Jeremy... Do you mean a blow through MAF in general or that specific placement of a blow through MAF?

I'll try to cover both bases just in case

There are many advantages to running a blow through MAF on a setup like this.
Lets start by looking at the problems with a standard suck through big MAF setup.

First of all, the MAF sensor is very sensitive to outside turbulence. With all the pre-turbo big MAF setups such as Perrin's, you have an open filter element placed directly on the MAF housing which is also usually placed in the engine bay. With an open filter so close to the sensor, the MAF readings get severely thrown off if/when stronger air currents make their way through the filter. For a severe example of this, you could try blowing a strong fan directly on filter while the car is idling. Chances are the car will die almost instantly or atleast stutter a good deal trying to keep idling. Now imagine what happens every time the radiator fans kick on while you're standing at a stop light. While you can tune around some of this, its pretty much impossible to maintain a solid idle and consistant MAF readings with a setup like that. The other major issue here is that the IAT which is incorporated into the MAF heats up very quickly from the hot engine bay air and causes the ECU to pull timing in response which results in inconsistant performance.

Some people have tried re-locating the MAF into the fender to try to alleviate these issues. While doing so does shield the MAF from the turbulance of the fans and the heat of the engine bay, brings us to another problem with this sort of setup. Think about the distance from the MAF sensor to the throttle body. In a stock configuration, intake air is metered by the MAF, passes through a relatively short inlet, then the turbo, the short TMIC piping, and the TMIC itself before getting to the throttle body. Now consider the path the air must take with a FMIC setup and a big MAF in the fender. The air enters the intake and is metered by the MAF in the fender, it then travels through the relatively long inlet to the turbo, next it goes through one long FMIC pipe, through a significantly bigger FMIC, and then through another long FMIC pipe, before it finally reaches the throttle body. Its easy to see that with this kind of setup you have significantly increased the dellay between when the air is metered and when it actually reaches the engine. This creates an inherent lag in the system and makes fueling much more difficult and inconsistant. Sadly there are no tuning parameters yet available to define/alter such a dellay so there is very little that can be done to tune around this problem. Of course there are also the obvious issues with putting a sensitive electronic part in a somwhat volatile area like the fender. The fender liners on our Subaru's aren't horribly high quality and have been known to get ripped out on occasion which would leave the intake and sensor exposed to the elements and potentially cause even more severe problems.

Now lets see how the blow through MAF alleviates these issues.
Since the MAF sensor is now in an enclosed environment, there is no problem with outside turbulance throwing off the readings. The idle will now be nice and stable and the MAF readings will be consistant no matter how much wind there is or if the radiator fans happen to kick on.

Since the MAF is now significantly closer to the throttle body, the inherent lag in the previous setup has also been eliminated. The ECU can now respond very quickly and accurately to air flow changes which results in very crisp and smooth throttle response that can't really be accomplished with a suck through type setup.

This brings us to another important advantage of the blow through setup. Remember that the ECU uses the IAT for boost and ignition trims. With a pre-turbo MAF, the IAT is a relatively irrelevant reading since its usually just measuring ambient temps at best and under hood temps at worst. The ECU has no idea what the temps are actually like post IC. If the IC is heatsoaked, the IAT does no necessarily change and the ECU cannot pre-emptively pull timing. It must instead rely on the knock sensor before it can react to the heatsoaked IC. With the blow through setup, the IAT is now located after the IC and directly before the throttle body. The readings it reports to the ECU are now much more relevant. If the IC is heatsoaked, the IAT reports a high temp to the ECU which can pull some timing in response to avoid knock before it ever happens and to keep the car running at its best no matter how hot the engine bay is.

Finally, since I now don't have to worry about MAF placement pre-turbo, we were free to run a complete 4" intake instead of necking down the inlet to l3" right before the turbo. I believe this has a lot to do with the amazing response this setup car has.

The only disatvantage I've found with the blow through setup is that the sensor reacts a little differently under boost. This requires a bit more attention when re-calibrating the MAF, but its really nothing major at all. The overall drivability, response, and consistancy gained with the blow through setup is well worth the bit of extra time spent to propperly calibrate it.


Now that we've covered that, lets talk about the specifc placement of the blow through MAF. My main criteria for placement was this:

it must be on the cold side IC piping

it should be downstream a relatively long straight section and atleast an inch upstream of any bend to minimize turbulence

it should also be upstream of atleast one bend before the throttle body, yet it should be close enough to the throttle body to provide good response

finally I did not want to put the sensor in the fender nor did I want to have to extend the MAF sensor wiring

The placement I chose seemed a logical choice and it has proven to work very well on this car. The only thing I may change is to extend the larger diameter piping and move the transitions further away from the MAF to allow the air to expand more completely before it reaches the sensor. This should allow for even more consistant readings and will also allow me to use the same size housing for higher flow rates.


I know I covered a lot here, but that was a really loaded question . Let me know if I missed something or if I can clarify anything about this as I did a good deal of research and contemplation to get this setup right.


Thanks