Toyota Tuning Guide
Engine/Platform Overview:
The Toyota T24A-FTS, used in the Tacoma and 4Runner, is a direct- and port-injected, twin-scroll single turborcharged, twin variable-DOHC 2.4L 4-cylinder. Airflow measurements are done via MAF. The Denso ECU uses a conventional torque-based control strategy and, if equipped, offers automatic transmission control as well. The Tacoma is available with either a 6-speed manual or an 8-speed torque-converter automatic. The 4Runner is available only with an automatic.
Both Tacoma and 4Runner are available as hybrids, with a single electric motor located between the engine and transmission. The electric motor on the hybrids is a separate control module from the combined engine/transmission controller and is not flash supported at this time.
Boost control features a pneumatic actuator and a 2-port boost control solenoid. Charge air travels through an air-to-air intercooler, and is measured for charge air temperature and throttle inlet pressure before passing through the throttle body and into the intake manifold.
For further information on this engine, check out the link below. Note – this document highlights details for the transverse applications of the T24A-FTS (Highlander, Crown, RX) which differs slightly from the longitudinal applications in the Tacoma and 4Runner.
https://www.toyota-club.net/files/faq/21-09-20_faq_t24-engine_en.htm
4D Tables:
Before discussing tunable tables and controls, it is important to understand Toyota/Denso’s use of 4-dimensional tables. These tables are frequently used, so understanding how they work is key for adjusting the intended data points. Let’s start by reviewing 1D, 2D, and 3D tables:
1D: This type of table is just a single value. There are no lookup axes, meaning the output of this table lookup is constant. Ex:
2D: This type of table is commonly referred to as a “curve”. It features one lookup axis and calibration data that may change across that axis, allowing for the output of this table lookup to vary based on one input parameter. If the input axis does not exactly match one of these breakpoints, the ECU will blend the values of two adjacent cells using proportional weighting. In this example, this table controls the maximum amount of load based on current engine speed. Ex:
3D: This type of table is commonly referred to as a “map”. It features two lookup axes (X and Y), with the calibration data referred to as the “Z” data. This means that the table output can vary based on two input parameters. If the input axes do not exactly match one of the breakpoints for the X or Y breakpoints, the ECU will blend the values of the adjacent Z-data cells using proportional weighting. In this example, this table controls the requested torque based on driver pedal position and engine speed. Ex:
4D: This type of table is presented as several copies of a 3D table that lookup and blend using an additional axis – the W axis. The W axis acts as a third input parameter to adjust the table lookup output. Let’s consider the Ignition Efficiency 4D table:
Starting with Ignition Efficiency W[0], we see 3D table with input axes of Ignition Retard (X-axis) and Engine Load (Y-axis). The output of this lookup is Ignition Efficiency (Z-data).
Looking at the image of the table tree, we can see that 10 copies of this table exist – Ignition Efficiency W[0] -> Ignition Efficiency W[9]. Each 3D table is identical in size and uses the same lookup axes for X and Y but features different Z-data.
So how does the ECU know which Ignition Efficiency table to use at any given time? Reference the table Ignition Efficiency W Values. This table presents as a 2D curve and acts as a lookup axis for all 3D tables within this group. Each index value in the W Values table corresponds to – in this case – Engine Speed. So, if Engine Speed is 1200RPM the ECU uses the W[3] table; if Engine Speed is 4400RPM the ECU uses the W[7] table. And if Engine Speed falls between any of the W-axis breakpoints, the ECU will blend the output of the 3D lookup from adjacent W tables.
All 4D tables will be organized and named with the format shown above.
Torque Control:
Torque Request:
Base torque requests originate from one of three different tables – Base Torque Request (Normal), Base Torque Request (Eco), and Torque Request (WOT).
Base Torque Request (Normal/Eco) are 3D tables that use input axes of Accel Pedal Position (X-axis) and Engine RPM (Y-axis). The Eco table is used when the vehicle is in the Eco drive mode, and the Normal table appears to be used in all other drive modes.
Torque Request (WOT) is a 2D curve that takes over torque requests after APP eclipses a WOT threshold (not currently defined). This table will be active in all drive modes.
Tuning Tips: Increase torque requests as needed to increase engine output. Increasing torque requests in the Base (Normal/Eco) tables can help to increase throttle sensitivity, but first consider changing Axle Torque Limits (see following section). While Torque Request (WOT) is the source for torque requests at WOT, we recommend increasing torque requests in the base tables at the highest APP X-axis breakpoints to match the WOT table.
Axle Torque Limits:
The output of the torque request tables can be limited by the Axle Torque Limit tables. Limit tables are drive mode and gear specific. These tables consider torque multiplication based on gear ratio through the transmission and final drive ratio, which explains why the table values are so high in lower gears.
Tuning Tips: These tables have a significant impact on throttle sensitivity and powertrains responsiveness. Even with a stock Base Torque Request (Normal/Eco) table, these Axle Torque Limit tables will significantly limit torque at low- to mid-throttle. These tables can also limit maximum torque at WOT, so adjust as necessary to hit target output. We have a spreadsheet tool that will convert axle torque limit values into engine torque, per gear, to make this translation easier to visualize.
Torque Model:
Once Requested Torque is passed through the Axle Torque Limits, it is used as an input to the torque model. The torque model is the fundamental step where a torque request is translated into load, as well as the inverse – load back into torque. The combination of Torque to Load and Load to Torque allows for closed loop control of torque, should ignition and/or lambda efficiency be lower than ideal.
The most important element to understand is the translation of torque into load. The Torque to Load tables are 4D – the W axis appears to be a gasoline AFR, but this is pending further investigation. The W[0]-W[3] 3D tables use input axes of Requested Torque (X-axis), and Engine Speed (Y-axis). Based on the current amount of requested torque and engine speed, a certain amount of engine load will be requested. Essentially, this model states that – with ignition timing at MBT and lambda efficiency at 100% - this much load will produce this much torque.
Engine load is a representation of air mass consumed per cylinder, per engine cycle. Through a combination of volumetric efficiency and mass air flow measurements, the ECU can translate this target load into a target intake manifold pressure. This produces the boost target.
The purpose of the inverse torque model (Load to Torque) is to account for torque loss due to ignition and/or lambda inefficiency. When actual spark angle is less advanced than the known MBT angle – such as high load where the engine is knock limited based on fuel octane, or when charge air temperature gets very hot – the torque model uses the Ignition Efficiency tables to estimate how much less torque is being produced due to that spark angle. Final spark angle is compared to MBT angle at that load and RPM, and the difference is loaded into the Ignition Efficiency tables. This table outputs the percent torque loss and feeds it back into the inverse torque model so that additional load can be commanded to compensate.
Tuning Tips: There is NO NEED to modify the Torque to Load, Load to Torque, or Ignition Efficiency tables until torque requests eclipse the maximum breakpoint value for the Requested Torque X-axis. This will likely be more than what the stock turbo can produce.
Load Limiters
Once the torque model outputs a target load, that load value can be limited by several different limit tables. This occurs before the translation of target load into target boost. The load request from the torque request/torque model will be compared to all of the outputs of the following tables, and the ECU will select the lowest output to use as the final Load Desired. Only then will Load Desired be translated into a target boost pressure.
Load Limit Ambient Pressure: This is a 3D table that uses input axes of Engine Speed and Ambient Pressure (in ATM) to allow load limiting.
Tuning Tips: Note that load limits decrease as barometric pressure decreases. It is important to maintain this relationship as you increase the load limit to avoid turbo overspeed conditions. Reference OTS values – OTS maps were tested up to ~7000ft of elevation and approximately 0.75 ATM barometric pressure.
Load Limit Ignition Correction: This is a 3D table that uses input axes of Knock Retard and Engine Speed to allow load limiting. Reference the monitors ‘Load Limit Correction Axis’ and ‘Load Limit Ignition Correction Table Data’ to see the X-axis and table output, respectively.
Tuning Tips: Stock values at the 0* Knock Retard axis breakpoint may limit power, so increase that breakpoint as needed. We recommend otherwise leaving this table alone to avoid running too much load when there are significant knock corrections occurring.
Engine Load Limit 5: This is a simple 2D curve that limits load versus Engine Speed.
Tuning Tips: This is a quick and easy table to use to limit maximum load.
Load Limit 3D: This is a 3D table that uses currently unknown input axes to limit load.
Tuning Tips: Tune this pragmatically and conservatively as the function of the derate is not currently understood.
Load Limit Inferred Temperature CLT: This 3D load limit table uses input axes of Coolant Temperature and Engine Speed. The main purpose of this table appears to be limiting engine load when the engine is cool.
Tuning Tips: You may need to increase the values found at the highest Coolant Temp. (Y-axis) breakpoint, depending on power goals. Otherwise, this table can be left unmodified.
Load Limit Inferred Temperature IAT: This table is not used in the factory calibration and has the load limit pushed well out of the way.
Boost Control
Once torque is translated into target load, target load is translated into target boost pressure. This target boost pressure value can be limited by some boost limit tables before becoming a “final” boost target. Target Wastegate Duty Cycle is controlled using final target boost and actual manifold pressure.
Boost Target Limit 4D: This 4D set of tables allows for limiting of boost pressure, even if torque and load limits are high enough to command more boost. The W-axis is Coolant Temp., X-axis is Engine Speed, and Y-axis is Rail Exponent Pressure. The latter is currently pending further investigation. The Z-data represents boost pressure limits – this is gauge pressure relative to ambient, not absolute pressure.
Tuning Tips: These tables are generally far enough out of the way to not require modification. But if you choose to increase boost significantly, these tables may limit power. Raise as needed to hit power goals.
Wastegate Control:
Target Wastegate Duty Cycle is controlled using final target boost and actual air mass flow. The ECU also creates a modeled turbo speed that can help to inform decisions to increase boost further.
Wastegate Position Target 4D: These tables command feed-forward WGDC, before closed loop correction with P, I, and D components of boost error control. The W-axis represents Engine Speed, although the individual 3D tables that the W-axis blends do not change target WGDC. The X-axis is Air Mass Flow, and Y-axis is Target Boost.
Tuning Tips: Stock values work well with the stock intake and stock turbo around stock boost pressure. Increasing boost significantly may move the engine into a different part of the WGDC map that requires tuning. Watch Throttle Inlet Pressure Actual vs. Throttle Inlet Pressure Desired, and Throttle Angle – if TIP Actual is significantly above TIP Desired, the throttle will close. In these areas of operation, you would want to decrease WGDC to avoid the throttle closure. Similarly, if TIP Actual is significantly below TIP Desired, try increasing the WGDC values.
Cam Control:
The T24 uses variable timing on the intake and exhaust camshafts. We have defined the Exhaust Cam Timing Target and Intake Cam Timing Target tables. However, yet to be defined are the desired cam position overlap tables. Regardless of the calibration of the Exhaust and Intake cam timing tables, these values will be overridden by the cam overlap command, presumably to avoid valve-to-valve contact. This system will continue to be investigated to allow for greater tunability.
Fuel Control:
The T24 uses both direct and port injection systems to deliver fuel. The ECU will always command a stoichiometric air fuel ratio unless the component protection system is activated. Component protection seeks to use fuel enrichment to decrease the modeled temperatures of various exhaust components. Various temperature thresholds exist to activate different component protection modes – as temperatures get higher, enrichment gets greater. Once enrichment occurs, fueling enters open-loop and fuel trims deactivate.
Target Lambda: Four temperature enrichment thresholds exist to activate different component protection tables – AFR Enrichment Threshold 1, 2, 3 and 4. These correspond to activate AFR Target Component Protection 1, 2, 3, and 4 modes. Modes 1 and 2 are 4D tables that vary based on W-axis of Engine Speed, but X- and Y-axes are pending investigation. Modes 3 and 4 are 3D tables that use inputs of engine speed and load to target a specific lambda.
Tuning Tips: Use caution modifying these tables – lowering temperature thresholds could cause unintended and unnecessary enrichment.
MAF: The T24 uses a MAF sensor to measure airflow. If an aftermarket intake with a larger diameter MAF sensor housing is used, modify the MAF Scaling table and log the MAF Frequency monitor.
Tuning Tips: At low to mid-load, you will be able to reference fuel trims to correct the MAF curve. At WOT, the engine enters open-loop fueling and will report zero fuel trims. Reference AFR Actual and AFR Desired to correct the MAF curve at higher loads.
Ignition Timing
Final ignition advance is the sum of several base tables, compensations, knock learning, and knock corrections. Final timing is compared back to modeled MBT, the difference of which is used for ignition efficiency calculations within the torque model. These engines are very knock limited on standard pump gas, but yield significant power gains with additional timing advance. Careful calibration of timing is key for consistent, safe power gains.
Ignition Timing Base 1-4: These 3D tables control base timing advance prior to compensations/corrections. Timing advance figures shown are significantly lower than what the final timing advance will be, as compensations will add a significant amount of timing.
Tuning Tips: Tune these to a safe margin from the knock threshold. Stock values are adequate for 91 OCT, 93 OCT can pick up a bit more, but lower octane fuels like 87 OCT should have timing reduced pre-emptively. Higher ethanol contents should allow for significant timing advance increase.
Knock Correction Angle: This monitor reports timing corrections applied to all cylinders. The default value for Knock Correction Angle is -3*, which can be adjusted using the table Knock Control Base. Timing corrections can be both positive and negative – if the engine detects no knock, and there is room to add timing, this Knock Correction Angle can work upwards to -1*, effectively adding 2* of timing advance. Knock events will cause Knock Correction Angle to decrease. The rate at which timing is reduced due to knock is controlled by the table Knock Control Change Rate, which stock is -0.25*, meaning a quarter degree is taken out of timing at minimum per knock event.
Tuning Tips: Knock Correction Angle will generally only increase towards -1* from the base value of -3* when timing compensations – like Charge Air Temperature - work ignition advance downwards. When charge air temperature is well controlled and below ~130*F, this Knock Correction Angle will hold steady at -3*, unless knock events occur at which point it will work further negative. However, if charge air temperature gets above ~130*F, you will most likely see that value increase towards -1*.
Knock Learning:
Global timing adjustments can be made via the knock learning system as knock is detected repeatedly under varying load. The output of this system can be monitored via Knock Correction Learned. Keep an eye on this value as you develop your calibration and test on the street. After a reflash, it will revert to a default value. More tables and monitors related to this system are pending investigation.
Transmission Tuning:
Shift Schedules:
Upshift and downshift schedules determine when gear shifts occur based on throttle position and output shaft speed of the transmission. Each drive mode uses its own set of upshift and downshift schedules to tailor vehicle response to the intended purpose of that drive mode. There are roughly 14-15 different sets of drive mode shift schedules, some of which may not be active on a given trim. We have labelled Normal, Sport, and Tow drive modes in all ROMs – these three will be active on all trim levels.
We have attached a spreadsheet tool that allows for conversion of Output Shaft Speed into Engine Speed and Vehicle Speed to more easily understand when shifts will occur.
This spreadsheet tool converts upshift schedules and downshift schedules. On the third tab, it also shows the overlap between upshift and downshift thresholds:
Tuning Tips: For upshift schedules, if measured output shaft speed of the transmission exceeds the threshold shown in the shift schedule at a given APP, the upshift will occur. To increase the shift point, increase the OSS threshold.
For downshift schedules, if measured OSS exceeds the threshold shown in the shift schedule for a given APP, the downshift will occur. To make a downshift happen sooner, increase the OSS threshold.
Torque Converter Control:
Torque converter clutch lock and unlock schedules are defined per drive mode and gear, and allow for the calibration of when the converter is allowed to unlock and slip, or lock and hold. Like shift schedules, converter control uses OSS thresholds and APP to determine lockup behavior.
Tuning Tips: An unlocked converter can improve vehicle response and turbo response, but can also create an elastic driving experience and poor fuel economy. Be careful to calibrate lock and unlock tables together to avoid overlap which would produce a cycling-behavior between lock and unlock states.
Dyno Setup:
Dyno setup is very straightforward on the Tacoma/4Runners, but does have one ‘gotcha’ for doing a complete pull. The transmission’s Manual mode isn’t a true manual mode. Even if in Manual, if you go WOT before ~4000RPM, the transmission will still command a downshift.
This is because, even in Manual mode, the transmission is still using the shift schedules from the drive mode you’re currently in – Normal, Sport, Eco, etc. The manual shifting option will allow you to lock out automatic upshifts, but will not prevent automatic downshifts.
In order to do a pull through the full rev range, you will need to modify the downshift shift schedule for gear you’re planning to use and the drive mode you’re planning to use. For OTS development, we tested primarily in 4th gear and Normal drive mode. For dyno testing, we modified the Normal drive mode 4->3 downshift thresholds to be very low values – this avoided the undesired automatic downshift. Example values shown below compared to stock:
Stock:
Dyno Test Values:
Charge Air Temperature/Fan Setup:
The intercooler is located relatively deep behind the grill in the main cooling stack. On the street with real-world airflow, the stock intercooler does a pretty good job - definitely room for improvement, but not horrible even with more boost thrown at it. However, even with a big ~32,000cfm 40mph fan on our dyno, we easily saw 170F charge air temperatures through a 14 second 4th gear pull. This is much worse than what you’ll see on the street and can have negative impacts on likelihood of knock. Having a good fan setup is key on dyno, and so is validating the tune on the street.
Copyright 2025 © COBB Tuning Products LLC. All Rights Reserved. | www.cobbtuning.com