Tuesday, May 10, 2011
"calibrated" MAFs and why they suck...
They work by having the output voltage reduced relative to the airflow, either by mucking about with the signal conditioning electronics in the sensor, to report that less air is flowing than really is--or by placing a stock sensor in a larger diameter housing so that sampled flow represents a smaller portion of the total flow.
Both of these procedures result in the MAF "lying" to the PCM about how much air is being ingested by the engine; telling it that less air than is really flowing is coming in.
This is how they allow you to run larger injectors without changing the MAF curve or injector parameters in the tune.
For example, let's say you install 42lb/h injectors in a 2003 GT that cam with 21lb/h units, and install a MAF "calibrated" for the 42lb/h injectors--no changes to the tune are needed as the calibrated MAF lies to the PCM and tells it that 50% less air is flowing.
The PCM says OK and calculates an injector pulse that would create the target AFR based upon 21lb/h injectors (which it believes are still installed). This pulse will be 50% shorter than would be needed with 21lb/h injectors at the real airflow, however since the real injectors are 42lb/h the target AFR is achieved with no change to the tune.
But there is a fly in this ointment...
The PCM also uses incoming air flow, and how much the engine could consume at the current rpm¹, to calculate engine load. The calculated load is then used in nearly every other engine control operation performed by the PCM, in particular ignition timing and fueling.
But as the incoming air flow is incorrect (the "calibrated" MAF is lying to the PCM), the calculated load is also incorrect which is not good. Let's look at just a single example of how this incorrect load can really mess things up.
Here is the Spark Borderline Table from the stock tune for my '03 GT (I have added the 10° base timing to the table values to make things more clear).
This table defines the maximum spark advance to be allowed at various load and rpm combinations. If the PCM calculated advance is higher than the table value the table value will be used, if not the PCM's calculated value is used.
Note that at 40% load, at 5000 rpm, up to 42° advance will be permitted; however that at 80% load only 25° is allowed.
Now imagine that you are really running at 80% load, however your "calibrated" MAF is telling the PCM that 50% less air than is really flowing is coming in. The PCM says "Cool, 40% load, let me pump that timing up to oh, let's say 38°"
Guess what happens to your engine at 80% load and 38° advance? It ain't pretty...
Many other engine control calculations are load based (AFR based on ECT and load being one).
The bottom line is that the only reason, other than to mess with your tuner's mind, to buy a "calibrated" MAF is if you are not planning on tuning your engine properly (or at all) using a modern tuning system. They are legacy kludges that should all be gathered up and tossed into the dustbin of automotive history.
The truth is that any MAF that meets the engine's air flow needs, can be used with any injectors, and any tune--IF you have the MAFs actual transfer function and a tuning system that let's you enter that map into the tune...
Here is an excerpt from the Pro-M website that reveal's the desireability of proper tuning [emphasis aded]:
"As an added bonus, All Pro-M meters are supplied with a transfer function sheet. This is the air mass vs. voltage data for your particular mass air flow meter. This is invaluable information for anyone who desires to have their PCM professionally tuned. It allows the tuner to have the EXACT air mass vs. voltage data for your meter, resulting in a perfectly accurate reading by your PCM. No other method of calibration will provide you with this important information.
Lack of this information will result in a "tune" that is, at best, an educated guess."
So you see, here is the proof that calibrated MAFs are only for those who do not wish to properly tuner their engine...
¹ - This explanation is close enough for this discussion, however it is actually a bit more complex than this.