Ford Engine Tuning and Examples

The Ford Mustang is arguably the most popular vehicle for the do-it-yourself tuner. Ford refers to its PCM as an EEC, or electronic engine control. Starting with the 1988 California specification, and 1989 50-state versions, the Mustang has been equipped with a mass air, sequential EFI system. Other truck and passenger car applications soon followed suit. Various iterations of the EEC have been released with increasing clock speed and capabilities. The EEC-IV systems used on the 1989–1995 OBD-I vehicles were extremely well received by the aftermarket community for their ease of programming and relatively simple control strategy. The mass air based system allowed a large amount of flexibility and ability to adapt at elevated power levels. With a scaled MAF and larger injectors, it was not uncommon to see unmodified EECs supporting over 600 hp. Although drivability was not ideal, the engine operation was acceptable to the performance enthusiast who valued quarter mile ETs over street manners. The advent of custom tuning software for these EECs allowed experienced calibrators the opportunity to deliver tremendous horsepower and excellent street manners with the stock EEC hardware. Programmable “chips” were designed as modules that could be plugged into the J3 service port opposite the wiring harness on the EEC to hold new operational code for the EEC. When the EEC is booted, the J3 port is examined for data. If an after-market module with a valid program is plugged in, the EEC reads the file on it and operates entirely based upon data stored on the new module. If no module is present, the EEC reverts to the program hard coded onto the stock board from the factory.

This tech tip is from the full book, ENGINE MANAGEMENT: ADVANCED TUNING.

For a comprehensive guide on this entire subject you can visit this link: LEARN MORE ABOUT THIS BOOK HERE

SHARE THIS ARTICLE: Please feel free to share this article on Facebook, in Forums, or with any Clubs you participate in. You can copy and paste this link to share: https://musclecardiy.com/performance/ford-engine-tuning-and-examples/

The 1996 Mustang brought in the era of modular engines and OBD-II controls with the EEC-V. Software companies responded with programming tools that offered the same flexibility on the newer EEC-V with the benefit of the finer control strategy from Ford available to the after-market tuner as well. The EEC-V family retained the popular J3 port available for aftermarket chip interface, but could now also be flash programmed via the OBD port under the dash. The 2002 EEC-V added another twist with an internal limit to the maximum calculated MAF rate that was reduced to a little over 1,700 kg/hr. Since many supercharged applications can exceed this, scaling of the MAF/injector size/engine displacement became necessary at high power levels on the later versions.

In 2005, Mustang moved to a new family of engine controller known internally as the “Oak” family, with variants spreading across the entire product line. This new processor remains mass air based, but now includes torque-based ETC control. Modifications are still possible, but the torque-based ETC strategy poses a significant challenge to inexperienced calibrators who wish to drastically increase engine power. The trusty J3 port on this processor is gone as well. The only way to change calibration data is by flash programming. To further complicate things, the new CAN (controller area network) protocol is used for communication rather than the older KWP (key word protocol) as seen in earlier OBD-II EECs. This limits access to scan tools and programmers that have been updated with the newer CAN communication strategy.

This screen shot from the SCT Advantage software shows the MAF transfer function of a 2001 Mustang Cobra. The EEC actually processes in A/D (analog to digital) counts, but SCT has done the math to show the calibrator actual voltage in the table as well for comparison.

The SCT tuning package puts everything at your fingertips to modify and datalog almost any modern Ford vehicle.(Nate Tovey)

Ford EECs convert most voltage inputs from sensors (MAF, TPS, etc.) into an A/D count before actual processing. Some software packages normalize this to only show voltage values in the editor; others leave it as counts. The actual scaling depends upon the clock speed of the processor being used, but it always works out such that 0 to 5 v reference signals become 0 to 1,023 A/D counts. Clock speed of the processor determines the sampling rate, so time-based sensor inputs such as MAF must be corrected before processing. Many actual raw MAF transfer functions are listed “pound mass per clock tick” rather than “pounds per hour,” so this can lead to confusion when attempting to copy a MAF transfer function from one model into another if the normalization is not correct.

The Ford EEC-IV/V uses a Hitachi manufactured MAF sensor with a heated wire element. This design provides a temperature-compensated mass flow measurement directly to the processor. The post in the center of the MAF helps to reduce the effect of standing waves in the inlet tract on actual MAF measurement. Actual output of this MAF sensor ranges from 0 to 5 v, with a 0 lb/hr= 0 v intercept. The EEC only recognizes a maximum input value of 5 v, even though most sensors continue to increase output voltage with respect to flow all the way up to battery voltage. To prevent “pegging” the EEC’s MAF input, the range of the sensor being used should be selected to match the intended maximum engine airflow rate. Many stock calibrations also limit the maximum recognized MAF input to about 4.7 v to account for build tolerances and voltage drift. Since the slope of the MAF transfer function is so steep in this range, the addition of ~0.2 v worth of range (4.9 v max) can usually safely allow for more measurement capacity with the same hardware.

The later 2002 and up EEC-V controllers employ another unique variable Ford calls “max air charge multiplier” that must be changed when adding a supercharger. This variable limits the effective maximum load calculations. Most naturally aspirated applications have this set to 0.9. Changing this to 1.9 (the upper limit in the software) allows the EEC to properly compute loads well beyond 100%. This is also sometimes accompanied by the variable “anticipated air charge multiplier” which should be set slightly lower at approximately 1.8.

A cold air kit installed in front of the factory MAF alters the way air flows into the sensor. The new MAF output versus actual airflow must be entered into the EEC for proper fuel, load, and spark calculations. (Nate Tovey)

Over the years, many of the OE parameters have become accessible to the public. Companies like Superchips Custom Tuning and Diablosport offer excellent calibration tools that display all the essential calibration parameters in a friendly Windows-based environment. These software and hardware packages supply everything needed to start from the OEM calibration and change any scalars, functions, or tables needed to accommodate the performance enthusiast’s needs. The appropriate starting file is selected by noting the “catch code” marked on both the EEC and the inside of the passenger doorjamb. Once this file is opened up in the editor, changes can be made and saved as a new tuning file. This new tuning file can then be either flashed to a chip to be plugged into the external J3 port of the EEC (the opening with an exposed edge connector opposite of the wiring harness) or stored in a handheld programmer for direct EEC reflashing via the OBD-II port under the dash. Let’s look at a couple of specific examples:

Example 1:

A new MAF sensor is installed along with a cold air kit on a 1999 Mustang GT automatic. “CHH2” base file is loaded and saved with a new tuning file name to avoid overwriting the original. Under the “functions” group, there is a listing for MAF transfer function. If the new MAF has a known output, this can be copied into the transfer function as a starting point. SCT offers a handy list of almost every MAF available for Mustangs that can be pasted into the editor in seconds. At this point, the tune should be relatively close, at least enough to start and drive the vehicle. As long as the error between actual mass flow and MAF sensor output is within about 10%, the closed loop learning of the EEC fills in the gaps enough for most conditions. However, the proper recalibration should include modeling of the airflow to accommodate the new actual transfer function of the MAF sensor and cold air kit.

One of the most popular upgrades for the 5.0L Mustang crowd is the Vortech supercharger. When recalibrated with the proper MAF and fuel system, normal driving can be as good as stock with almost double the power.

Since the inlet plumbing upstream of the MAF skews its output, the proper method is to set target λ = 1, force open loop, and map each MAF voltage (or A/D count) break point for actual airflow under part load. Repeat for WOT at a richer lambda to build the upper portion of the MAF transfer function and return to closed loop operation. This ensures proper load calculation and fuel delivery under all conditions. Once this is known, a new WOT lambda can be set in the base fuel table (at 70 to 90% load) along with a new borderline spark advance value at WOT. Ford calibrates the Mustang GT for 87-octane fuel, so there should be room for a couple more degrees of advance with 91-octane or higher. The result should be an increase of about 15 hp over stock, depending on how aggressive one gets with the spark advance tuning.

This tech tip is from the full book, ENGINE MANAGEMENT: ADVANCED TUNING.

For a comprehensive guide on this entire subject you can visit this link: LEARN MORE ABOUT THIS BOOK HERE

Example 2:

A 9 psi maximum centrifugal supercharger, MAF, and injectors are added to a 1992 Mustang GT manual. “A9L” base file is loaded and saved with a new tuning file name to avoid overwriting the original. Again, starting with the OEM file for this vehicle saves a lot of time sorting out much of the background functions. We focus on the parameters that have been directly affected by the addition of the supercharger and fuel system.

First the injector constants must be adjusted to accommodate the new hardware. If this example uses Ford/Bosch 42 lbs/hr (green top) injectors, the appropriate values can be copied from their original application (1999 F150 Lightning). This means a low slope of 50.0, a high slope of 42.3, break point of 2.5e-5, and the corresponding voltage compensation curve (as seen in Figure A-1). These changes can be easily made by either loading the SCT value file for “42-lb injectors” or selecting “Bosch 42 lbs/hr” injectors in the Diablosport tuning wizard. After a quick check to the scalars tabs to confirm the new values are in place, we can proceed. If a different tuning software package is used, these values must all be entered manually. Remember, properly modeling the fuel delivery is absolutely critical to the accuracy of airflow modeling to be done later.

Figure A-1 The voltage compensation curve of the 42 lbs/hr injector is inserted to replace the original values. The SCT software highlights values higher than the stock reference in red, and lower values in blue.

A typical WOT timing curve for a centrifugally supercharged 5.0L engine is shown using the SCT Advantage software.

Idle speed should be increased slightly to prevent possible stalling. The factory setting of 672 rpm may not be able to support the drag of the supercharger. Changing the target speed to about 750 rpm goes a long way toward providing a stable idle. Since the car is likely running on 93-octane fuel, the ignition lead can also be increased at light load to offset the slower burn rate. The values in the base ignition table can be increased for all loads below about 80% by about 3 to 4 degrees, or this particular EEC allows us to simply enter 4 degrees as a scalar under “closed throttle spark adder” and “part throttle spark adder” for quicker results.

The base timing table only maps loads up to about 90%, so this table can be left essentially unchanged for much of its range. The top row of this table should have its high RPM values reduced slightly to smooth the transition into boost. The 1989–’93 EEC provides a specific table for WOT spark advance, so this is reshaped to reduce spark advance in areas where the cylinder pressure is expected to be higher. On later EECs, the base table is rescaled to include loads in excess of 100% and the appropriate cells are changed to optimize timing under boost.

With the centrifugal super-charger, WOT ignition timing below 2,500 rpm can be left unaltered. As boost is expected to build near 3,500 rpm, timing is reduced from about 22 to 20 degrees to allow for the slightly faster burn rate as density increases. Near 6,000 rpm, manifold pressure is highest along with intake temperatures due to compression from the supercharger. Timing is changed from the factory setting of 26 degrees down to about 19 degrees. Later dynamometer testing reveals exactly how much ignition advance can be added back in before detonation, but this should be done only after delivered lambda matches the desired safe target ratio.

Next, a good starting point for the MAF transfer function must be chosen. If using a “scaled” MAF for this application such as a Pro-M 80 mm piece that is “calibrated for 42 lbs/hr injectors,” a simple multiplier can be applied to the stock transfer function. This multiplier should be the difference in injector sizes. In this case, a multiplier of 42.3/19.6 » 2.16 is applied to the Y-axis (actual mass flow) of the transfer function as a starting point for later calibration. The SCT software has a large number of MAF sensor transfer functions available in the value file library to make this step quicker and slightly more accurate. Regardless, the actual transfer function as installed in this specific vehicle must still be mapped on the dynamometer.

Now open loop operation is forced by setting the closed loop enable temperature to the maximum allowed and the target fuel maps must be reset to ease mapping of the new MAF. The target lambda for loads under 90% is set to 1.0 (14.68 air/fuel), and the vehicle is driven on the dynamometer to measure the error in the MAF transfer function. Lambda values are used to correct the MAF curve in the software until errors are less than about 3% for all light load conditions. Target lambda is then set to l» 0.78 (11.4 air/fuel) for loads above 70% and sweeps are made to map the top end of the MAF transfer function. Again, corrections are applied to the MAF curve in the software until error is less than 3%. Once delivered lambda is equal to target lambda, timing can be advanced slightly under load to find best power or the detonation threshold. Since these EECs lack any form of knock control, it is advisable to stay about 3 degrees below the knock limit for any street driven car to prevent future damage.

After this mapping has been done, the base fuel table can be restored to a smooth transition from λ=1 at idle and cruise to the desired ratio under high load. Closed loop operation can be reinstated by setting the enable temperature back to stock. If the vehicle has a slight hesitation on tip in, the acceleration enrichment can be increased to offset the temporary lean condition on transition. This is done by changing the “AE multiplier vs throttle position” curve. This compensates for the mechanical lag in the MAF measurement caused by the longer path between the meter and throttle body. It also compensates for the increased instantaneous flow that comes with positive pressure built up against the outside of the throttle blade from the compressor.

At this point, the vehicle is ready to be test driven to check for any other issues.

Written by Greg Banish and posted with permission of CarTech Books