scaling, stop and obtain a matching definition or validated measurement rather than guessing an address or map name.
Named EDC16 examples and axes
These names are search aids for a matching definition, not universal addresses or verified raw conversion factors. A curve changes the reported sensor flow; air-path targets, substitute values and smoke limits have different jobs.
Identifier and scopeInput and outputHow to use it
AFSCD_dmLin_CUR, VAG EDC16 listAUTODTC's VAG EDC16 list labels this a 32×1 MAF linearisation curve. It does not give signal units, output scaling, exact ECU variant or software ID.Search the matching definition for this identifier. Verify whether its input is voltage, frequency or period, and whether its output is kg/h, g/s or scaled raw counts. Change the verified flow values to match the replacement sensor and housing; do not assume 32 points or a particular scale in another file.
EDC16C39, Alfa 159 1.9 JTDm exampleAliant Auto's example describes the MAF conversion as input frequency versus air mass in kg/h and lists AFSCD_dmLin_CUR among useful IDs. It does not explicitly link the ID to the displayed curve or publish its numerical axes and scaling, so the association is a likely candidate, not a verified map definition.The author suggests data from the donor 2.4 JTDm MAF and intake in this specific conversion. Confirm the exact sensor, housing and installation, sensor signal, units and both ECU definitions; compare the donor curve at the recipient's actual breakpoints and interpolate only if supported. Validate the result in service.
AFSCD_mAirPerCylDfl_MAP, VAG listThe same list calls this an 8×8 substitute air-mass map used on error; its precise axes, units and software are not given.This is a fallback value, not the healthy sensor's transfer curve. Do not change it to compensate for an inaccurate MAF. Check which signal and quality state the logger displays.
FlMng_qAFSCDSmoke_MAP, VAG listThe list describes a 13×16 MAF-based smoke fuel limiter but does not supply exact axes, output units or software ID. Confirm these in the particular definition.A consumer of air mass, not the sensor linearisation. Recheck its operating range and any active limiting after the airflow is corrected; do not scale it as a substitute for calibrating the MAF.
AFSCD_dmAirSubsVal_MAP, BMW E90 335d EDC16CP35The documented case identifies a substitute-air lookup indexed by engine speed and injected quantity when required inputs fail. Its dimensions and numeric scaling are not established there.Check sensor-quality and selected-signal channels when interpreting a MAF log. A plausible modelled or fallback value is not proof the meter itself is calibrated. Do not edit this fallback to fix the sensor curve.
Applying a donor frequency curve: illustrative EDC16C39 method
Assume the exact recipient definition has been independently confirmed as signal frequency in Hz against kg/h and permits interpolation; these values are invented. At a recipient breakpoint of 4,000 Hz, its current curve says 300 kg/h. A verified donor curve for the complete replacement sensor and housing has 3,500 Hz → 310 kg/h and 4,500 Hz → 370 kg/h. Linear interpolation gives:
fraction = (4,000 − 3,500) / (4,500 − 3,500) = 0.5
candidate flow = 310 + 0.5 × (370 − 310) = 340 kg/h
change at this point = 340 / 300 − 1 = 13.33%
Enter 340 kg/h at the existing 4,000 Hz point, after converting to the actual editor's units and encoding. Repeat across the documented range, preserving the original axis unless the software definition and procedure explicitly support changing it. This example does not establish an EDC16C39 stock or donor value. Use the ECU's documented interpolation rule; do not extrapolate beyond the donor data. If the actual axis is period in microseconds, use that axis and its interpolation rule: frequency and period have an inverse relationship, so directly interpolating in Hz can give a different answer. Check for clipping and headroom, restore normal operation, then log raw signal, measured and selected airflow, quality state, RPM, boost, intake temperature and EGR operation. Validate against an independent reference where possible.
EDC17: MAF linearisation and related maps
The AUTODTC VAG EDC17 map list identifies the names below, but does not identify a specific software version or document the breakpoints, units, table dimensions or raw scaling. Treat the names as candidates to verify in the matching ECU definition. EDC17 variants and sensors do not all use the same signal type or map set.
Candidate nameRole, axes and action
AFSCD_dmLin_CURListed as MAF linearisation. The list does not establish its input axis or output scale for a particular ECU. Confirm whether the actual input is voltage, frequency, period or another representation, and the output's units. For a verified change of sensor and housing, the corresponding flow points are the primary candidate for calibration.
FlMng_qAFSCDSmoke_MAPListed as a MAF-based smoke limiter. The available list does not establish its axes, units or selection logic. It consumes air-mass information and limits fuel; it is not the transfer curve. Check its operation and coverage after a MAF change rather than applying the same percentage to it.
AirCtl_mDesBase0_MAP, AirCtl_mDesBase1_MAP, AirCtl_mDesBase2_MAPListed as air-control/EGR maps for different gear groups. Exact axes and output units must be checked in the application. They are downstream targets, not measured sensor calibration. Check air-path behaviour with emissions controls functioning; do not alter them merely to make a MAF log agree.
AFSCD_swtAirSysVal_CListed as an air-system configuration selector. Its precise values and logic are not established by that list. Confirm the active sensor and air-system configuration; do not change the switch to mask a sensor fault.
Illustrative EDC17 calculation. Suppose a matching definition independently confirms a direct sensor-period axis in microseconds and airflow in kg/h. At 250 µs the recipient curve has 400 kg/h, and an independently verified flow measurement for the replacement sensor and housing at that exact period is 420 kg/h:
factor = 420 / 400 = 1.05
candidate AFSCD_dmLin_CUR output at 250 µs = 400 × 1.05 = 420 kg/h
The period, flow values and verified definition are hypothetical; they are not a claim that a particular EDC17 file uses those breakpoints. Apply the candidate only if that exact file's definition and sensor test support it; measure several points and retain the existing axis unless the definition and method justify changing it. 250 µs corresponds to 4,000 Hz, but interpolation in period is not equivalent to interpolation in frequency. Do not use requested/actual diesel air mass as an independent flow reference. Re-log sensor input, measured and selected air mass, relevant quality state, RPM, pressure, temperature and EGR response. See Bosch ASMod air system model for the distinction between measured and modelled air mass.
MED9.1 2.0 TFSI: period curve and air-mass correction
The MxT project's MED9 description identifies MSHFMTPH as an air-mass transfer characteristic against sensor period in microseconds, with flow in kg/h for the supported MED9 workflow. Its MED9 support is marked alpha and the project says that matching binaries were not supplied for its contributed MED9 definitions, so independently verify the particular ECU software, axes and values. A MED9.1 2.0 TFSI discussion also identifies this period-based signal path; it is community information, not a Bosch function sheet.
Map or parameterAxes, role and action
MSHFMTPHFor the cited MED9 example, period in µs → air mass flow in kg/h. It is the transfer-curve candidate for a validated sensor or housing change. Confirm the actual file's axis direction, breakpoints, output scaling and interpolation before replacing any values; do not import ME7's MLHFM layout or MLOFS arithmetic by assumption.
KFKHFMCommunity discussions identify this as an HFM correction map. Verify both axis variables, dimensionless scaling and where it acts in the matching software; the available examples do not establish universal axes or calculation order. Inspect its contribution and change it only for a separately evidenced correction. A curve change and a correction-map change can compound; avoid applying the same error twice.
KRKATE and fuel-pressure functionsThe cited MxT documentation describes KRKATE as a direct-injection fuel constant with fuel-pressure correction, not a MAF transfer curve. Check injector data, high- and low-pressure supply and actual versus requested rail pressure before attributing a mixture error to air measurement. Change fuel calibration only for a separately established need.
Illustrative MED9.1 correction. Suppose a confirmed MSHFMTPH point at 300 µs reads 360 kg/h. Under a simplified stable petrol operating condition with final commanded lambda 0.80, independently measured lambda 0.84, no masking feedback and verified injector and fuel pressure data, a provisional factor is 0.84 / 0.80 = 1.05. If diagnosis supports a MAF error and the verified transfer output is directly proportional, the candidate at that period is 360 × 1.05 = 378 kg/h. Retain KFKHFM in this example, review adjacent supported points and re-log lambda, raw period, airflow, load, fuel pressure and corrections. The values are invented, and the ratio is not proof of the fault or a universal formula; other corrections and software calculation order may affect it. Do not use idle trims to infer high-load accuracy.
MED17: identify the air strategy first
MED17 is a family, not one air-measurement strategy. Some applications use a MAF, while the MxT MED17 guide describes a pressure-based adaptive air model for its covered applications, including the EA855 EVO workflow. Do not treat its statements as proof that every MED17 lacks a MAF. Identify the exact ECU software, sensor hardware, logging channels and active strategy before choosing a calibration.
Name or functionAxes, role and application
MAF transfer characteristic, where fittedIdentify the exact sensor-signal input and air-mass output in the matching definition. The sources consulted do not establish a single map name or axis for all MAF-equipped MED17 variants. Voltage, frequency or period and physical units must be verified in the specific software. A sensor/housing change calls for a validated transfer characteristic, not an assumed ME7 or MED9 map.
rk_w, MxT MED17 workflowMxT calls this a relative fuel-mass correction and bins logs by engine speed nmot_w and load rl_w. The guide uses a user-selected map definition and a multiplicative output. It does not prove that rk_w is the calibration's Bosch symbol in every file, nor that every MED17 has the same axes. This is fuel correction in the cited workflow, not MAF sensor linearisation.
fupsrls_w / pbrint_wMxT identifies these as signals associated with its adaptive pressure/load model. Treat them as logged model quantities, not automatically as editable calibration maps. Their availability and behaviour are software-dependent.
MAP sensor characteristic and injector calibrationIn a pressure-based strategy, verify the matching MAP sensor conversion, its pressure reference and units, along with temperature, injector and fuel-pressure data. Exact map names and axes require a matching definition. An error in any can resemble an air-model or fuel-trim error.
Illustrative fuel-correction calculation, not a MAF calibration: In the MxT-described workflow, suppose a verified speed/load cell has a relative fuel multiplier of 1.000 and its established combined correction is +4%, where positive means more fuel. The documented operation is 1.000 × (1 + 4/100) = 1.040. Apply only if the selected map is confirmed to act as that multiplier in the exact software and diagnosis supports it; re-log at the same operating point. MxT describes one method for combining its particular STFT and LTFT signals; other ECU/logger conventions may differ. Correcting fuel does not establish that calculated air mass, load or torque is accurate. Check pressure, leaks, injector data, fuel composition, fuel pressure and feedback state first. The numbers are invented. See the MxT fuel-trim algorithm.
MD1: identify the sensor and the matching definition
MD1 variants differ by manufacturer, engine and software. Some use an air-mass sensor and other air-path measurements; do not assume that an EDC17 or MED17 map name, axis or address transfers to MD1. An example vehicle listing identifies the DS 1.5 BlueHDi MD1CS003 and lists a throttle air-mass-flow calculation, but does not provide an independently verified MAF transfer-map name, axis or scaling. A throttle airflow model is not automatically the MAF sensor characteristic.
Function to identifyAxes and action
MAF sensor linearisation, if fittedFind the exact ECU-specific characteristic in a trustworthy matching definition: raw sensor input (voltage, frequency, period or digital quantity) and mass-flow output. Verify units, breakpoints, encoding and active path. A named MD1 MAF map and its axes could not be confirmed from the available sources; AFSCD_dmLin_CUR is an EDC16/17 search name, not a verified MD1 identifier here.
Sensor diagnostic bounds and selected-air qualityCheck raw sensor range, plausibility and whether a diagnostic value is measured, modelled or substituted. The actual names, axes and thresholds must come from that software. An unchanged, valid range should remain enabled; a flow curve edit cannot repair physical saturation or wiring faults.
Air-path targets, smoke/fuel limit and airflow modelsCheck which functions consume the corrected air mass and their coverage, using matching documentation and logs. Their names, inputs and units are not established by the vehicle listing. Do not alter these functions simply to conceal an inaccurate sensor or to force agreement with an air-mass target.
Conditional arithmetic example: If an independently verified MD1 definition for the installed hardware identifies a direct flow-output point at a particular sensor signal as 500 kg/h, and a calibrated reference for that same signal and complete sensor/housing measures 525 kg/h, the candidate output is 500 × (525 / 500) = 525 kg/h. This is a hypothetical calculation, not a documented MD1 map or breakpoint. Use only after verifying the signal representation, linearisation, units, correction chain and interpolation for that software; gather multiple points and do not extrapolate beyond the reference. If those details cannot be verified, do not edit a guessed map. Requested versus actual diesel air mass is not an independent reference. Confirm measured versus modelled air, operating state, EGR response and diagnostics in validation logs.
MD1CS003: specific diagnostic evidence and definition gap
A ScanDoc MD1CS003 diagnostic menu for Peugeot 301 lists a “mass air flow sensor replacement” service function. This is evidence that the tool exposes a replacement procedure for that application; the listing does not explain the procedure, prove it changes the transfer curve or establish that every MD1CS003 vehicle uses it. Consult the current vehicle-specific service instructions before use and record the adaptation state before and after.
A separate ScanDoc Peugeot Traveller MD1CS003 example labels air-flow channels in kg/h and mg/stroke and a target in mg/stroke. Its snapshot contains implausible values in several unrelated channels, so its numerical readings are not calibration evidence. The labels are useful when selecting logger channels, but do not identify map axes, prove signal quality or establish the ECU's internal conversion.
For a specific Peugeot 308 1.5 BlueHDi 130, a seller listing identifies MD1CS003 hardware 9818703480 and software upgrade 9693382980. It does not publish a verified MAF transfer-curve symbol, axis or scaling. Matching hardware alone is insufficient: obtain the exact software and calibration identification and a compatible definition, then verify its map semantics and raw-to-physical conversion before calculating changes. No variant-specific MD1 MAF map name, axis or example values were independently verified in this research; do not transfer the EDC17 AFSCD_dmLin_CUR name or a guessed address to this file.
General validation workflow
Record the original calibration, ECU identification, sensor part number and intake arrangement. Establish whether MAF, speed density or a blend is active in each test region. A speed-density estimate is another model and is useful for comparison only when independently validated.
Correct faults first: unmetered-air and exhaust leaks, sensor supply and grounds, contamination or damage, fuel pressure, injector characterization, fuel composition and lambda sensor or logger errors can all distort the result. Check both banks separately where available.
Use a documented transfer curve for the complete sensor and housing as the starting point. A bore-area ratio is only a rough starting estimate: placement, bends, screens, pulsation and flow distribution matter. Check signal headroom at idle and peak flow; a saturated sensor cannot reveal additional airflow.
At stable, warm closed-loop petrol operating points, log raw voltage or frequency, calculated airflow, RPM, load, requested and measured lambda, short- and long-term corrections and control state. Group samples by signal, reject transients and allow for sensor and exhaust delay. Interpret correction signs and the way learned and immediate trims combine for that ECU; a trim is evidence of a mixture discrepancy, not proof that the MAF caused it.
For regions where feedback is absent or cannot measure the commanded rich mixture, compare final requested lambda with a calibrated, suitably installed wideband while monitoring fuel pressure and engine safety. Feedback and learned corrections can mask an error; establish their actual state and follow the ECU-specific procedure. Do not infer high-load accuracy from idle trims.
Make small, justified changes in well-sampled regions, preserve a smooth physically plausible curve, and re-log. Under a simplified open-loop petrol model with correct injector and pressure data and no active fuel correction, measured lambda divided by commanded lambda is an approximate multiplier for the reported airflow: for example 1.05 / 1.00 suggests about 5% more reported air. This is a diagnostic approximation, not a universal formula: offsets, filtering, blending and ECU correction conventions can change how the stored table must be edited. Investigate large or inconsistent errors before applying it.
Restore any temporary test settings and retain diagnostic and engine-protection functions. Verify idle, part load, transitions and safely controlled higher load, both banks, fuel trims, final lambda, calculated load, plausibility and headroom. Recheck under representative temperatures and conditions. Do not fill unmeasured regions with invented certainty.
See ME7 MAF and lambda log review for a worked logging approach and EcuTek's BRZ intake rescaling guide for its specific MAF and speed-density comparison. Modifying emissions-related calibrations may be restricted by local rules; retain applicable monitoring and verify the requirements for the vehicle's use.
Open each version to inspect formatting, tables and photographs.
Revision 7: MAF calibration across ECU types
What MAF calibration does
The mass airflow (MAF) calibration translates the sensor signal into an estimate of air mass for the installed sensor and intake. Depending on the ECU, this estimate contributes to cylinder charge, load, fuel, ignition and torque calculations. It is not simply a fuel adjustment. A replacement sensor, housing, airbox or pipework can change the relationship between signal and flow. Identify the exact ECU software, sensor and housing before selecting a curve. See airflow metering for sensor types.
A transfer curve may express voltage or frequency against grams per second, kg/h or a scaled value. Confirm axes, units, offsets, limits and any later corrections in the actual definition. Some sensors report frequency rather than voltage, and a frequency signal is not necessarily a mass measurement: a Karman vortex meter measures volume and needs the relevant density calculation. A different sensor or housing cannot be assumed to share a curve. The Haltech MAF documentation describes voltage and frequency inputs and sensor curves.
Examples across ECU strategies
| System | Calibration approach and qualification |
| Bosch ME7 petrol | In the documented ME7 examples MLHFM converts the MAF signal; MLOFS and additional correction functions affect the result. The effective air mass also affects load. Verify the exact software and calculation order before scaling; see ME7 fueling and MAF calibration. |
| GM example using frequency | Some GM applications use airflow versus MAF frequency. EFILive's Calc.VE tutorial shows a frequency-indexed MAF calibration. Other airflow models may participate depending on the operating state and ECU; confirm the active strategy rather than assuming MAF is the sole input. |
| Subaru BRZ / Toyota 86 example | EcuTek documents a voltage-to-g/s scaling map for the specified BRZ/FRS/GT86 family. Its tuning guide discusses changes for the housing or intake. This is an application example, not a table definition for every Subaru or Toyota ECU. |
| Nissan 350Z VQ35 example | EcuTek describes a multi-stage conversion on the covered ROMs: voltage to a percentage, then a maximum-flow conversion and further airflow corrections. Check bank, ROM and the actual calculation rather than assuming a single direct voltage-to-g/s table. See EcuTek's VQ35 strategy guide. |
| Diesel air management | Air mass can be used for air-path and EGR monitoring and fuel-quantity limiting, depending on the ECU. Diesel operation is generally lean and petrol-style stoichiometric fuel-trim scaling is not a general diesel method. Compare specified and actual air mass under the manufacturer's test conditions and diagnose the air path before changing a curve. The Ross-Tech TDI procedure provides an application-specific logging example; see also Volkswagen TDI air-mass measurement. |
| Standalone ECU | Some support a configured voltage or frequency input and a loaded sensor curve. Haltech documents these choices and a MAF fuel-load mode; other standalone systems may instead use speed density or another load strategy. Follow the specific controller's documentation and use data for the installed sensor and housing. Haltech MAF setup. |
Worked examples: calculating and applying a correction
These figures are invented to illustrate the arithmetic; they are not recommended calibration values. First establish the fuel and sensor health, ECU strategy, exact table axes, units and correction sign as described below. A mixture error does not prove a MAF error. Retain the original file, make changes only in the evidenced region, and re-log the result.
1. Subaru BRZ / Toyota 86: voltage and closed-loop trims
The EcuTek guide describes this application's voltage-to-g/s table. Suppose a stable 2.00 V bin has a table value of 30.0 g/s and an established combined fuel correction of +6%: in this example positive means the ECU adds fuel. If diagnosis supports an underestimated MAF and the applicable correction acts proportionally, the starting estimate is:
factor = 1 + 6 / 100 = 1.06
new airflow at 2.00 V = 30.0 × 1.06 = 31.8 g/s
Apply 31.8 g/s to the corresponding voltage point, review neighbouring bins using their own adequate stable samples, and check interpolation and curve shape. Do not copy 6% across the entire curve. Log again at similar conditions and check the bank and learning state. This example assumes the displayed +6% is already the effective combined correction; do not blindly add STFT and LTFT. Their sign, representation, learning regions and combination depend on the ECU and logger. See EcuTek's BRZ MAF guide.
2. GM frequency-output example: measured versus commanded lambda
For a GM application with a frequency-to-airflow table, as shown in the EFILive Calc.VE tutorial, suppose the table gives 200 g/s at 8,000 Hz. During a stable point the final commanded lambda is 0.80 and a calibrated, time-aligned wideband reads 0.84. Assume the MAF is the active load input, injector and pressure data are correct, feedback and learned fuel corrections are accounted for, and the error is attributable to the MAF:
approximate factor = measured lambda / commanded lambda
= 0.84 / 0.80 = 1.05
candidate airflow = 200 × 1.05 = 210 g/s at 8,000 Hz
The mixture is 5% leaner in this simplified ratio; the candidate adds 5% to reported air and consequently calculated fuel under the stated model. Apply only after confirming the table and active airflow model for that operating system. Inspect neighbouring frequency bins and re-test with a wideband, fuel pressure and load monitoring. If lambda becomes dangerously lean, stop the test and diagnose; do not continue a loaded pull to populate a histogram. The same calculation can suggest a reduction: 0.76 / 0.80 = 0.95, or 5% less, under the same assumptions.
3. Bosch ME7: include the MLOFS offset
The ME7 guide describes a documented relationship in which effective air mass equals MLHFM minus MLOFS. Suppose a particular point has MLHFM 120 units, MLOFS 10 units, and a justified factor of 1.05. Use the actual definition's units and confirm that this relationship applies to the software:
old effective air mass = 120 − 10 = 110
new effective air mass = 110 × 1.05 = 115.5
new MLHFM value = 115.5 + 10 = 125.5
Write the calculated 125.5, rounded according to the editor's precision, into the applicable MLHFM point; retain MLOFS in this example. Directly multiplying 120 by 1.05 gives 126 and changes the offset's contribution as well. This arithmetic illustrates the offset, not a claim that every ME7 has a value of 10 or that every correction is applied at this stage. Verify calculation order, corrections and load effects, then re-log. The MxT ME7 guide documents one software-specific way to group and analyse trim data.
4. Nissan 350Z VQ35: a percentage stage
EcuTek's VQ35 guide describes voltage-to-percentage and percentage-to-g/s stages on covered ROMs, followed by further corrections. For arithmetic only, suppose a particular point reads 40% and a separately established 1.05 factor is appropriate at that stage. The candidate is 40% × 1.05 = 42%. This does not by itself establish the final g/s: trace the subsequent conversion, bank and corrections, and confirm the editing tool's representation and recommended adjustment location. Do not assume a percentage table is a direct airflow table or change a shared maximum-flow scalar to fix one local bin.
5. Standalone ECU: choosing the correction location
Haltech documents a configurable sensor curve and, in MAF tuning mode, a Fuel Base correction factor that starts at 1.000. Suppose a valid MAF curve point reads 50.0 g/s and an independent flow reference for the same installed sensor and housing establishes 52.5 g/s: 52.5 / 50.0 = 1.05, so the candidate sensor-curve point is 52.5 g/s. Alternatively, a base-fuel factor of 1.050 would apply a 5% correction in the documented fuel calculation, but the resulting effects and conditions are not necessarily equivalent throughout the ECU. Do not apply both 1.05 changes for the same error: 1.05 × 1.05 = 1.1025, or 10.25% combined. Use the appropriate correction location and verify the resulting fuel and load behaviour. See Haltech MAF setup and Fuel Base documentation.
6. Diesel: use a known flow reference
Suppose an independent, appropriate flow-bench reference measures 440 kg/h for the complete sensor and housing at a stable signal, while the ECU conversion at that signal is 400 kg/h. If the definition is a direct, proportional flow value at that signal, 440 / 400 = 1.10, giving a candidate of 440 kg/h. Repeat at multiple flow points and verify the sensor, units, setup and reproducibility before changing the curve. In a vehicle, a requested 440 and measured 400 kg/h are not an independent 1.10 calibration factor: EGR, boost, leaks, restrictions and control conditions can explain the difference. Follow the vehicle-specific test, such as the Ross-Tech TDI logging procedure, and retain emissions and diagnostic functions.
Which maps to find: ME7 and EDC16
The MAF transfer curve is the primary calibration for a changed sensor or housing. Other maps may need checking because they consume the resulting air mass, but they are not automatically changed. Exact names, units, axes, availability and addresses depend on the ECU variant, software and definition. Identify the software and confirm the function against its matching documentation before editing.
Bosch ME7: example map roles
| Map or parameter | Action and reason |
| MLHFM | Primary MAF voltage-to-air-mass characteristic in the documented ME7 implementations. For a verified sensor/housing change, revise the points for which a valid new transfer curve is available. Confirm the definition's axis, output units and interpolation. |
| MLOFS | Air-mass offset in the documented relationship. Check it when calculating new MLHFM values; it does not necessarily need changing. The worked ME7 example above shows how to scale the effective value while retaining the offset. |
| KFKHFM | Additional air-mass correction in the documented ME7 strategy. Inspect its axes, values and role so its contribution is not overlooked or applied twice. Change it only where the application-specific evidence supports that correction; do not automatically use it to hide injector or pressure errors. |
| KRKTE, TVUB and FKKVS | Fuel and injector calibration functions, not MAF transfer curves. Check the injector flow, voltage offset and fuel-system correction when diagnosing mixture error; modify them only for a separately established fuel-system change or error. |
For example, if the identified MLHFM point is 120, MLOFS is 10 in the same units and an independently justified effective-air factor is 1.05, enter (120 − 10) × 1.05 + 10 = 125.5 at that point, subject to the actual map's scaling and precision. Review adjacent points and downstream calculated load; re-log. This is an illustrative calculation, not a stock map value. The map roles and offset caveats are explained in ME7 fueling and MAF calibration; the MxT ME7 guide describes a logging-based MLHFM workflow for its supported applications.
Bosch EDC16: identify the specific sensor conversion
For an analogue EDC16 application, look in the matching definition for the MAF sensor characteristic / air-mass linearisation: a voltage input and an airflow output. This is the candidate curve to modify when a documented replacement sensor and housing have a different transfer function. The published Trifonov and Hadjiivanov sensor study experimentally reports an EDC16U1 diagnostic voltage-to-flow characteristic. It does not identify the calibration map's Bosch name, address or raw scaling, so those cannot be inferred from the diagnostic values. A separate VAG EDC16 map list calls one MAF linearisation curve AFSCD_dmLin_CUR; it does not identify the software or provide its axis scaling. Do not assume that all EDC16 files use this name, call the curve MLHFM, or share ME7's offset.
For example, that study reports 43.91 g/s at 3.0 V for its tested EDC16U1 setup. If an independently calibrated reference for a different, installed sensor and housing instead measures 46.11 g/s at the same signal, the illustrative ratio is 46.11 / 43.91 ≈ 1.0501. A direct g/s table point would therefore be set to approximately 46.11 g/s, provided the correct table and its scaling have been verified. If the definition uses kg/h, multiply by 3.6: 43.91 g/s is 158.08 kg/h and 46.11 g/s is 166.00 kg/h. These are a published baseline and a hypothetical replacement measurement, not a transferable curve. Build and verify the whole curve from multiple known points rather than extrapolating a single ratio.
Also inspect, but do not automatically alter, the MAF-based fuel-quantity or smoke-limiting maps, EGR air-mass targets and plausibility ranges in the exact file. They use or compare air mass; they do not convert the sensor signal into air mass. Correcting the transfer curve can change the values seen by those functions, so verify their behaviour in logs. Requested versus actual airflow is not an independent calibration reference. Some EDC16 variants use frequency-output sensors, so a voltage curve cannot be assumed for the whole family. If the definition does not reliably identify the conversion curve or its scaling, stop and obtain a matching definition or validated measurement rather than guessing an address or map name.
Named EDC16 examples and axes
These names are search aids for a matching definition, not universal addresses or verified raw conversion factors. A curve changes the reported sensor flow; air-path targets, substitute values and smoke limits have different jobs.
| Identifier and scope | Input and output | How to use it |
AFSCD_dmLin_CUR, VAG EDC16 list | AUTODTC's VAG EDC16 list labels this a 32×1 MAF linearisation curve. It does not give signal units, output scaling, exact ECU variant or software ID. | Search the matching definition for this identifier. Verify whether its input is voltage, frequency or period, and whether its output is kg/h, g/s or scaled raw counts. Change the verified flow values to match the replacement sensor and housing; do not assume 32 points or a particular scale in another file. |
| EDC16C39, Alfa 159 1.9 JTDm example | Aliant Auto's example describes the MAF conversion as input frequency versus air mass in kg/h and lists AFSCD_dmLin_CUR among useful IDs. It does not explicitly link the ID to the displayed curve or publish its numerical axes and scaling, so the association is a likely candidate, not a verified map definition. | The author suggests data from the donor 2.4 JTDm MAF and intake in this specific conversion. Confirm the exact sensor, housing and installation, sensor signal, units and both ECU definitions; compare the donor curve at the recipient's actual breakpoints and interpolate only if supported. Validate the result in service. |
AFSCD_mAirPerCylDfl_MAP, VAG list | The same list calls this an 8×8 substitute air-mass map used on error; its precise axes, units and software are not given. | This is a fallback value, not the healthy sensor's transfer curve. Do not change it to compensate for an inaccurate MAF. Check which signal and quality state the logger displays. |
FlMng_qAFSCDSmoke_MAP, VAG list | The list describes a 13×16 MAF-based smoke fuel limiter but does not supply exact axes, output units or software ID. Confirm these in the particular definition. | A consumer of air mass, not the sensor linearisation. Recheck its operating range and any active limiting after the airflow is corrected; do not scale it as a substitute for calibrating the MAF. |
AFSCD_dmAirSubsVal_MAP, BMW E90 335d EDC16CP35 | The documented case identifies a substitute-air lookup indexed by engine speed and injected quantity when required inputs fail. Its dimensions and numeric scaling are not established there. | Check sensor-quality and selected-signal channels when interpreting a MAF log. A plausible modelled or fallback value is not proof the meter itself is calibrated. Do not edit this fallback to fix the sensor curve. |
Applying a donor frequency curve: illustrative EDC16C39 method
Assume the exact recipient definition has been independently confirmed as signal frequency in Hz against kg/h and permits interpolation; these values are invented. At a recipient breakpoint of 4,000 Hz, its current curve says 300 kg/h. A verified donor curve for the complete replacement sensor and housing has 3,500 Hz → 310 kg/h and 4,500 Hz → 370 kg/h. Linear interpolation gives:
fraction = (4,000 − 3,500) / (4,500 − 3,500) = 0.5
candidate flow = 310 + 0.5 × (370 − 310) = 340 kg/h
change at this point = 340 / 300 − 1 = 13.33%
Enter 340 kg/h at the existing 4,000 Hz point, after converting to the actual editor's units and encoding. Repeat across the documented range, preserving the original axis unless the software definition and procedure explicitly support changing it. This example does not establish an EDC16C39 stock or donor value. Use the ECU's documented interpolation rule; do not extrapolate beyond the donor data. If the actual axis is period in microseconds, use that axis and its interpolation rule: frequency and period have an inverse relationship, so directly interpolating in Hz can give a different answer. Check for clipping and headroom, restore normal operation, then log raw signal, measured and selected airflow, quality state, RPM, boost, intake temperature and EGR operation. Validate against an independent reference where possible.
EDC17: MAF linearisation and related maps
The AUTODTC VAG EDC17 map list identifies the names below, but does not identify a specific software version or document the breakpoints, units, table dimensions or raw scaling. Treat the names as candidates to verify in the matching ECU definition. EDC17 variants and sensors do not all use the same signal type or map set.
| Candidate name | Role, axes and action |
AFSCD_dmLin_CUR | Listed as MAF linearisation. The list does not establish its input axis or output scale for a particular ECU. Confirm whether the actual input is voltage, frequency, period or another representation, and the output's units. For a verified change of sensor and housing, the corresponding flow points are the primary candidate for calibration. |
FlMng_qAFSCDSmoke_MAP | Listed as a MAF-based smoke limiter. The available list does not establish its axes, units or selection logic. It consumes air-mass information and limits fuel; it is not the transfer curve. Check its operation and coverage after a MAF change rather than applying the same percentage to it. |
AirCtl_mDesBase0_MAP, AirCtl_mDesBase1_MAP, AirCtl_mDesBase2_MAP | Listed as air-control/EGR maps for different gear groups. Exact axes and output units must be checked in the application. They are downstream targets, not measured sensor calibration. Check air-path behaviour with emissions controls functioning; do not alter them merely to make a MAF log agree. |
AFSCD_swtAirSysVal_C | Listed as an air-system configuration selector. Its precise values and logic are not established by that list. Confirm the active sensor and air-system configuration; do not change the switch to mask a sensor fault. |
Illustrative EDC17 calculation. Suppose a matching definition independently confirms a direct sensor-period axis in microseconds and airflow in kg/h. At 250 µs the recipient curve has 400 kg/h, and an independently verified flow measurement for the replacement sensor and housing at that exact period is 420 kg/h:
factor = 420 / 400 = 1.05
candidate AFSCD_dmLin_CUR output at 250 µs = 400 × 1.05 = 420 kg/h
The period, flow values and verified definition are hypothetical; they are not a claim that a particular EDC17 file uses those breakpoints. Apply the candidate only if that exact file's definition and sensor test support it; measure several points and retain the existing axis unless the definition and method justify changing it. 250 µs corresponds to 4,000 Hz, but interpolation in period is not equivalent to interpolation in frequency. Do not use requested/actual diesel air mass as an independent flow reference. Re-log sensor input, measured and selected air mass, relevant quality state, RPM, pressure, temperature and EGR response. See Bosch ASMod air system model for the distinction between measured and modelled air mass.
MED9.1 2.0 TFSI: period curve and air-mass correction
The MxT project's MED9 description identifies MSHFMTPH as an air-mass transfer characteristic against sensor period in microseconds, with flow in kg/h for the supported MED9 workflow. Its MED9 support is marked alpha and the project says that matching binaries were not supplied for its contributed MED9 definitions, so independently verify the particular ECU software, axes and values. A MED9.1 2.0 TFSI discussion also identifies this period-based signal path; it is community information, not a Bosch function sheet.
| Map or parameter | Axes, role and action |
MSHFMTPH | For the cited MED9 example, period in µs → air mass flow in kg/h. It is the transfer-curve candidate for a validated sensor or housing change. Confirm the actual file's axis direction, breakpoints, output scaling and interpolation before replacing any values; do not import ME7's MLHFM layout or MLOFS arithmetic by assumption. |
KFKHFM | Community discussions identify this as an HFM correction map. Verify both axis variables, dimensionless scaling and where it acts in the matching software; the available examples do not establish universal axes or calculation order. Inspect its contribution and change it only for a separately evidenced correction. A curve change and a correction-map change can compound; avoid applying the same error twice. |
KRKATE and fuel-pressure functions | The cited MxT documentation describes KRKATE as a direct-injection fuel constant with fuel-pressure correction, not a MAF transfer curve. Check injector data, high- and low-pressure supply and actual versus requested rail pressure before attributing a mixture error to air measurement. Change fuel calibration only for a separately established need. |
Illustrative MED9.1 correction. Suppose a confirmed MSHFMTPH point at 300 µs reads 360 kg/h. Under a simplified stable petrol operating condition with final commanded lambda 0.80, independently measured lambda 0.84, no masking feedback and verified injector and fuel pressure data, a provisional factor is 0.84 / 0.80 = 1.05. If diagnosis supports a MAF error and the verified transfer output is directly proportional, the candidate at that period is 360 × 1.05 = 378 kg/h. Retain KFKHFM in this example, review adjacent supported points and re-log lambda, raw period, airflow, load, fuel pressure and corrections. The values are invented, and the ratio is not proof of the fault or a universal formula; other corrections and software calculation order may affect it. Do not use idle trims to infer high-load accuracy.
MED17: identify the air strategy first
MED17 is a family, not one air-measurement strategy. Some applications use a MAF, while the MxT MED17 guide describes a pressure-based adaptive air model for its covered applications, including the EA855 EVO workflow. Do not treat its statements as proof that every MED17 lacks a MAF. Identify the exact ECU software, sensor hardware, logging channels and active strategy before choosing a calibration.
| Name or function | Axes, role and application |
| MAF transfer characteristic, where fitted | Identify the exact sensor-signal input and air-mass output in the matching definition. The sources consulted do not establish a single map name or axis for all MAF-equipped MED17 variants. Voltage, frequency or period and physical units must be verified in the specific software. A sensor/housing change calls for a validated transfer characteristic, not an assumed ME7 or MED9 map. |
rk_w, MxT MED17 workflow | MxT calls this a relative fuel-mass correction and bins logs by engine speed nmot_w and load rl_w. The guide uses a user-selected map definition and a multiplicative output. It does not prove that rk_w is the calibration's Bosch symbol in every file, nor that every MED17 has the same axes. This is fuel correction in the cited workflow, not MAF sensor linearisation. |
fupsrls_w / pbrint_w | MxT identifies these as signals associated with its adaptive pressure/load model. Treat them as logged model quantities, not automatically as editable calibration maps. Their availability and behaviour are software-dependent. |
| MAP sensor characteristic and injector calibration | In a pressure-based strategy, verify the matching MAP sensor conversion, its pressure reference and units, along with temperature, injector and fuel-pressure data. Exact map names and axes require a matching definition. An error in any can resemble an air-model or fuel-trim error. |
Illustrative fuel-correction calculation, not a MAF calibration: In the MxT-described workflow, suppose a verified speed/load cell has a relative fuel multiplier of 1.000 and its established combined correction is +4%, where positive means more fuel. The documented operation is 1.000 × (1 + 4/100) = 1.040. Apply only if the selected map is confirmed to act as that multiplier in the exact software and diagnosis supports it; re-log at the same operating point. MxT describes one method for combining its particular STFT and LTFT signals; other ECU/logger conventions may differ. Correcting fuel does not establish that calculated air mass, load or torque is accurate. Check pressure, leaks, injector data, fuel composition, fuel pressure and feedback state first. The numbers are invented. See the MxT fuel-trim algorithm.
MD1: identify the sensor and the matching definition
MD1 variants differ by manufacturer, engine and software. Some use an air-mass sensor and other air-path measurements; do not assume that an EDC17 or MED17 map name, axis or address transfers to MD1. An example vehicle listing identifies the DS 1.5 BlueHDi MD1CS003 and lists a throttle air-mass-flow calculation, but does not provide an independently verified MAF transfer-map name, axis or scaling. A throttle airflow model is not automatically the MAF sensor characteristic.
| Function to identify | Axes and action |
| MAF sensor linearisation, if fitted | Find the exact ECU-specific characteristic in a trustworthy matching definition: raw sensor input (voltage, frequency, period or digital quantity) and mass-flow output. Verify units, breakpoints, encoding and active path. A named MD1 MAF map and its axes could not be confirmed from the available sources; AFSCD_dmLin_CUR is an EDC16/17 search name, not a verified MD1 identifier here. |
| Sensor diagnostic bounds and selected-air quality | Check raw sensor range, plausibility and whether a diagnostic value is measured, modelled or substituted. The actual names, axes and thresholds must come from that software. An unchanged, valid range should remain enabled; a flow curve edit cannot repair physical saturation or wiring faults. |
| Air-path targets, smoke/fuel limit and airflow models | Check which functions consume the corrected air mass and their coverage, using matching documentation and logs. Their names, inputs and units are not established by the vehicle listing. Do not alter these functions simply to conceal an inaccurate sensor or to force agreement with an air-mass target. |
Conditional arithmetic example: If an independently verified MD1 definition for the installed hardware identifies a direct flow-output point at a particular sensor signal as 500 kg/h, and a calibrated reference for that same signal and complete sensor/housing measures 525 kg/h, the candidate output is 500 × (525 / 500) = 525 kg/h. This is a hypothetical calculation, not a documented MD1 map or breakpoint. Use only after verifying the signal representation, linearisation, units, correction chain and interpolation for that software; gather multiple points and do not extrapolate beyond the reference. If those details cannot be verified, do not edit a guessed map. Requested versus actual diesel air mass is not an independent reference. Confirm measured versus modelled air, operating state, EGR response and diagnostics in validation logs.
MD1CS003: specific diagnostic evidence and definition gap
A ScanDoc MD1CS003 diagnostic menu for Peugeot 301 lists a “mass air flow sensor replacement” service function. This is evidence that the tool exposes a replacement procedure for that application; the listing does not explain the procedure, prove it changes the transfer curve or establish that every MD1CS003 vehicle uses it. Consult the current vehicle-specific service instructions before use and record the adaptation state before and after.
A separate ScanDoc Peugeot Traveller MD1CS003 example labels air-flow channels in kg/h and mg/stroke and a target in mg/stroke. Its snapshot contains implausible values in several unrelated channels, so its numerical readings are not calibration evidence. The labels are useful when selecting logger channels, but do not identify map axes, prove signal quality or establish the ECU's internal conversion.
For a specific Peugeot 308 1.5 BlueHDi 130, a seller listing identifies MD1CS003 hardware 9818703480 and software upgrade 9693382980. It does not publish a verified MAF transfer-curve symbol, axis or scaling. Matching hardware alone is insufficient: obtain the exact software and calibration identification and a compatible definition, then verify its map semantics and raw-to-physical conversion before calculating changes. No variant-specific MD1 MAF map name, axis or example values were independently verified in this research; do not transfer the EDC17 AFSCD_dmLin_CUR name or a guessed address to this file.
General validation workflow
- Record the original calibration, ECU identification, sensor part number and intake arrangement. Establish whether MAF, speed density or a blend is active in each test region. A speed-density estimate is another model and is useful for comparison only when independently validated.
- Correct faults first: unmetered-air and exhaust leaks, sensor supply and grounds, contamination or damage, fuel pressure, injector characterization, fuel composition and lambda sensor or logger errors can all distort the result. Check both banks separately where available.
- Use a documented transfer curve for the complete sensor and housing as the starting point. A bore-area ratio is only a rough starting estimate: placement, bends, screens, pulsation and flow distribution matter. Check signal headroom at idle and peak flow; a saturated sensor cannot reveal additional airflow.
- At stable, warm closed-loop petrol operating points, log raw voltage or frequency, calculated airflow, RPM, load, requested and measured lambda, short- and long-term corrections and control state. Group samples by signal, reject transients and allow for sensor and exhaust delay. Interpret correction signs and the way learned and immediate trims combine for that ECU; a trim is evidence of a mixture discrepancy, not proof that the MAF caused it.
- For regions where feedback is absent or cannot measure the commanded rich mixture, compare final requested lambda with a calibrated, suitably installed wideband while monitoring fuel pressure and engine safety. Feedback and learned corrections can mask an error; establish their actual state and follow the ECU-specific procedure. Do not infer high-load accuracy from idle trims.
- Make small, justified changes in well-sampled regions, preserve a smooth physically plausible curve, and re-log. Under a simplified open-loop petrol model with correct injector and pressure data and no active fuel correction, measured lambda divided by commanded lambda is an approximate multiplier for the reported airflow: for example 1.05 / 1.00 suggests about 5% more reported air. This is a diagnostic approximation, not a universal formula: offsets, filtering, blending and ECU correction conventions can change how the stored table must be edited. Investigate large or inconsistent errors before applying it.
- Restore any temporary test settings and retain diagnostic and engine-protection functions. Verify idle, part load, transitions and safely controlled higher load, both banks, fuel trims, final lambda, calculated load, plausibility and headroom. Recheck under representative temperatures and conditions. Do not fill unmeasured regions with invented certainty.
See ME7 MAF and lambda log review for a worked logging approach and EcuTek's BRZ intake rescaling guide for its specific MAF and speed-density comparison. Modifying emissions-related calibrations may be restricted by local rules; retain applicable monitoring and verify the requirements for the vehicle's use.
Revision 7: MAF calibration across ECU types
What MAF calibration does
The mass airflow (MAF) calibration translates the sensor signal into an estimate of air mass for the installed sensor and intake. Depending on the ECU, this estimate contributes to cylinder charge, load, fuel, ignition and torque calculations. It is not simply a fuel adjustment. A replacement sensor, housing, airbox or pipework can change the relationship between signal and flow. Identify the exact ECU software, sensor and housing before selecting a curve. See airflow metering for sensor types.
A transfer curve may express voltage or frequency against grams per second, kg/h or a scaled value. Confirm axes, units, offsets, limits and any later corrections in the actual definition. Some sensors report frequency rather than voltage, and a frequency signal is not necessarily a mass measurement: a Karman vortex meter measures volume and needs the relevant density calculation. A different sensor or housing cannot be assumed to share a curve. The Haltech MAF documentation describes voltage and frequency inputs and sensor curves.
Examples across ECU strategies
| System | Calibration approach and qualification |
| Bosch ME7 petrol | In the documented ME7 examples MLHFM converts the MAF signal; MLOFS and additional correction functions affect the result. The effective air mass also affects load. Verify the exact software and calculation order before scaling; see ME7 fueling and MAF calibration. |
| GM example using frequency | Some GM applications use airflow versus MAF frequency. EFILive's Calc.VE tutorial shows a frequency-indexed MAF calibration. Other airflow models may participate depending on the operating state and ECU; confirm the active strategy rather than assuming MAF is the sole input. |
| Subaru BRZ / Toyota 86 example | EcuTek documents a voltage-to-g/s scaling map for the specified BRZ/FRS/GT86 family. Its tuning guide discusses changes for the housing or intake. This is an application example, not a table definition for every Subaru or Toyota ECU. |
| Nissan 350Z VQ35 example | EcuTek describes a multi-stage conversion on the covered ROMs: voltage to a percentage, then a maximum-flow conversion and further airflow corrections. Check bank, ROM and the actual calculation rather than assuming a single direct voltage-to-g/s table. See EcuTek's VQ35 strategy guide. |
| Diesel air management | Air mass can be used for air-path and EGR monitoring and fuel-quantity limiting, depending on the ECU. Diesel operation is generally lean and petrol-style stoichiometric fuel-trim scaling is not a general diesel method. Compare specified and actual air mass under the manufacturer's test conditions and diagnose the air path before changing a curve. The Ross-Tech TDI procedure provides an application-specific logging example; see also Volkswagen TDI air-mass measurement. |
| Standalone ECU | Some support a configured voltage or frequency input and a loaded sensor curve. Haltech documents these choices and a MAF fuel-load mode; other standalone systems may instead use speed density or another load strategy. Follow the specific controller's documentation and use data for the installed sensor and housing. Haltech MAF setup. |
Worked examples: calculating and applying a correction
These figures are invented to illustrate the arithmetic; they are not recommended calibration values. First establish the fuel and sensor health, ECU strategy, exact table axes, units and correction sign as described below. A mixture error does not prove a MAF error. Retain the original file, make changes only in the evidenced region, and re-log the result.
1. Subaru BRZ / Toyota 86: voltage and closed-loop trims
The EcuTek guide describes this application's voltage-to-g/s table. Suppose a stable 2.00 V bin has a table value of 30.0 g/s and an established combined fuel correction of +6%: in this example positive means the ECU adds fuel. If diagnosis supports an underestimated MAF and the applicable correction acts proportionally, the starting estimate is:
factor = 1 + 6 / 100 = 1.06
new airflow at 2.00 V = 30.0 × 1.06 = 31.8 g/s
Apply 31.8 g/s to the corresponding voltage point, review neighbouring bins using their own adequate stable samples, and check interpolation and curve shape. Do not copy 6% across the entire curve. Log again at similar conditions and check the bank and learning state. This example assumes the displayed +6% is already the effective combined correction; do not blindly add STFT and LTFT. Their sign, representation, learning regions and combination depend on the ECU and logger. See EcuTek's BRZ MAF guide.
2. GM frequency-output example: measured versus commanded lambda
For a GM application with a frequency-to-airflow table, as shown in the EFILive Calc.VE tutorial, suppose the table gives 200 g/s at 8,000 Hz. During a stable point the final commanded lambda is 0.80 and a calibrated, time-aligned wideband reads 0.84. Assume the MAF is the active load input, injector and pressure data are correct, feedback and learned fuel corrections are accounted for, and the error is attributable to the MAF:
approximate factor = measured lambda / commanded lambda
= 0.84 / 0.80 = 1.05
candidate airflow = 200 × 1.05 = 210 g/s at 8,000 Hz
The mixture is 5% leaner in this simplified ratio; the candidate adds 5% to reported air and consequently calculated fuel under the stated model. Apply only after confirming the table and active airflow model for that operating system. Inspect neighbouring frequency bins and re-test with a wideband, fuel pressure and load monitoring. If lambda becomes dangerously lean, stop the test and diagnose; do not continue a loaded pull to populate a histogram. The same calculation can suggest a reduction: 0.76 / 0.80 = 0.95, or 5% less, under the same assumptions.
3. Bosch ME7: include the MLOFS offset
The ME7 guide describes a documented relationship in which effective air mass equals MLHFM minus MLOFS. Suppose a particular point has MLHFM 120 units, MLOFS 10 units, and a justified factor of 1.05. Use the actual definition's units and confirm that this relationship applies to the software:
old effective air mass = 120 − 10 = 110
new effective air mass = 110 × 1.05 = 115.5
new MLHFM value = 115.5 + 10 = 125.5
Write the calculated 125.5, rounded according to the editor's precision, into the applicable MLHFM point; retain MLOFS in this example. Directly multiplying 120 by 1.05 gives 126 and changes the offset's contribution as well. This arithmetic illustrates the offset, not a claim that every ME7 has a value of 10 or that every correction is applied at this stage. Verify calculation order, corrections and load effects, then re-log. The MxT ME7 guide documents one software-specific way to group and analyse trim data.
4. Nissan 350Z VQ35: a percentage stage
EcuTek's VQ35 guide describes voltage-to-percentage and percentage-to-g/s stages on covered ROMs, followed by further corrections. For arithmetic only, suppose a particular point reads 40% and a separately established 1.05 factor is appropriate at that stage. The candidate is 40% × 1.05 = 42%. This does not by itself establish the final g/s: trace the subsequent conversion, bank and corrections, and confirm the editing tool's representation and recommended adjustment location. Do not assume a percentage table is a direct airflow table or change a shared maximum-flow scalar to fix one local bin.
5. Standalone ECU: choosing the correction location
Haltech documents a configurable sensor curve and, in MAF tuning mode, a Fuel Base correction factor that starts at 1.000. Suppose a valid MAF curve point reads 50.0 g/s and an independent flow reference for the same installed sensor and housing establishes 52.5 g/s: 52.5 / 50.0 = 1.05, so the candidate sensor-curve point is 52.5 g/s. Alternatively, a base-fuel factor of 1.050 would apply a 5% correction in the documented fuel calculation, but the resulting effects and conditions are not necessarily equivalent throughout the ECU. Do not apply both 1.05 changes for the same error: 1.05 × 1.05 = 1.1025, or 10.25% combined. Use the appropriate correction location and verify the resulting fuel and load behaviour. See Haltech MAF setup and Fuel Base documentation.
6. Diesel: use a known flow reference
Suppose an independent, appropriate flow-bench reference measures 440 kg/h for the complete sensor and housing at a stable signal, while the ECU conversion at that signal is 400 kg/h. If the definition is a direct, proportional flow value at that signal, 440 / 400 = 1.10, giving a candidate of 440 kg/h. Repeat at multiple flow points and verify the sensor, units, setup and reproducibility before changing the curve. In a vehicle, a requested 440 and measured 400 kg/h are not an independent 1.10 calibration factor: EGR, boost, leaks, restrictions and control conditions can explain the difference. Follow the vehicle-specific test, such as the Ross-Tech TDI logging procedure, and retain emissions and diagnostic functions.
Which maps to find: ME7 and EDC16
The MAF transfer curve is the primary calibration for a changed sensor or housing. Other maps may need checking because they consume the resulting air mass, but they are not automatically changed. Exact names, units, axes, availability and addresses depend on the ECU variant, software and definition. Identify the software and confirm the function against its matching documentation before editing.
Bosch ME7: example map roles
| Map or parameter | Action and reason |
| MLHFM | Primary MAF voltage-to-air-mass characteristic in the documented ME7 implementations. For a verified sensor/housing change, revise the points for which a valid new transfer curve is available. Confirm the definition's axis, output units and interpolation. |
| MLOFS | Air-mass offset in the documented relationship. Check it when calculating new MLHFM values; it does not necessarily need changing. The worked ME7 example above shows how to scale the effective value while retaining the offset. |
| KFKHFM | Additional air-mass correction in the documented ME7 strategy. Inspect its axes, values and role so its contribution is not overlooked or applied twice. Change it only where the application-specific evidence supports that correction; do not automatically use it to hide injector or pressure errors. |
| KRKTE, TVUB and FKKVS | Fuel and injector calibration functions, not MAF transfer curves. Check the injector flow, voltage offset and fuel-system correction when diagnosing mixture error; modify them only for a separately established fuel-system change or error. |
For example, if the identified MLHFM point is 120, MLOFS is 10 in the same units and an independently justified effective-air factor is 1.05, enter (120 − 10) × 1.05 + 10 = 125.5 at that point, subject to the actual map's scaling and precision. Review adjacent points and downstream calculated load; re-log. This is an illustrative calculation, not a stock map value. The map roles and offset caveats are explained in ME7 fueling and MAF calibration; the MxT ME7 guide describes a logging-based MLHFM workflow for its supported applications.
Bosch EDC16: identify the specific sensor conversion
For an analogue EDC16 application, look in the matching definition for the MAF sensor characteristic / air-mass linearisation: a voltage input and an airflow output. This is the candidate curve to modify when a documented replacement sensor and housing have a different transfer function. The published Trifonov and Hadjiivanov sensor study experimentally reports an EDC16U1 diagnostic voltage-to-flow characteristic. It does not identify the calibration map's Bosch name, address or raw scaling, so those cannot be inferred from the diagnostic values. A separate VAG EDC16 map list calls one MAF linearisation curve AFSCD_dmLin_CUR; it does not identify the software or provide its axis scaling. Do not assume that all EDC16 files use this name, call the curve MLHFM, or share ME7's offset.
For example, that study reports 43.91 g/s at 3.0 V for its tested EDC16U1 setup. If an independently calibrated reference for a different, installed sensor and housing instead measures 46.11 g/s at the same signal, the illustrative ratio is 46.11 / 43.91 ≈ 1.0501. A direct g/s table point would therefore be set to approximately 46.11 g/s, provided the correct table and its scaling have been verified. If the definition uses kg/h, multiply by 3.6: 43.91 g/s is 158.08 kg/h and 46.11 g/s is 166.00 kg/h. These are a published baseline and a hypothetical replacement measurement, not a transferable curve. Build and verify the whole curve from multiple known points rather than extrapolating a single ratio.
Also inspect, but do not automatically alter, the MAF-based fuel-quantity or smoke-limiting maps, EGR air-mass targets and plausibility ranges in the exact file. They use or compare air mass; they do not convert the sensor signal into air mass. Correcting the transfer curve can change the values seen by those functions, so verify their behaviour in logs. Requested versus actual airflow is not an independent calibration reference. Some EDC16 variants use frequency-output sensors, so a voltage curve cannot be assumed for the whole family. If the definition does not reliably identify the conversion curve or its scaling, stop and obtain a matching definition or validated measurement rather than guessing an address or map name.
Named EDC16 examples and axes
These names are search aids for a matching definition, not universal addresses or verified raw conversion factors. A curve changes the reported sensor flow; air-path targets, substitute values and smoke limits have different jobs.
| Identifier and scope | Input and output | How to use it |
AFSCD_dmLin_CUR, VAG EDC16 list | AUTODTC's VAG EDC16 list labels this a 32×1 MAF linearisation curve. It does not give signal units, output scaling, exact ECU variant or software ID. | Search the matching definition for this identifier. Verify whether its input is voltage, frequency or period, and whether its output is kg/h, g/s or scaled raw counts. Change the verified flow values to match the replacement sensor and housing; do not assume 32 points or a particular scale in another file. |
| EDC16C39, Alfa 159 1.9 JTDm example | Aliant Auto's example describes the MAF conversion as input frequency versus air mass in kg/h and lists AFSCD_dmLin_CUR among useful IDs. It does not explicitly link the ID to the displayed curve or publish its numerical axes and scaling, so the association is a likely candidate, not a verified map definition. | The author suggests data from the donor 2.4 JTDm MAF and intake in this specific conversion. Confirm the exact sensor, housing and installation, sensor signal, units and both ECU definitions; compare the donor curve at the recipient's actual breakpoints and interpolate only if supported. Validate the result in service. |
AFSCD_mAirPerCylDfl_MAP, VAG list | The same list calls this an 8×8 substitute air-mass map used on error; its precise axes, units and software are not given. | This is a fallback value, not the healthy sensor's transfer curve. Do not change it to compensate for an inaccurate MAF. Check which signal and quality state the logger displays. |
FlMng_qAFSCDSmoke_MAP, VAG list | The list describes a 13×16 MAF-based smoke fuel limiter but does not supply exact axes, output units or software ID. Confirm these in the particular definition. | A consumer of air mass, not the sensor linearisation. Recheck its operating range and any active limiting after the airflow is corrected; do not scale it as a substitute for calibrating the MAF. |
AFSCD_dmAirSubsVal_MAP, BMW E90 335d EDC16CP35 | The documented case identifies a substitute-air lookup indexed by engine speed and injected quantity when required inputs fail. Its dimensions and numeric scaling are not established there. | Check sensor-quality and selected-signal channels when interpreting a MAF log. A plausible modelled or fallback value is not proof the meter itself is calibrated. Do not edit this fallback to fix the sensor curve. |
Applying a donor frequency curve: illustrative EDC16C39 method
Assume the exact recipient definition has been independently confirmed as signal frequency in Hz against kg/h and permits interpolation; these values are invented. At a recipient breakpoint of 4,000 Hz, its current curve says 300 kg/h. A verified donor curve for the complete replacement sensor and housing has 3,500 Hz → 310 kg/h and 4,500 Hz → 370 kg/h. Linear interpolation gives:
fraction = (4,000 − 3,500) / (4,500 − 3,500) = 0.5
candidate flow = 310 + 0.5 × (370 − 310) = 340 kg/h
change at this point = 340 / 300 − 1 = 13.33%
Enter 340 kg/h at the existing 4,000 Hz point, after converting to the actual editor's units and encoding. Repeat across the documented range, preserving the original axis unless the software definition and procedure explicitly support changing it. This example does not establish an EDC16C39 stock or donor value. Use the ECU's documented interpolation rule; do not extrapolate beyond the donor data. If the actual axis is period in microseconds, use that axis and its interpolation rule: frequency and period have an inverse relationship, so directly interpolating in Hz can give a different answer. Check for clipping and headroom, restore normal operation, then log raw signal, measured and selected airflow, quality state, RPM, boost, intake temperature and EGR operation. Validate against an independent reference where possible.
EDC17: MAF linearisation and related maps
The AUTODTC VAG EDC17 map list identifies the names below, but does not identify a specific software version or document the breakpoints, units, table dimensions or raw scaling. Treat the names as candidates to verify in the matching ECU definition. EDC17 variants and sensors do not all use the same signal type or map set.
| Candidate name | Role, axes and action |
AFSCD_dmLin_CUR | Listed as MAF linearisation. The list does not establish its input axis or output scale for a particular ECU. Confirm whether the actual input is voltage, frequency, period or another representation, and the output's units. For a verified change of sensor and housing, the corresponding flow points are the primary candidate for calibration. |
FlMng_qAFSCDSmoke_MAP | Listed as a MAF-based smoke limiter. The available list does not establish its axes, units or selection logic. It consumes air-mass information and limits fuel; it is not the transfer curve. Check its operation and coverage after a MAF change rather than applying the same percentage to it. |
AirCtl_mDesBase0_MAP, AirCtl_mDesBase1_MAP, AirCtl_mDesBase2_MAP | Listed as air-control/EGR maps for different gear groups. Exact axes and output units must be checked in the application. They are downstream targets, not measured sensor calibration. Check air-path behaviour with emissions controls functioning; do not alter them merely to make a MAF log agree. |
AFSCD_swtAirSysVal_C | Listed as an air-system configuration selector. Its precise values and logic are not established by that list. Confirm the active sensor and air-system configuration; do not change the switch to mask a sensor fault. |
Illustrative EDC17 calculation. Suppose a matching definition independently confirms a direct sensor-period axis in microseconds and airflow in kg/h. At 250 µs the recipient curve has 400 kg/h, and an independently verified flow measurement for the replacement sensor and housing at that exact period is 420 kg/h:
factor = 420 / 400 = 1.05
candidate AFSCD_dmLin_CUR output at 250 µs = 400 × 1.05 = 420 kg/h
The period, flow values and verified definition are hypothetical; they are not a claim that a particular EDC17 file uses those breakpoints. Apply the candidate only if that exact file's definition and sensor test support it; measure several points and retain the existing axis unless the definition and method justify changing it. 250 µs corresponds to 4,000 Hz, but interpolation in period is not equivalent to interpolation in frequency. Do not use requested/actual diesel air mass as an independent flow reference. Re-log sensor input, measured and selected air mass, relevant quality state, RPM, pressure, temperature and EGR response. See Bosch ASMod air system model for the distinction between measured and modelled air mass.
MED9.1 2.0 TFSI: period curve and air-mass correction
The MxT project's MED9 description identifies MSHFMTPH as an air-mass transfer characteristic against sensor period in microseconds, with flow in kg/h for the supported MED9 workflow. Its MED9 support is marked alpha and the project says that matching binaries were not supplied for its contributed MED9 definitions, so independently verify the particular ECU software, axes and values. A MED9.1 2.0 TFSI discussion also identifies this period-based signal path; it is community information, not a Bosch function sheet.
| Map or parameter | Axes, role and action |
MSHFMTPH | For the cited MED9 example, period in µs → air mass flow in kg/h. It is the transfer-curve candidate for a validated sensor or housing change. Confirm the actual file's axis direction, breakpoints, output scaling and interpolation before replacing any values; do not import ME7's MLHFM layout or MLOFS arithmetic by assumption. |
KFKHFM | Community discussions identify this as an HFM correction map. Verify both axis variables, dimensionless scaling and where it acts in the matching software; the available examples do not establish universal axes or calculation order. Inspect its contribution and change it only for a separately evidenced correction. A curve change and a correction-map change can compound; avoid applying the same error twice. |
KRKATE and fuel-pressure functions | The cited MxT documentation describes KRKATE as a direct-injection fuel constant with fuel-pressure correction, not a MAF transfer curve. Check injector data, high- and low-pressure supply and actual versus requested rail pressure before attributing a mixture error to air measurement. Change fuel calibration only for a separately established need. |
Illustrative MED9.1 correction. Suppose a confirmed MSHFMTPH point at 300 µs reads 360 kg/h. Under a simplified stable petrol operating condition with final commanded lambda 0.80, independently measured lambda 0.84, no masking feedback and verified injector and fuel pressure data, a provisional factor is 0.84 / 0.80 = 1.05. If diagnosis supports a MAF error and the verified transfer output is directly proportional, the candidate at that period is 360 × 1.05 = 378 kg/h. Retain KFKHFM in this example, review adjacent supported points and re-log lambda, raw period, airflow, load, fuel pressure and corrections. The values are invented, and the ratio is not proof of the fault or a universal formula; other corrections and software calculation order may affect it. Do not use idle trims to infer high-load accuracy.
MED17: identify the air strategy first
MED17 is a family, not one air-measurement strategy. Some applications use a MAF, while the MxT MED17 guide describes a pressure-based adaptive air model for its covered applications, including the EA855 EVO workflow. Do not treat its statements as proof that every MED17 lacks a MAF. Identify the exact ECU software, sensor hardware, logging channels and active strategy before choosing a calibration.
| Name or function | Axes, role and application |
| MAF transfer characteristic, where fitted | Identify the exact sensor-signal input and air-mass output in the matching definition. The sources consulted do not establish a single map name or axis for all MAF-equipped MED17 variants. Voltage, frequency or period and physical units must be verified in the specific software. A sensor/housing change calls for a validated transfer characteristic, not an assumed ME7 or MED9 map. |
rk_w, MxT MED17 workflow | MxT calls this a relative fuel-mass correction and bins logs by engine speed nmot_w and load rl_w. The guide uses a user-selected map definition and a multiplicative output. It does not prove that rk_w is the calibration's Bosch symbol in every file, nor that every MED17 has the same axes. This is fuel correction in the cited workflow, not MAF sensor linearisation. |
fupsrls_w / pbrint_w | MxT identifies these as signals associated with its adaptive pressure/load model. Treat them as logged model quantities, not automatically as editable calibration maps. Their availability and behaviour are software-dependent. |
| MAP sensor characteristic and injector calibration | In a pressure-based strategy, verify the matching MAP sensor conversion, its pressure reference and units, along with temperature, injector and fuel-pressure data. Exact map names and axes require a matching definition. An error in any can resemble an air-model or fuel-trim error. |
Illustrative fuel-correction calculation, not a MAF calibration: In the MxT-described workflow, suppose a verified speed/load cell has a relative fuel multiplier of 1.000 and its established combined correction is +4%, where positive means more fuel. The documented operation is 1.000 × (1 + 4/100) = 1.040. Apply only if the selected map is confirmed to act as that multiplier in the exact software and diagnosis supports it; re-log at the same operating point. MxT describes one method for combining its particular STFT and LTFT signals; other ECU/logger conventions may differ. Correcting fuel does not establish that calculated air mass, load or torque is accurate. Check pressure, leaks, injector data, fuel composition, fuel pressure and feedback state first. The numbers are invented. See the MxT fuel-trim algorithm.
MD1: identify the sensor and the matching definition
MD1 variants differ by manufacturer, engine and software. Some use an air-mass sensor and other air-path measurements; do not assume that an EDC17 or MED17 map name, axis or address transfers to MD1. An example vehicle listing identifies the DS 1.5 BlueHDi MD1CS003 and lists a throttle air-mass-flow calculation, but does not provide an independently verified MAF transfer-map name, axis or scaling. A throttle airflow model is not automatically the MAF sensor characteristic.
| Function to identify | Axes and action |
| MAF sensor linearisation, if fitted | Find the exact ECU-specific characteristic in a trustworthy matching definition: raw sensor input (voltage, frequency, period or digital quantity) and mass-flow output. Verify units, breakpoints, encoding and active path. A named MD1 MAF map and its axes could not be confirmed from the available sources; AFSCD_dmLin_CUR is an EDC16/17 search name, not a verified MD1 identifier here. |
| Sensor diagnostic bounds and selected-air quality | Check raw sensor range, plausibility and whether a diagnostic value is measured, modelled or substituted. The actual names, axes and thresholds must come from that software. An unchanged, valid range should remain enabled; a flow curve edit cannot repair physical saturation or wiring faults. |
| Air-path targets, smoke/fuel limit and airflow models | Check which functions consume the corrected air mass and their coverage, using matching documentation and logs. Their names, inputs and units are not established by the vehicle listing. Do not alter these functions simply to conceal an inaccurate sensor or to force agreement with an air-mass target. |
Conditional arithmetic example: If an independently verified MD1 definition for the installed hardware identifies a direct flow-output point at a particular sensor signal as 500 kg/h, and a calibrated reference for that same signal and complete sensor/housing measures 525 kg/h, the candidate output is 500 × (525 / 500) = 525 kg/h. This is a hypothetical calculation, not a documented MD1 map or breakpoint. Use only after verifying the signal representation, linearisation, units, correction chain and interpolation for that software; gather multiple points and do not extrapolate beyond the reference. If those details cannot be verified, do not edit a guessed map. Requested versus actual diesel air mass is not an independent reference. Confirm measured versus modelled air, operating state, EGR response and diagnostics in validation logs.
MD1CS003: specific diagnostic evidence and definition gap
A ScanDoc MD1CS003 diagnostic menu for Peugeot 301 lists a “mass air flow sensor replacement” service function. This is evidence that the tool exposes a replacement procedure for that application; the listing does not explain the procedure, prove it changes the transfer curve or establish that every MD1CS003 vehicle uses it. Consult the current vehicle-specific service instructions before use and record the adaptation state before and after.
A separate ScanDoc Peugeot Traveller MD1CS003 example labels air-flow channels in kg/h and mg/stroke and a target in mg/stroke. Its snapshot contains implausible values in several unrelated channels, so its numerical readings are not calibration evidence. The labels are useful when selecting logger channels, but do not identify map axes, prove signal quality or establish the ECU's internal conversion.
For a specific Peugeot 308 1.5 BlueHDi 130, a seller listing identifies MD1CS003 hardware 9818703480 and software upgrade 9693382980. It does not publish a verified MAF transfer-curve symbol, axis or scaling. Matching hardware alone is insufficient: obtain the exact software and calibration identification and a compatible definition, then verify its map semantics and raw-to-physical conversion before calculating changes. No variant-specific MD1 MAF map name, axis or example values were independently verified in this research; do not transfer the EDC17 AFSCD_dmLin_CUR name or a guessed address to this file.
General validation workflow
- Record the original calibration, ECU identification, sensor part number and intake arrangement. Establish whether MAF, speed density or a blend is active in each test region. A speed-density estimate is another model and is useful for comparison only when independently validated.
- Correct faults first: unmetered-air and exhaust leaks, sensor supply and grounds, contamination or damage, fuel pressure, injector characterization, fuel composition and lambda sensor or logger errors can all distort the result. Check both banks separately where available.
- Use a documented transfer curve for the complete sensor and housing as the starting point. A bore-area ratio is only a rough starting estimate: placement, bends, screens, pulsation and flow distribution matter. Check signal headroom at idle and peak flow; a saturated sensor cannot reveal additional airflow.
- At stable, warm closed-loop petrol operating points, log raw voltage or frequency, calculated airflow, RPM, load, requested and measured lambda, short- and long-term corrections and control state. Group samples by signal, reject transients and allow for sensor and exhaust delay. Interpret correction signs and the way learned and immediate trims combine for that ECU; a trim is evidence of a mixture discrepancy, not proof that the MAF caused it.
- For regions where feedback is absent or cannot measure the commanded rich mixture, compare final requested lambda with a calibrated, suitably installed wideband while monitoring fuel pressure and engine safety. Feedback and learned corrections can mask an error; establish their actual state and follow the ECU-specific procedure. Do not infer high-load accuracy from idle trims.
- Make small, justified changes in well-sampled regions, preserve a smooth physically plausible curve, and re-log. Under a simplified open-loop petrol model with correct injector and pressure data and no active fuel correction, measured lambda divided by commanded lambda is an approximate multiplier for the reported airflow: for example 1.05 / 1.00 suggests about 5% more reported air. This is a diagnostic approximation, not a universal formula: offsets, filtering, blending and ECU correction conventions can change how the stored table must be edited. Investigate large or inconsistent errors before applying it.
- Restore any temporary test settings and retain diagnostic and engine-protection functions. Verify idle, part load, transitions and safely controlled higher load, both banks, fuel trims, final lambda, calculated load, plausibility and headroom. Recheck under representative temperatures and conditions. Do not fill unmeasured regions with invented certainty.
See ME7 MAF and lambda log review for a worked logging approach and EcuTek's BRZ intake rescaling guide for its specific MAF and speed-density comparison. Modifying emissions-related calibrations may be restricted by local rules; retain applicable monitoring and verify the requirements for the vehicle's use.