Adding An Automotive Cold Air Intake

[Thomas] and a buddy were sucking down a few brews when they decided to hack their 2001 Chevy Cavalier for a bit better performance. If they could find a way to bring cooler air to the engine they speculated that they’d see an increase in efficiency. Instead of routing the air intake to a hood scoop, they took off the factory air filter and mounted a cold air filter in its place. PVC pipes were then used to create a delivery path from the front of the vehicle with the output in close proximity to the new filter. They tested their work and discovered a drop in intake temperature from 101 to 48 degrees Fahrenheit at 60 mph, and from 109 to 54 degrees Fahrenheit at 45 mph. Now the sedan runs better and generates more horsepower, all for around $35 in parts.

117 thoughts on “Adding An Automotive Cold Air Intake

  1. Can I just add,

    How cute, people who think they know about cars… CAI’s are a favorite of the honda wannabe kids. Unless your car is horribly designed by idiots a CAI will not help you. Just like how a K&N filter wont help you, or any miracles in a bottle, or any special sparkplugs….

    Hack a day is not the place to find REAL automotive information. Go find a website where real people are doing real stuff to their cars.

    Posted at 5:48 am on Mar 17th, 2010 by fartface

    I know of a miracle in a bottle. K Seal, I poured this into my radiator on my Nissan people carrier with suspected warped Head, confirmed by a Nissan dealer mechanic. Coolant would escape the radiator cap, and the engine would overheat. A further test confirmed it by a colour change in the test tool. Repair estimate just over £1000, K-Seal fixed it for £15, the engine temp never rose above normal. The engine outlived the car!!
    Now, about this Hack. Stop slagging each other off. People want to learn and improve their cars. The stock intake is designed to keep engine noise to a minimum and reduce emissions. Obviously you have to mod other things to make a whole improvement, I find car forums full of people who scorn rather than teach, Hackaday provides information on a Hacked device, the OP’s car has been Hacked. Somebody do a Hack on there intake and do a write up. Provide us with hard evidence of what you have measured and teach us how to Hack our intakes. I personaly did my last vehicle, because it would jump out of first gear and I often stalled the engine when I pulled away in 2nd gear. Instead of revving more, I changed the air intake instead of my driving style and hardly ever stalled the engine.


  2. It works just fine.
    This sort of thing has been around forever.

    Off-road fans often relocate their intake above the car.
    As for having the CAI in the location that they do is not uncommon either. Just avoid driving in a flood.
    As for heat soak, I don’t know how well PVC will hold up after a long trip. I’d wrap it up with some header wrap.

    Most cars today also have an air silencer box on the intake to reduce noise. It also reduces airflow and soaks heat as a trade off.

    My G35 has the same motor (VQ35DE) as the 350Z. The G35 being a luxury sport car has the silencer on the intake. the 350z dose not, and because of that the 350z gets a slightly better increase in HP.
    I mention this because most CAI systems remove or bypass the silencer.

    ALSO commercial CAI systems that mount the filter in a low location such as our friends in this Hackaday post come with bypass filters to let water out so that it wont enter in the intake manifold.

    I think this hack would help increase airflow, but heat soak from the plastic material would slightly void any gains after a long trip.

    I myself have made my own intake. My gas mileage have improved by .5 miles per gallon since according to the information on my car’s computer.

    As for the people saying this wont work. Compare the intake on a 350z and G35. It looks like it’s been working for years now.

    I’ve also seen some rednecks rout an air conditioner duct to cool the intake…


    Might be a good reference article.

    As for the repeated fingerwagglings over what does or does not do what? Hacking has a tradition to uphold. We used to be in the forefront of showing “Proof Of Concepts” as truly observable Fait Accompli demos. Yes, there’s always an element of not the final build methods so to speak. But lest we forget?

    The first variants of however many different Hacks we share on some basic concept ALL add to our data bases of what “works and why or not”

    Post new info- cool- dissing etc in a non-productive fashion is Zero Cool ok?

  4. O2 sensor as a good reference for what ?

    The problem with the CAI stuff is that people use flawed methods for determining it works. Like the on board mileage computer, its set to read a certain grams per second flow rate, it knows that x amount of grams at y rpm, uses z amount of fuel.
    Then you change the domain for one of the input parameters and the calculations are now incorrect, especially since its measuring flow rate and not density, it can do some calculations based on IAT.

    Even measuring a few HP on a dyno is generally useless since dyno’s are very low resolution and a lot of things affect their operation, you’d have to recreate the exact setup each time, dynos do have ways of changing for ambient temp etc, but not for engine temperature.

    Though its getting more common to use wide bands, most cars only use narorow band sensors, which are great for determining if a car is running its fuel mixture at 1.4.6:1 lambda 1 for petrol, but outside of that very ‘narrow band’ all it can do is say if its rich or lean, and not by how much.
    So the car’s ecu sees that feedback and adjusts the short and long term trims for closed loop operation to try to keep fuel mixture at the closest it can for the perfect chemical burn to keep emissions low, this may be not the most efficient fuel consumption ratio.

    ecu tables for a MAF car usually work on flow rate or load (which is some derivative of flow rate) and RPM, so if you somehow get more air past the MAF sensor at the same RPM it might push the ecu into reading a cell above or below (if you lower the flow rate by making the tube the maf sensor is in larger) however the maps are tuned on the engines observed VE, so when you artificially change that, you’re not necessary in the optimum range anymore.

    as other people noted if you add bends and remove factory pipes, you’ve also go the problem of non laminar airflow going across the sensor, this can cause a rougness in idle as as well as have the same effect as changing the size of the tube the MAF is in, which again causes the ecu to incorrectly determine the correct cell for its calibration tables.

    the ecu uses the maf flow rate to determine what temperature compensation values to use as well, so you can end up getting the wrong ones of those too.

    hacking is good, but getting into something thats so commonly pushed as a good thing with flawed testing methods means people can end up with worse systems overall.

    Just read the legalese on the people selling CAI’s. You might get zero gain, find a supercharger retailer that has the same legalese (ie zero gain).

    There is an alternative medicine that works, its called medicine.


  5. Hi, Automotive Engineer here.

    I’m sorry, but cold air intake systems are more or less a joke. If you’d like to understand why, install air temperature probes inside your engine, Yeah, yeah, you’ll have to drill some holes, but here’s what you’d find:

    Yes, you are getting cooler air in the INTAKE TUBE. However, as soon as you enter the throttle body, you are faced with lots of thermally conductive metal which is at operating temperature – As the air travels through the throttle body, intake runners, and eventually into the cylinder head, it heats up. The temperature difference between air entering from a true “cold air intake” and your factory designed airbox is negligible. A *real* dyno test will show this – I’m not talking about the doctored dyno tests that cold air intake manufacturers show you; I’m talking about an objective test which plots your torque curve vs RPM, with the same conditions each time.

    IMO, the small difference is not worth it. If you take the integral of the curve on some engine condigiurations, you will find that you have hardly gained anything, and have simply “moved” power from a lower RPM to a higher one, or vice versa.

    Cold air intakes are for kids who like bling bling under their hoods. Nothing beats a properly tuned (professionally modified or stock) engine management system. These claims of huge power/economy increases with cold air intakes are 100% bunk!!!

  6. Of course the big question is “did he really drop the temps 50f,” because I doubt he got more than a few degrees for his trouble.

    Amazingly car manufacturers aren’t stupid, and neither are the guys selling after market bolt ons.

  7. Although you do get a cooler sounding intake, your ECU will compensate your fuel flow to match the MAF (Mass Air Flow) and IAT (Intake Air Temperature) average to derive the same power efficiency. To get a better hack on your power bandwidth, you need to attack the electrical side of the house. These changes can be made as simply as adding a resistor in parallel or series in line with the sensors depending on the engine……

  8. I agree with jim – there’s no simple way to gain much more efficiency from engines, because the engineers usually know what they’re doing.

    But I did do this on my rav4 (for less), and did find slightly higher manifold pressures and lower intake temps:

    The downside is you have to clean your air filter more often. But it does sound a little better and I did possibly yield 5hp more.

  9. honestly you guys should pick up some Dsport mag go the the back and read the test and tune, i’ve seen older cars that might have a few years on them gain 5-6 peak horse power from a full synthetic oil, proper air intake, and a spark plug change, I saw a gain of gas mileage in my jeep when i got rid of the paper filter and got a K&N filter, i have no understanding of any hacking but i do know my cars, i’m Sr20det swapping a s13 and rebuilding a KA24E for a pathfinder. I disagree with most of you the engineers know how to get a motor efficient as cheaply as possible, many motors have the ability to do %100 more horse power it just takes a little work and a good bit of money

  10. Seems this is a perfect example of to many people and too little knowledge. There is a reason import 4 bangers are so efficient and powerful. They use Turbos to get high compression, which equals power, but does nothing to hurt efficiency, where as US manufacturers use various Piston, connecting rod, and cylinder dome configurations to get higher compression for power. The US method is more wear on the engine, and lowers efficiency. The Jap method is more elegant, as it provides the power when you need it, while reducing fuel consumption and wear. The most efficient engine is a low compression engine, which is what a Jap engine is when the turbo is not spooled up. I am generalizing a bit, but if you want specifics, look at the NA 4G63 mitsubishi engine compared to the turbo 4G63 build. Both get great mileage. Both have cold air intake. They improve the cold air intake and add the turbo, with no loss on efficiency, and great power gains. Now take a same year US model 2.0 liter engine and compare (compare NA models to NA, or Turbo to Turbo)it to the 4G63. It lacks the power and efficiency of the Jap designed engine. The biggest issue is the stock compression ratio. Any external engine mods net you more power by increasing either efficiency or air flow. Increasing air flow does not harm efficiency, so either way, you get power with no loss, and possible gains. Any internal changes to net more power cause a decrease in efficiency with two exceptions. The first goes back to air flow. Porting heads and opening up valves nets power gains, while not causing losses in efficiency if taken by itself. The second is removing moving mass from the engine. Boring out cams, shaving mass off crank lobes, removing rotating parts that can be done away with (i.e. the balancing shaft of a mitsubishi 4G63), or even oil scrapers to remove excess oil from crank lobes. The only internal mods left all increase compression. This kills efficiency, but does net HP. You want more power, better gas mileage, and no loss on efficiency, try these things…. Cold air intake. Aluminum is NOT the best choice under the hood. It is great outside the engine compartment, but bad inside for the same reason. It has a high rate of heat exchange. You want that outside the hot engine compartment, but not in it. Plastic actualy is best inside, unless you use jet coat or another ceramic or other based heat shield on the intake under the hood. Try reducing driveline weight. I knocked something like 60 pounds off my drive shaft on a 3000GT VR4 by using a single peice carbon fibre shaft to replace the 3 piece steel one. The added gain was a rebuildable drive shaft, where as the stock one was not. Driveshaft was garbage because a $20 u-joint was not replacable. On my Hummer H3, 3000GT and eagle talon Turbos, I see gains from using lighter wheel and tire combinations. I also use these same techniques on bikes. If the vehicle is new, and has not seen many oil changes, flush everything and go with a high grade synthetic (do not do this on an old engine It will knock) The decrease in friction will improve efficiency, and net less wear for a longer life.

  11. someone stated this wasn’t like over-clocking. It is the same thing. Over-clocking creates more heat. Getting more power from an engine creates more heat. In both cases, you have to have systems in place to disipate this heat, or you get a failure. Any engine mod should be followed by a period of higher awareness of your engine temps. It is best to have a guage set that monitors head, block, and manifold temps. Or get a data logger that records it all on newer cars.

  12. I’m writing this because I want to learn some things about my car & having to read a ton of ‘trash talk’ is sure slowing that goal down.
    A few people have noticed that all these posts give more information about the writer than the subject.
    Yeah, the webmaster is the winner, bringing out a huge crowd of people with relationship problems (name calling, etc. – ’cause maybe you can’t talk to your ‘Old Lady’ this way?).
    This attitude is not helping any of the forums I’ve tried, but hey, I guess it’s better to vent here than punch somebody out – or not.

  13. I have a 2004 Ford Taurus wagon with 115000 miles & a 3.0 24V V6. EPA 18-25 mpg. The stock air intake system has not been altered except for a K&N filter added 6 years ago. I have tested my proprietary system for one year with the following results:
    1) Intake temp. to engine from +2 deg. above ambient to 8 deg. below.
    2) intake humidity to engine from 18 deg. to 60 deg.
    3) Avg. mpg from 19 to 21. Hwy. from 24 to 27.
    4) Power & performance like a new V8.(Not dyn)
    5) Suspect better tail pipe emissions. Can’t find any place near Kansas City to test.
    Details of simple cheap system are secret. May sell to Ford for $2 million if they are interested! John Schroeder

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.