MS3 Custom CAI

When it is time to clean my MS-CAI filter for the first time (soon - this NC clay is a b****) I am going to convert the thing to an SRI with a box picking up air at the front. Mostly because I don't want to deal with the cleaning headache a second time. The thing is, I believe that if I shorten it, I will regret doing so if I don't put some sort of straightener in the air path. Do any of the major chains (or websites) carry airflow straighteners for do-it-yourself types?
 
^ Dude that is just a bunch of bulls*** all wrapped in a neat package with a "Mazda Tech Answer" bow on it.

What you have here is completely wrong.

MAZDASPEED intakes are made by AEM, and while they are nice, it has been proven time and again that they actually cause problems all by themselves, typically due to their incorrect MAF metering section. This slight mis-calibration throws off some cars more then others, but ask around and you will see what's up. Truth is, whether it is an engineered SRI like Cobb's, or an engineered CAI like CP-E, you'll find the main factors in those being so successful and seamless in our cars to be the same thing: the MAF section is identical to stock in size.

Mazda didn't tune for anything with this intake, or any other. This computer learns the A/F, MAF inputs, and driving habits amongst other things and eventually creates its own trims depending on those variables.

Lastly, probably the reason why CAI's are working here while SRI's aren't is NOT that too much air is entering the intake! This is just silly. It is likely an issue of the short ram not being designed properly to straighten the airflow in time (unlike a long tube will tend to do), and this non-linear airflow is disrupting the MAF! The Cobb SRI doesn't have this problem, because they engineered a velocity stack to straighten the flow out. Simple.

This is your first post, but seriously, I'd reconsider telling anybody you are a Mazda tech until you've got the facts straight, not just what someone told you.

This is not to flame you or insult you, but to correct the mis-information that is coming out of your post.
(iagree)
Hey fourthmeal I am getting i kick out of your posts.
Everytime I read a post containing wrong info or something that just has to be warned againts I scroll down and see you have written exactly what I wanted say. People need to do thier research on thier mods, alot of engineering has gone into these cars.
 
Do any of the major chains (or websites) carry airflow straighteners for do-it-yourself types?

Well, I searched, but couldn't find anything.

I went back to the spectre website, looking for a straightener, and found a kit to convert the spectre maf pipe ID to the stock maf ID. The stock ms3 maf ID is 2 3/4 in., spectre is 3 in. The kit comes with 3 different size inner pipes, and one is 2 3/4 in.

I will be getting this kit for my custom cai and I'll update the first post.
 
i was bored today and thnking about doing the SRI.

what i wanted to chime in on was that the stuttering issues people may have could be caused by the inaccurate diameter of the MAFs housing. ok we know that. so what to do.

im putting the "backstory" in quotes

well, it just so happened that when i first turbocharged my 240sx, i used a hacked mafs setup. it used the principle that going from stock 270cc injectors to a set of sr20det 370cc injectors that dropped in the fuel rail, was an increase of ~37%. luckily, late model 240sx's used plastic mafs housings with a probe style, hotwire MAFs. DSMs had done this before as well, so what we did was cut the MAFs housing to just the probe and square area above it that housed the circuit board/electronics.

Doing some calculations with radius, diameter, pi, and using our 37% increase from stepping up to the larger injectors, we were able to find out the size that the "new" MAFs needed to be. the stock MAFs was ~2.4xx" inner diameter, and after the calculations, we came up with a figure that a 37% larger MAFs housing would have to be ~2.875". perfect. we then placed our MAFs probe in a 3" outer diameter pipe, with 1/8" walls (.125"), making the inner diameter 2 7/8" (2.875").

this was all done to bypass having to buy an s-afc,and cut the fuel trims by 37%. plus, being an s-afc tricks the MAFs signal going to the ECU, the timing maps remained unchanged and can cause problems when starting to really fool with s-afc "tunes". (the hacked mafs setup we used, we just used a .75 degree of timing retard per lb of boost, and we deviated from the stock 20 degrees by moving base to 16* BTDC and were only planning to run 7-10psi, and it worked out perfectly, for being way cheaper and definitely safe.

anyways, we found there were slight inaccuracies in guessing the MAFs housing, being that a "percentage" increase involving cross sectional area and volume and flow patterns really couldn't be spot on with simple calculations (whereas going from 270cc to 370cc on the injector side of the equation was rather straightforward).

what we did was develop a "shim" to slide into the hacked MAFs housing, effectively reducing the diameter of the MAFs housing and kinda getting the fuel trims back to close to normal. before, they weren't way off, but they were slightly lean. the driveability test passed with flying colors, and the wideband increased noticeably when the shim was inserted.

all the shim was, was a piece of 2.875" OD piping with 1/8" wall thickness (mild steel exhaust pipe you can buy anywhere - stainless might be better to not build corrosion in humid climates), and it had an ID of 2.75", and the car ran a bit better, at least slightly richer to keep peace of mind.

here's some pictures of it, i'm going to attempt the custom SRI (because CAI vs SRI is a dumb argument and it really isn't much of a difference anyways). and while redspeed found the correct MAFs housing size, this could be a viable option as well...though i'm not sure how well it would work with the plastic spectre pipes, not sure how easy it will slide in and out.

these are just thoughts, and luckily newer MAF sensors are mounted on the pipe with screws, so it's easier to remount them (we had to get creative). but short of buying the right sized pipe, this is the only thing that will effectively "fill" the area without being a worry that it will get sucked into the intake tract or worse.
 

Attachments

  • mafsdiameter.webp
    mafsdiameter.webp
    29.5 KB · Views: 190
  • mafshimout.webp
    mafshimout.webp
    50.2 KB · Views: 202
  • mafshimgoingin.webp
    mafshimgoingin.webp
    48.2 KB · Views: 188
  • mafshimin.webp
    mafshimin.webp
    50.1 KB · Views: 213
Excellent post ispypsi! (2thumbs)

The spectre kit that I'm trying to get is the same principle that you mentioned.
 
FWIW, I just put on the short ram, basically just the MAFs kit and an AEM filter (no oil, bout the best flowing one out there).

I'm not a believer in butt dynos, but I could definitely tell it's an improvement. Like alot. Unfortunately, as soon as I hit 2nd gear, it falls on it's face. 2nd gear spools and 3rd gear seems sluggish throughout. I only took it about a tenth of a mile down my driveway and back, so not really sure yet. I had the battery disconnected for about 30 minutes during the install, so I'm debating whether to leave it on for a few days to let it settle out or not. The bogging is just like the others mentioned, very hesitant. I don't feel like messing with the CAI, so if this doesn't work out, I'm likely to buy the BEGi kit, the one with the little enclosure deal. (Plus it's about the cheapest out there for what you get). And maybe pair it with an ATP inlet pipe, if they revise it to fit better - i hear its got some issues.

Edit - other option to hold me over is to put top half of the stock airbox back on and ziptie the filter to it (aka ghetto rigged). might help a little since the bottom of the box and the tiny 2"x3" hole in there is all the breathing and that just plain sucks. or doesnt. ha.
 
:D

im hoping to get it as my license plate soon. most people don't understand it.

and then still don't understand it when i say "pound per square inch"....
 
:D

im hoping to get it as my license plate soon. most people don't understand it.

and then still don't understand it when i say "pound per square inch"....

My favorite "trick" plate was one I read about. Apparently it took a couple of years for the registry to catch on and revoke it, but a woman had one that was "3m ta3". Doesn't look like much unless you imagine it in your mirror.
 
I really want to do this mod but with polished aluminum instead of plastic. So far my grand total on parts is around $165 (with out shipping or handling). Has anyone done this without putting the 90 Degree at the end? From the brand name CAI's I've been looking at some put it there and some don't. It looks like that might also cut back on some of the turbulence too. If I don't put the 90 in there then I think a splash gaurd of some sort would be a worthy investment. Let me know what you think.
 
I don't think you will have any problems with a 90* elbow. Mine been there for a year now w/o any issues. Problem is space. If you want your filter to come up and give you a better chance against hydrolock, you need the 90* elbow in there.
 
I did a version of this, but short ram since I didn't want to spend money on the extra piping if the MAF section wasn't going to work. I got a CEL for lean idle, as well as a bucking problem at mid RPM high throttle situations. I took it off for a week or so and fabricated two metal "straighteners" to mimic the stock air flow straighteners (after first trying aluminum screen material that didn't work). Well, the straighteners did their thing and the short ram has been on for over a week CEL free.

I thought about extending it down with some 3" mandrel bends I have laying around in the garage, but I think I'd rather make an "induction box" like Rally Sport Direct has for some other turbo cars.
 

New Threads and Articles

Back