Pulled mine of a 1994 4.6 Thunderbird non SC.
I took some measurements to compare effective opening surface area of a 55mm w/ no sampling tube vs a 70mm w/ sampling tube.
70mm w/ out sampling tube: area = 3848 mm2
70mm sampling tube (calculated it is a rectangle 8x70mm): area = 560 mm2
70mm w/ sampling tube: unobstructed area = 3288 mm2
55mm area= 2375 mm2
Good job! I've been meaning to do that very test for a long time. Thanks for saving me the work.
I hope you guys understand that it's the transfer function of the MAF that really matters. And about the only way to know that transfer function is to either flowbench test a MAF, or to look in the tuning software for the vehicle that MAF belongs to.
There are some that are very close, such as the 89-93 and 94-95 Mustangs. Those are pretty close, and should work. Others are way off. I can tell you that just about every 4.6 MAF I've seen has a transfer function that hits 5v at around 44.5 lbs/minute. The stock 5.0L MAFs peg 5v at around 35 - 36 lbs/minute.
Also noteworthy, each lb per minute of airflow is likely to support about 10 HP. So if a MAF "pegs" 5v at 36 lb/min, it will likely support about 360 gross FWHP. However, there's another problem here. The stock MAFs on some calibrations are limited to 4.79 volts in the tune, which is 31.48 lb/min (stock Fox MAF). So without tuning, or a calibrated MAF, that's still the upper limit for airflow, so the HP limitation for a 55mm MAF is about 300 - 310 FWHP.
The MAF sensor should NEVER be swapped to a different housing, unless you are going to flow or dyno tune the MAF transfer function.
4.6 MAFs should never be used on a 5.0L motor unless you're dyno tuning the car, or building a custom chip that can accomodate those changes.
Now if you're going to the dyno and want a cheap reliable MAF upgrade, the 4.6 MAFs are dirt cheap, and can support over 400HP electronically. And their 80mm housing (despite the big bar in the middle of them) probably flows enough CFM to get the job done.
Aftermarket MAFs like C&L (which I use all the time in mail order chips) benefit airflow considerably because they don't have that divider bar in the middle. However, Ford used that bar to prevent backflow thru the MAF, so there may be some perceived benefit to keeping this divider.
You should also realize that the O2 sensors generally can "fix" any tune inaccuracy up to 25% richer or leaner than stoich, at idle and part throttle. 12% for Fox bodies. That's a pretty big margin, so chances are unless you are WAY off on the MAF, the car will "learn" how to drive and idle right eventually. BUT... and this is a big one... adaptive learning does not 100% apply to WOT. The narrowband O2 sensors don't work at richer air/fuel ratios, so they cannot monitor or correct for WOT.
At best, they learn at half throttle, middle RPMs, and apply those learned corrections to full throttle. That in itself is a bad idea that gets turned off when we tune. Moral of story: car can run and drive great, and still be lean as hell at WOT. Don't trust the car just because it runs good. The only way to truly know is with a wideband air/fuel ratio meter.