440HP+; An engineering approach
My approach is as an engineer, with quite some years modding my own cars, some common sense, and some learning by reading and of course trial and error. Also, sorry for my English; I´m not a native speaker.
AIR FILTER:
In a normal NA car, replacing the air filter by a high flow one should add very little extra power, if any (it depends on engine size, but it´s normally between 0-3 HP).
In a “family” turbo car, it could add something between 2-5 HP, and some 5-10 HP in a factory-performance turbo car. This is not precise, and is mostly based on hundreds of dyno results made by independent testers.
Anyway, a “performance air filter” is not just any cone shaped filter, and here is where I want to go deep. A cheap or wrong type of air filter could easily rob power from your engine.
Allowing more air to an engine by means of the air filter, basically depends on two parameters:
- Material´s air allowance (how much air your new filter allows per square inch of surface)
- The new surface size
In our case, we have a good quality CORRUGATED PAPER 6,5 L x 6.0 base x 2.5 ID OEM filter.
That makes a total breathable surface of 2*Phi*3.25*6 inches, equal to 122.5 Sq. In.
The K&N option, from which I have stepped away lately, has the exact size and shape (mantle surface), but a supposedly more breathable surface. They claim it to be some 10% more air yielding, but numerous field tests say they give you more sound than power.
I decided to go with a Vibrant Performance 10925, brand from which I have read impressive results. Only I was fooled by a wrong information on size on Summit Racing, and when it finally arrived, It was way smaller than OEM, so size would eat air allowance; I never even tried it. I devoted some time to search for a brand name filter, that would have good proven results as a brand, and that has a bigger mantle size (Also our airboxes are very restrictive in filter size). Finally, I found what seemed to be the very best option: MISHIMOTO MMAF-4006. It has not only the longest length allowed by our airbox, but also the biggest size in the market that fits there. As a plus (and not minor), it has a second air entrance on the farthest cap, giving even more breathable surface.
It´s total mantle (breathable surface) is: ((6.438+5.25)/2/2)*2*Phi*6 + (4/2/2)*2*Phi*4 = 110.15 + 25.15 = 135.3, which is some 10% more surface. JUST THIS is pure gain, supposing the surface is at least as breathable as the OEM corrugated paper. Obviously, one can think that a good quality cotton layers specialty surface will allow more air, so a good guess is that the filter alone will pass some 15-20% more air.
As if that are not good enough news, let me add that the OEM flange has a section of 2.5 inches (air outlet). This filter, has a 4” one, so the less restrictions, the more air. Please note that we are only adding air FEASIBILITY. Feasibility, because there are many instances in our engine where air/gasses are strangled, and unless we improve them all, we will not be taking advantage of all done so far.
So far for the filter.
FLANGE/TURBO HOSE:
In order to use this kind of filters, one has to replace the OEM 2.5” by a 4” one. There are very few options in the market, and they are more or less the same, but thinking in the next step (turbo air feeding hose), I bought the whole kit (MST performance´s MB-A4501). The only other equivalent option in the market is the Mishimoto kit, but it replaces the whole airbox by a rustic, not OEM looking one, and I can´t afford it (local MOT inspection bans modified vehicles).
This kit allowed me two key things:
1) A 4” flange to as to fit my filter
2) A high flow low restraints turbo feeding hose, which replaces the terribly bad engineered OEM hose.
NOTE: This kit INCLUDES a 4” inlet air filter; it´s just I just don´t trust this filter´s performance over a big brand one, and also my Mishimoto filter has a bigger surface. Also, I have NOT installed the provided airbox extension (I really don´t trust it will add performance).
AIRBOX/PCM SEAL:
If you take out the airbox lid, and then the air filter, you will see that all around the PCM there is a big (half an inch) gap that does not seal to the airbox. Our airbox is sucking hot air from the engine bay all the time. If we are taking all this hassle to get more and cooler air, you must seal this gap. I used a thick rubber foam strip. I can´t think of a counterproductive issue here, because with cooler air in the box, the PCM should be equally cooled or even better.
TURBO:
I have NOT modified the turbo itself. I just took off the heat shield (it prevents some heat to hit the hood, but it does not prevent the heat to spread in the engine bay, thus hotter engine components, and hotter air inlet hose). I wrapped it with a turbo blanket (those fiber/lava rock ones). It fits tight like a sock, the engine/hood is notably cooler, and now I can even touch the turbo after a run. This also contributes to keep exhaust gasses hotter within the turbo and when going out, thus helping a little with spin speed at low end.
DOWNPIPE:
After researching all products that I found, I decided to go with the ANTZ Performance (Australia). It provides by far the bigger pipes at the turbo outlet (3.5”), BUT IT ALSO IS THE ONLY ONE THAT KEEPS THAT SECTION UP UNTIL THE CAT CONVERTER, just where we need the bigger sections, because of the hotter air. I also covered the downpipe with fiber/lavarock tape right until the cat converter.
CAT CONVERTER:
I can´t simply delete the converter due to strict laws here, but I replaced it with a High Flow one, made also by ANTZ Performance. I continued with the mid-pipe (bigger section) provided by Antz. It really make the turbo spool sooner (low end torque is noticeably improved).
CUSTOM MAP:
All of the above will give you just a few ponnies, unless you put on your ECU a good performance map, made custom for these mods, so, after some research, I ended up buying a map from a Mercedes specialist in the US (VelosDesignWerks), which is installed using a special tool provided with the map (remember I live in Chile, so getting every piece of hardware here is a PITA).
FINAL RESULTS:
I have not dynoe´d my car, but theoretically, this solid stage 2 setup, should go over 440 HP and 560 NM.
Also, last week, I tried to test my setup against a 2021 BMW M3, and well, I could not win him in fact, but up until 100 MPH, he could not leave me behind as well, so you can speculate.
Hope this helps.


