1.9 “Fogged” Airbox Modification

Pros: Increased Performance, Proven Results, Retains Stock Airfilter
Cons: The stock airbox is expensive to replace, so don’t mess up
Cost: Roughly $20, and 3 to 4 hours of your time

In the year and a half since I first posted the DIY instructions to modify the M44 Z3s airbox I’ve found that I get the same sorts of questions. I figured a document that tied all the info together would help me and those also interested in the modification. This primarily deals with the Fogged Airbox mod for the M44 but also talks about how the fuel injection system on the Z3s works as well as alternate intake products that are available and my thoughts on them. The instructions are slightly updated from my original plans to make assembly easier and to add an optional safety step.

Engine and Fuel Injection:

The fuel injection in the Z3s is a mass air fuel metering system. What this means is your car directly measures the weight of the air being draw into your engine and injects an appropriate weight of fuel (in the ratio of 14.7 parts air to each part of fuel) which it then ignites through your ignition system. Under cruising conditions and light acceleration your cars computer (DME) uses feedback from the oxygen (O2) sensors to fine tune the mixture.

The DME measures the weight of the air using a sensor called the Hot Film Meter. (HFM) The HFM is basically a film through which a current is passed to heat it. If you were to look at the inside of your HFM you will see that the film is very small and it actually only samples a small fraction of the air that passes through the meter. By measuring the temperature change of the film (and knowing the intake air temperature) the DME is able to calculate the mass (weight) of air that is flowing over the film. It can then calculate the total mass of air flowing through the HFM. This works if the airflow through the meter is smooth and evenly distributed, in other words laminar flow. The DME then uses this to determine how much fuel needs to be injected to keep the AFR at 14.7/1, also called Lambda=1. An AFR of 14.7/1 results in the most complete combustion of your fuel which gives you the least emissions. If an engine is running rich OR lean (more or less fuel to air) the engines emissions will increase. At full throttle the DME runs the engine richer for more power.

Now a key point… if you increase the airflow into your engine, the DME will measure it using the HFM (assuming smooth airflow) and will increase the fuel injected to compensate. The DME must do this or your engine would run leaner which would increase emissions. If you increase the amount of air, to keep the ratio between air and fuel at 14.7/1, the DME must add more fuel. If you increase airflow and burn more fuel you have increased your power output.


As delivered from BMW the M44 engine in the Z3s really is not as rev happy as say the engine in a Miata. Above 5000 rpm the power feels like it has reached a plateau and there isn’t much more to be gained by revving the engine higher. The reason for this is because up high your engine’s ability to breathe is being restricted. I wanted to see if I could correct this and verify an improvement.

After looking into the intake systems that were on the market or known to be ‘in the works,’ I determined that there was nothing out there that I would put on my car. The following are a few of the setups I found and some of my reasons for rejecting them:

Drop in K&N filter ($40): in my opinion, K&N filters do not filter as well as your stock paper filter. I will not put one on my car. Your filter is the first and only line of defense against letting dirt into your engine. The K&N is made from an oiled cotton gauze, if the oil dries out the ‘filter’ basically stops all filtering. I have heard it said that when a K&N filter gets wet the water can ‘wash’ the previously filtered dirt right through it and into your engine. The person who told me this raced motorcycles and said that when he switched from K&Ns back to paper filters his engines lasted much longer between rebuilds. Lastly, Greg Hudson had performed dyno tests comparing a new stock paper filter against a new drop in K&N filter… the result was a LOSS of power with the K&N. That demonstrates that the paper filter isn’t what is limiting airflow as the paper filter has a huge amount of surface area for the size of our engine. Swapping filters does nothing to address the real source of restriction in your intake.

K&N Filtercharger cone filter ($149): Has a lot of problems. It has the same filtering concerns as the drop in filter and the rest of the setup adds a few more problems. The cone filter can suck in hot radiator air which will hurt performance. This is because cooler air is denser air (heavier), the more air weight you can get into your engine the more fuel will be injected and the more power you will get out of your engine. The circular shield that some sell might help a little but it will also increase restriction as it blocks the filter somewhat. It still doesn’t get all that great of a source of cold air either. As with most other cone setups the tube that connects the filter to the HFM changes diameter at the cone filter and at the HFM. Those ‘adapters’ are not a good idea. When they increase or decrease in size it will cause turbulence and restrictions in the airflow. For the HFM to meter the air weight properly it needs a smooth flow of air through it. The connecting tube changes size right at the start of the HFM, this could cause turbulence which would result in improper metering by the HFM. If the HFM isn’t able to properly measure the airflow into the engine the DME is going to be injecting the wrong amount of fuel. This will keep the engine from running optimally and won’t be making the power it should. Additionally, the filter is right in front of the HFM and the airflow doesn’t have much space to ‘settle down’ so it will still be turbulent from passing through the filter itself when it hits the HFM. There will be more on this later.

Art of ROAR cone filter (~$200): Mostly the same problems as the K&N setup above.

ECIS (~$200, when available): Uses a K&N cone filter so it has all the same filtering concerns and most of the other problems except it has a shield behind/besides the filter to block off the filter from the rest of the engine compartment to help with the heat problems.

Dinan Cold air intake ($299): uses a K&N cone filter so it has all the same filtering concerns. It positions the filter behind your foglight to avoid the hot engine air problem but exposes it to more water. Has the same type of ‘adapters’ and the problems they can cause that all the above cone filter setups share.

What I was looking for was a setup that kept the paper filter (they filter well.. even when wet) that had a smooth airflow to the HFM for proper metering, and had a good source of cold air but that allowed the engine to breathe better.

Factory airbox setup:

On the stock M44 airbox the air intake is in a snorkel between the passenger side headlight and the edge of the radiator. The snorkel has a fairly narrow opening which may limit airflow somewhat plus it isn’t in a direct path of air, the hood support blocks it somewhat. From the snorkel the air flows through a 2 5/8″s insulated hose over the top of the radiator then down the other side where it connects to a plastic funnel on the airbox itself. Each turn causes restrictions.That funnel has a 1 3/4″ opening into the airbox itself and is ,IMO, the biggest source of restriction in the intake.

After the air is inside the airbox it has a good amount of area to expand to evenly pass through the filter then changes over to the HFM as BMW designed for proper air metering.

The solution… aka the ‘Fogged Airbox’:

It increases the air intake hose into the airbox to a 4″ diameter hose. That is about a 700% increase in surface area for air to flow through easier. The setup grabs its air from behind the foglight for a good source of cold air. It keeps the paper filter for maximum engine safety and the flow to the HFM is exactly as BMW designed it. In my opinion, it addresses the real problems with the factory airbox and doesn’t create any additional problems like the other setups do. It costs around $20 and takes about 3-4 hours or so to build and install.

Testing, Does it work?:

Instead of using formulas and/or a flow bench I tested the best way possible, the actual DME measured airflow into my engine. This is the ideal way to test because if your intake/engine has a bigger restriction elsewhere the formula/flow bench methods will give you misleading numbers that won’t be realized in the real world.

As was mentioned above the DME in the Z3s has a hot film sensor to calculate the actual mass (weight) of the air flowing across it based on the temperature of the film. The cooler the film is the more air mass that has passed over it. If the airflow is smooth through the HFM it will be metered properly. From this and a few other sensors the DME can calculate the mass of the air being taken into the engine. In other words, your DME knows exactly what the airflow is… all I need to do is have the DME tell me that information on before/after runs and compare the two.

Getting that data is actually very easy, provided you have the proper tools. OBDTOOL plugs into the OBDII port on your car and can grab data directly from the DME. For this test I need to grab airflow (expressed as pounds per minute) and engine RPM and log that, in real time, to a file.

Grabbing the data is easy. You set OBDTOOL to log the data to a file, drive at about 1000rpm in second or third, then just floor it till redline. When you are logging two data points (airflow + RPM) OBDTOOL is able to receive 2-3 updates per second. To get more data points (for more accurate airflow curves) I made multiple runs and merged the data points together. The higher the gear you are in the more data points you receive as well.

For the first set of data I made 3 runs with my car in its normal modified state. That includes the Fogged airbox , a +3mm big bore throttle body and a Remus exhaust. My car is a ’96 without traction control so it doesn’t have the secondary throttle body in the air stream.

Since I’m interested in just the effect of the airbox mod that is all I changed for the second set of runs.

There is one slight problem though… I can’t get my airbox back to exactly stock as I modified my personal airbox. The factory stock setup has that 1 3/4″ diameter inlet into the airbox which is the largest source of the restriction in the intake system. So I made a 4″ diameter plate (size of the Fogged’ airbox inlet) that has a 1 3/4″ hole in the center to simulate having the stock inlet in the airbox. This plate fits right in front of the rodent screen on my inlet and is held in place by it with tape around the outside edges to seal better. Because the airflow on the incoming side of the restrictor plate is still less restrictive (because it is still attached to the 4″ hose of my setup vs. a 2 5/8″ hose and the stock snorkel ) the airflow numbers recorded are probably higher then if the airbox was fully a stock setup. Still, the test will give a fairly accurate result of the real world results of opening up that stock 1 3/4″ inlet to 4″s.

Looking at the graph (click on the picture to the right for a larger view), at low RPMs the Fogged’ box seams to flow a bit better than stock. In the midrange the flow is basically stock, then at 4000 rpm or so the difference begins to grow and above 4800 it grows very quickly. The difference is about an 8.5% increase of air flow at 6400 RPM. That means your engine is burning about 8.5% more fuel there too… which also means you are producing about 8.5% more power! In fact because you are getting more air in the cylinders at higher RPM (increased volumetric efficiency) you will have a higher effective compression ratio and you will extract slightly more power from the air and fuel you were already burning before the mod in addition to the extra power for the extra air and fuel.

Now my thoughts as to why the ‘Fogged’ midrange flow is nearly stock. Simply because in that range the stock airbox is not the limiting factor. I assume if there is a problem (the cylinders could be filling to capacity) that it is the intake manifold in that range. Above 4800 rpm the DISA setup has changed over the butterfly in the manifold and the stock airbox was limiting flow again.

As an aside I also tested the intake air temperature. According to my center console computer it was 70 degrees during the test. The intake air temp. (as read from the DME using OBDTOOL) was 75 degrees while just cruising around and when I floored it for a few seconds it dropped to 73 degrees and stayed there.

I’ve tested a similar mod on a 318ti and recorded about a 5% airflow increase. That car was a ’97 with the additional throttle body in the airflow (read extra restriction) and it the rest of the engine was totally stock.

I’ve also done before/after 0-60mph runs using a G-Tech Pro. With just the Remus exhaust I was able to get a best time of 8.1 seconds. After the airbox, throttle body and Garrett Lim’s software (with stock 6500 rpm rev-limiter) my best times dropped to 7.45 seconds. My software was overwritten last December, by the way. Since then I have run stock software with the box.

Tools you need to build it:

Something to cut a hole in the plastic airbox. You could use a 4 1/8″ hole saw, reciprocating saw, Dremel or whatever else you feel comfortable with.

A Dremel with a heavy duty cutting wheel, a grinding stone and if you are using it to make the main hole the dime sized metal bladed circular saw and one of the grinding stones.

Eye protection (you are going to need it!)

10 mm socket and socket wrench

Hot glue gun with extra strength glue

Regular screwdriver

Permanent marker

Scissors and Wire cutters

Optional but recommended: Propane torch and plumbing solder, only needed if you are going to install the rodent screen in the inlet

Stuff to buy:

Go to the nearest Home Depot or the like and in the ventilation section get an aluminum 5″ to 4″ duct adapter. This is a tubular piece to allow a 5″ hose to plug into a 4″ hose and should cost about $3.

Optional but recommended: While at Home Depot in the ventilation department you need a ‘Rodent Blocker’ which is just a square piece of heavy mesh which is meant to go on your clothes dryers exhaust to keep mice from getting into you house. Figure a buck or two for this.

Go to a Pep Boys or whatever and in their ventilation section they should have a 4″x72″ air intake type of plastic hose for under $15. Anything similar should work OK but make sure it’s flexible. The stuff at Pep Boys worked really well and is very smooth on the inside for better air flow. Also get a 4.25″ hose clamp for a buck.

Very optional: can of flat black spray paint… high temperature outdoor gas grille paint works OK

Making the inlet:

First thing you need to do is to make the 4″ inlet out of the adapter you bought at Home Depot. To do this on the expansion funnel (between the 4″ and 5″ tubes) you need to cut/drill the two rivets. This lets you separate the adapter into 3 pieces, the 4″ tube, the expansion funnel and the 5″ tube. We only need the 4″ tube with the ‘rib’ on the side that was connected to the expansion funnel which opened up to the 5″ piece. The 4″ diameter tube will become your inlet into the airbox.

Optional but recommended: Cut down the ‘Rodent Blocker’ to a circle that will just barely fit inside the ‘rib’ on the inlet and solder it in place. This is a safety step to be sure a rodent doesn’t climb up your airhose and decide to make a nest in your airbox.

Very optional: spray paint the intlet. Paint doesn’t stick very well so it will take multiple coats. I’ve found that it helps if after each coat of paint you ‘bake’ it on with a hot air gun but even then it can come off fairly easily.

Completed inlet along with the airhose and hose clamp. This is all the parts required for the Fogged airbox.

Adapting the airbox:

It would be helpful if you have the car up on ramps or jack stands but it isn’t required. You need to get the bottom half of the airbox out of the car. To do that release the four clips holding the top of the airbox on and move the top aside. Take out the air filter and put it somewhere clean till you are done. On the fender side of the airbox there are two 10 mm nuts holding the air box in place. Remove them.

Above the radiator is the big black plastic box which holds the stock intake hose. Take the cover off of this by removing the (4) 10 mm bolts. Two are right in front of the plastic box in plain sight, the other two are right behind the plastic box. Lift the cover of that off and put it out of the way. You will now see the stock hose and where it connects to the snorkel on the airbox. Just pull that hose off and the bottom of the airbox can be removed from the car. While you are at it disconnect the other end of the stock hose and remove it from the car as it’s not needed.

Now with the bottom half of the airbox out of the car remove the external snorkel from the box. There is a clip on it that will let the snorkel slide off the box. Notice how tiny the hose is at the end of the snorkel! Next the internal funnel needs to be removed. There is a clip on the top of that and another clip on the opposite side of the box that lets you take the plastic frame out along with the sound deadening material. Don’t worry if you break some of the plastic frame pulling it out… I left mine out and it’s not a problem.

Now comes the fun part, you need to cut a 4″ hole where the stock hole is. Put the 4″ side of the inlet over the hole, on the outside of the airbox, and draw a new circle on it with the marker. Don’t center the new hole over the old one. You basically want the new hole to be as far forward on the airbox as you can make it. You need the inlet to be forward so it can clear your steering reservoir when you reinstall the box.

Now PUT ON YOUR EYE PROTECTION!!!!! – Using whatever method you decided on, cut out that 4″ circle. This is the most time consuming part of the project as hot, nearly liquid plastic is going to be flying around so be careful! Be sure to also look inside the airbox as in a couple of places there are plastic ribs and such that you need to cut from the inside for the hole to come out. After you get the hole cut out test fit the inlet into it. The inlet should fit through the hole but stop at the ‘rib’ on the inlet. For the test it is easier if you just put the 4″ side through the air box from the outside but this isn’t how it will be when you are done. If you can’t get it to fit use the grinding wheel to smooth and enlarge the hole. It should be as tight a fit as you can make it without distorting the inlet. After you have the inlet fitting in the hole remove the inlet from the airbox. Scrub out the inside of the air box with a brillo pad or something similar to remove all the plastic bits that got thrown around when you cut the hole. After you get it all cleaned out dry it.

Now put the inlet into the airbox for real. The side with the rib goes inside of the box with the 4″ tube pointing out the hole. Only have the tube protrude from the box up to the rib on the tube hitting the inside of the airbox. The rib will keep the inlet from pulling through the box if you cut the hole properly. Now from the outside of the box seal the inlet to the box with the hot glue.

A slight air leak here isn’t critical, as it’s still before the air filter, but do the best you can. Do NOT use a silicone glue as it could cause problems with your O2 sensors.

After the glue has set, slide the 4″ air intake hose over the inlet protruding from the airbox. It is a very tight fit but if you take your time and work it around the inlet you will get it on. Secure it in place with the hose clamp. Do not overtighten the hose clamp as you could deform the inlet. You could cut the length down to about 3-4 feet now or you can wait till it’s installed then cut it.

Now reinstall the airbox into the car. The tube and hose should just fit into the car but it will fit if you cut the hole in the proper position. You can either put the old air intake cover back over top of the radiator and put back in the (4) 10 mm bolts or pull the bottom half of the plastic assembly off of the fan shroud and out of the car. I took it out of my car but if you want your intake to look stock leave it all in place.

Reattach the (2) 10 mm nuts holding the airbox in place. Put the air filter back into the airbox and secure the cover with the 4 clamps. Be sure you get a good seal. Now would be a good time to replace your air filter if you haven’t in 10k miles or so.

Route the hose pretty much straight down along the side of the radiator. This is easier to do from under the car but is manageable from above. Run the hose into the opening that leads into the front bumper, right in front of the wheel and place the hose opening behind the foglight. Trim the hose as needed to do this. You may want to disconnect your battery for a few minutes to reset the long term fuel trim (adaptation) that occurs in your DME but it isn’t really needed.

At this point you are done. You too now have a Fogged Airbox. Go take your car for a spin and run it up to redline a few times! You are going to love it!

Fogged Airbox FAQ:

(Q) Why not put the hose in front of the radiator or below the car or add a scoop or where ever?

(A) I’ve tried a scoop and it resulted in no measured airflow increase. It also greatly increases your chances of hydro-lock. Putting the hose behind the fog light gives good airflow and you would need to drive through a VERY deep puddle (several feet) before you would have a problem. Hydro-lock occurs when you fill your combustion chamber with enough water to stop the piston on its compression stroke. When this occurs you cause major damage to your engine. Even if you submerged your hose in water I don’t think the M44 creates enough vacuum to suck water up the 4″ diameter hose, then fill the airbox and finally draw it into the combustion chamber. If it was 1″ diameter hose, yes but not the 4″ hose. Your airbox has a drainage hole in it that would act as a vacuum relief if the main hose was submerged. Having said that it doesn’t mean I recommend that you drive through very deep puddles. You are not a U-Boat captain after all. You also want to limit the number of bends the hose needs to make as each bend will cause some restriction. Keep it simple and put the hose behind the foglight, it works.

(Q) Won’t rain get behind the fog light and get into the intake?

(A) Yes, but small amounts of water isn’t a problem. This can happen on the stock setup during driving in the rain or on very foggy nights. It basically wets your filter and since it is still a paper filter there is no problem.

(Q) Won’t I get check engine lights? IE… Dinan says that if the mod worked it would cause Check Engine lights as the DME won’t know how to deal with the airflow. They also claim the engine will run lean with any air intake on it and no ‘Stage II’ software in the car.

(A) Simple answer: No

Much longer answer: I’ve put around 15k miles on my car and never had a Check Engine light. No other Z3 owner that has done the mod has had lights that were caused by the Fogged airbox either. Remember how your fuel injection works, the HFM directly measures the airflow into your engine and injects the appropriate amount of fuel. It ‘knows’ how to deal with the increased airflow. This is assuming proper air mass metering by the HFM. I have done wide open throttle (WOT) runs recording my O2 sensor data vs. RPM to see what mixture I’m running at. At WOT my engines goes rich (O2 data at 0.8v) from right off idle and stays very consistent (+/- 0.05v) all the way to the rev-limiter. IOW, with the Fogged airbox the HFM is properly metering the airflow increase and the DME is adding fuel as needed throughout the rev range. This is with stock BMW software in my DME. If this was not the case the O2 voltage would drop off (go leaner) as I got above 4k RPM where the airflow increase occurs. On other types of fuel injection: air-flow metering or speed/density the DME would NOT be able to measure the additional air and they would need a ‘chip’ or ‘software’ to properly deal with the increase flow. On our cars this is NOT the case.

Even IF what Dinan claims was true (in that it applies to ALL intakes, it is not true) why wouldn’t the ‘Stage I’ software (which makes your engine run slightly richer at WOT and slight advances your timing) be enough? A theory on that: Assume for a second you have two intakes.. their actual airflow into the engine is the same. But one intake has smooth laminar flow at the HFM and the other intake has turbulence at the HFM. What would happen when you put them on your engine? Again, if you look at the inside of your HFM you will see that it actually only samples a small fraction of the air that passes through it. Based on that small sample it calculates the total air mass. This works fine if the airflow through the HFM is nice and smooth. In the case of the intake with smooth airflow through the HFM if you increase the airflow through the HFM your DME will properly measure the airflow increase and add fuel as required. In it’s maps it says for ‘A’ amount of air inject ‘X’ amount of fuel. Because the metering is accurate you are running at the proper air to fuel ratio.

Now what happens if you put on the intake that flows the same amount of air but causes turbulence at the HFM? Because the HFM only samples a small part of the airflow if the airflow through the HFM is turbulent the HFM will NOT properly meter the mass of air flowing through it and the DME will inject fuel based on the INCORRECT metering. You could end up running richer or leaner depending upon which direction the metering was off. The DME reads that you are taking in ‘B’ amount of air so it injects ‘Y’ amount of fuel. Since you are actually injesting ‘A’ amount of air (it was just mis-metered by your HFM because of turbulence) you should be injecting ‘X’ amount of fuel. But because it is in fact injecting ‘Y’ amount of fuel you are NOT running at the proper air to fuel ratio.

So, what can you do about this? Either smooth out your airflow for proper metering OR add a ‘correction’ (re-calibration) into the DMEs maps so that when it meters ‘B’ amount of air (still actually flowing ‘A’ amount) it in fact injects ‘X’ amount of fuel. This would result in you running at the proper air to fuel ratio.

If you ran that ‘correction’ on a car that meters its air properly (say a stock airbox, or an intake with smooth airflow that metered properly) you would end up running at the wrong air to fuel ratio. Hence, ‘Stage I’ and ‘Stage II’

(Q) What is laminar airflow and why is that important to your HFM?

(A) This is an attempt to explain it, not the best example but it was all I could think of. Picture six lines of people all running parallel to each other waiting to get on a subway.










RULE: When lines C and D move forward 1 person lines B and E move forward 1/2 a person and lines A and F move forward 1/4 of a person. Knowing this and starting at the front of the lines by watching the people in any single row (A,B,C,D,E or F) you can determine how many people total have gotten on the subway. For example if the first person from row A just got on the subway you can calculate that a total of 14 people have boarded the train. You use a small sample to determine the total knowing the ‘flow’ characteristics. This is what your HFM does but for air mass flowing into your engine.

(Q) Will I really feel the difference? Can it really work that well for so little money?

(A) Yes!!!

I have a very hard time convincing people that this works as well as it does. I’ve never heard from a person that has done it and not been thrilled at the difference. Above 5000 rpm it feels like a totally different engine. Now your engine pulls HARD all the way to redline. You might even hit the rev limiter a few times while you get accustomed to the difference.

(Q) Do you really think BMW screwed up on the airbox design that much?

(A) I think they met their design goal, so in that sense they didn’t screw up at all. Their engineers had to make a compromise between noise and airflow. The stock airbox acts like a ‘muffler’ for intake noise. Just like any other muffler this limits airflow somewhat in the pursuit of reduced noise. The Fogged mod is sort of like putting on a straight pipe for your intake. It lets air in easier but it also lets intake noise out easier. Under acceleration you will get a little bit of an intake roar but it sounds great and is perfectly appropriate in a sports car.

(Q) What do you get out of all of this?

(A) Besides the satisfaction of helping out other M44 Z3ers, nothing. 😉 To that end if you do Fogg your airbox please e-mail me or preferably post a message on the Z3 message board with your opinions on it.

(Q) But xxxx tried a K&N and noticed a huge difference in performance, why?

(A) Most of the people I’ve talked to that tried a K&N decided to give it a try when it was time to replace their paper filter. The reason they felt the increase in performance wasn’t from the K&N itself but because their paper filter was clogged and needed to be replaced. If they had just put in a new paper filter they would have felt the same thing. I replace my filters every 8-10k miles, I can tell when they need to be replaced as my gas mileage starts to drop.

Photographs courtesy of Rich Carlson, Robert Leidy and Tom Mosteller, thanks guys!

The Fogged airbox is copywritted by Shawn Fogg and is for individual use only.

Discuss this article and other Performance upgrades in the

///MZ3.Net discussion forum.

2 thoughts on “1.9 “Fogged” Airbox Modification

  1. Evan

    Hey what should I do about the second box that leads into the main airbox with the filter? Should I leave it alone or cover the hole so that no air can enter it? Thanks!

Comments are closed.