MAF or MAF less?

What is the best option for tuning a GM vehicle?

This is a common and somewhat misunderstood question regarding what will offer the customer the best overall performance.  The most appropriate answer depends on the use of the vehicle and the design of the intake.

For early LS1 vehicles, the P01 ECU had a small and less powerful processor, that was fine as it didn’t have a lot to do, like your laptop or home PC the speed of the ecu required depends on the number and complexity of tasks it has perform.  In addition to this the MAF used was relatively small in internal diameter and offered a restriction with only modest modifications, the removal of the MAF did not inhibit the performance of the ECU and did reduce the amount of restriction in the intake offered by the MAF sensor itself.

The standard early P01 ecu will run without the MAF without further changes but does loose some of its functionality, with the MAF removed the ECU logs a fault and defaults to the low octane timing table reducing the ability to modify the spark timing based on fuel quality and knock, without tuning this would result in a loss of performance.  Fortunately, the aftermarket saw to this with modified operating systems that restore the full functionality of adaptive timing system.

The later ECUs become a different story, the calibration data alone on an E38 or E67 ecu is 10 times larger than the early P01.  This means more and larger tuning Maps. The number of tasks has greatly increased with CAN messaging to the various vehicle control modules, this provides additional stress on the processor.   CAN messages are prioritised by the ECU and those deemed less important can drop out when the processor is stressed.  In addition, tighter tuning control for emissions and performance is added, the rpm pickup has twice the number of teeth and the camshaft sensor 4 times the number so the ECU is aware precisely the location of the rotating assembly at any time, this allows for more precise spark timing, misfire detection, knock detection and injection timing. 

All early P01 through to current GM Delphi ECUs operate on a similar principle where to relieve the processor from demanding Speed density calculation at high RPM the ECU switches the airflow it uses to solely the MAF, up to this point it uses a fancy blend of MAF and Speed density based on utilising the strengths of both methods.  The MAF requires no calculation as it is a sensor that measures Air flow, speed density is a calculation requiring manifold pressure, inlet air temperature and engine speed.  Within the P01, P59 and E40 ECU is a speed density efficiency Table, for the next generation a series of maps with data that form a speed density equation exist, the most current ECUs now use a Neural Network which still forms a more complex speed density output.  This calculation takes processor time, GM switch off this calculation at high speed, typically above 4000 rpm.  This frees up the processor for other tasks, importantly timing and knock control.

At lower engine speeds the ECU calculates the airflow and measure the airflow, it does this and determines the dynamic airflow, this is the most accurate airflow value as it uses the strengths of both methods.  The MAF is accurate at high flow but less at low flow, it also sits out the front of the intake so suffers from a small lag time and therefore cannot measure transient(changing) airflow as quickly. The MAF also needs a straight section of flow to be accurate or the air can buffet around the sensor element causing faulty readings. The speed density is calculated very quickly at low air speeds, it is therefore good for transient airflow changes and at low flow around idle and when large camshafts can disrupt the airflow due to reversion.

What this all means is that the choice to retrain the MAF depends on the application.  The MAF card used in LSA and later vehicles has the benefit of being able to be scaled to larger intakes, in most NA street car applications the stock MAF offers no restriction and causes no loss of performance.  Retaining the MAF with a head and cam package or boosted engines up to 800 HP is usually the best choice as the MAF pipe can be enlarged so no restriction is present and the ECU can utilise both the MAF and Speed density to produce the best dynamic airflow.  If the vehicle is correctly tuned for the two airflows, then it will be able to idle and drive at low speed better and will have more time for optimal timing processing above 4000rpm. In positive displacement superchargers that have very short intakes that are very large and very sensitive to restriction often a MAF cannot be used, not because of the potential of restriction but due to the inability to accurately measure the turbulent air in the short intake. 

In conclusion for early LS1 P01 ecu vehicles, the MAF is restrictive and offers little advantage. For later vehicles with E38, E67 and newer ECU’s retaining the MAF when correctly tuned results in better drive ability, faster ECU timing processing at high engine speeds for increased performance and lower stable idle speeds.  For all but the most high-power vehicles where the fitment of a MAF is not possible it is best to be run.