Mitsubishi Eclipse 3G Club banner

All you ever wanted to know about our MAF(s)

54K views 41 replies 14 participants last post by  Howl83  
#1 ·
Over the years of production Mitsubishi has largely used a Karmen Vortex Mass Air Sensor (until lately but the only ones relevant to us are KV). These sensors are special from all others, I won't get into their exact working because it is quite complicated and moreover irrelevant to their employment in these cars but the important things to consider are these:
1) Airflow must go STRAIGHT through (hence the honeycomb screen)
2) Airflow must be UNIFORM across the the cross section of the flow pattern
3) They are NOT designed for blow through (PERIOD!)
4) They are NOT adjustable
5) They are NOT cleanable (contrary to popular belief)

Other things to know is that the MAF also contains the barometer and intake air temperature sensors in these applications.

A lot of people don't like our MAF's because they provide a large restriction on the intake. This may be true but not to the extent the world seems to believe and there are things one can do to make the most of using these sensors, upgrading them and tuning them for maximum performance. This is the prerequisite for a few points I'm going to make here.

3g's came with two different MAF's throughout their production. The first and by far most common is the 501 MAF sensor.
Image

This sensor came on all GS's, RS's, and GT's over all years.

The other sensor is the 482.
Image

This came on all GTS's, automatic and manual. It also came in some Montero Sports.

Notice the differences, 501 top, 482 bottom.
Image


The 482 is geometrically identical to the 399 MAF that comes on the Evo. It however has different scalings. To my knowledge it has the same air metering capabilities, I have mine out close to 550 at the crank and I haven't topped it out yet. It provides a HUGE improvement in response and flow over the 501.

The 501 however is a different story. Of a few cars I've tuned and worked both suck through boosted and the big block naturally aspirated the 501 has failed to perform. It's as if it can't keep up with the air flow in these applications and had to be swapped out for the 482.

I've run into a number of tuning problems with both MAF's ultimately. Refer to the 5 key points I mentioned at the beginning of the post, unless all 5 are satisfied you will have to compensate in the tune for inaccuracy. The MAF's come attached directly to a straight flow air box providing for the flow requirements, the problem is when you upgrade to aftermarket intakes and upgrade engines a few things begin happening.

The first thing to note is that YOU CANNOT USE OILED AIR FILTERS WITH THESE MASS AIR SENSORS. The oil coats the filament and their metering abilities are eliminated. You can't clean these sensors either because the elements are coated with a special film that MAF cleaners dissolve.

Secondly, open filament air filters have a tendency to spin the air as it goes through the filter into an intake tube. The honey comb in the MAF helps take care of this, so that is not deletable.

The next problem, and most common results from installing just about any aftermarket intake. Here is a comparison between the MAF plate that comes in any AEM intake and a 4.5" plate that I bought off ebay.
Image

The result with the AEM intake plate is it focuses air into the center of the MAF. This can skew the density distribution of air through the cross section of flow through the MAF. This doesn't hurt things in stock form because the charge velocities on a stock engine or other low horsepower application aren't high enough work against an even air distribution.

However when bigger engines or forced induction begin pulling in more air the velocities through the MAF start climbing into a range where the elasticity of air causes the distribution to start changing. A small plate like the one that comes with the AEM (or ANY ebay intake) will ruin the meterability of the 482 because the opening is larger than the plate diameter. As I said earlier I have no idea why the 501 loses its functionality, but I have yet to be able to make one work well on anything more potent then a bolted on 3 liter, they all run lean to the point that it can't be compensated for.

On the 482 plates like these cause flat spots in the MAF curve, meaning chunks of your rpm range always run lean or rich. The fix I've found for these is the 4 or 4.5" plate.
Image


This is one of the reasons I will always recommend a custom intake. Put a filter right on the MAF with the 4.5" plate and a 3" pipe. What it makes for is an even air density across the maf and straight flow. As well as the least restriction you can get for the meterability. Like I said, I'm making something like 550 crank right now with the 482 and I'm using 100% stock MAF curves and settings.

On the 501.
Image


On the 482
Image


Swapping these MAF's REQUIRES a flash. There are 4 tables related to air metering with the maf. MAF Size, MAF Sensor Filtering, MAF Smoothing and MAF Scaling.

If you're going to be doing a MAF swap and doing any level of tuning you'll want to follow this tutorial for redefining your Scaling and Smoothing tables.
How-To: Rescale your MAF (without pics) - evolutionm.net

What this does is put the left column of your tables into HZ. This is the loggable readout given to the ECU by AirFlowHz in Evoscan. To tune the MAF you want to log your wideband, airflowhz, AFRmap, rpm and ecuload to see what your target AFR's are, when and how far off they are. Given properly dialed in injectors this method can be used to track down a MAF curve. If you have the appropriate intake however this should be unnecessary.

The tables I'm posting below are for a ROM CONVERTED BY THE ABOVE TUTORIAL.

For the 482:
Size: 358.8

Sensor Filtering:
192
241

Smoothing:
141
140
134
132
131
131
131
131
131
132
132
132
132
132
133
132
131
130
130
130
130

Scaling:
145
150
164
176
181
184
188
192
196
198
202
204
207
213
218
223
231
235
235
233
230

Since no one would possibly upgrade to the 501 I won't bother posting the settings for it. The 399 MAF is also swappable but I don't have the scalings for that on hand at the moment. If someone finds them feel free to post.
 
#2 ·
Justin, isnt that the smoothing/scaling for the 399 Evo MAF?

only ask because the GTS (482 MAF) tables that I have are different.

What I got:

For the 482:
Size: 358.8

Sensor Filtering:
192
236

Smoothing:
134
134
134
134
134
135
134
134
133
132
132
132
132
132
133
132
131
130
130
130
130

Scaling:
145
150
164
176
181
184
188
192
196
198
202
204
207
213
218
223
231
235
235
233
230
 
#3 ·
My Collection:

From top to bottom (501, 482, 399)
Image


501
Image


482
Image


399
Image


Adaptor bought from Advance Auto:
Image

Image


The fit:
Image


The size: (4")
Image



Adaptor that is sold with the Evo K&N intake:
Image

Image


The fit:
Image


The size: (6")
Image


All images appearing on club3g.com posted by user "Drunkoffjuic" aka Eric Nemcek, are the exclusive property of Eric Nemcek and are protected under the United States and International Copyright laws.

The images may not be reproduced, copied, transmitted or manipulated without the written permission of Eric Nemcek .

Use of any image as the basis for another photographic concept or illustration (digital, artist rendering or alike) is a violation of the United States and International Copyright laws. All images are copyrighted © 2006 - 2016 Eric Nemcek .
 
#5 ·
For the 399:
Size: 357.5

Sensor Filtering:
192
217

Smoothing:
115
115
123
129
128
127
128
128
127
127
126
125
125
125
124
121
121
121
121
121
121

Scaling:
145
152
165
175
182
187
191
195
198
200
203
206
208
217
222
226
233
236
237
235
232
 
#6 ·
#8 ·
:(
That's why I posted up the roms I am going off of. Just in case you didn't believe me. Proof is in the evidence. ;)

:agreed: people should be fine either way.


Sent from my iPhone using AutoGuide.com App
 
#9 ·
You can also modify the 501 MAF by pulling out the air horn, couple that with the Evo MAF adapter or simply modify the air horn and you can get a 482/399 MAF air flow. This info was posted over on GalantTuners.com as well. I have been running a modified 501 MAF now for the past year
 
#10 ·
- 501 -
Typically found in Mitsubishi vehicles with a 4G64 engine:

ECLIPSE 00-02 (All trims)
ECLIPSE 03-05 exc. GTS
GALANT 99-03
MONTERO SPORT 99-03 3.0L (6 cyl)
OUTLANDER 03
SEBRING 01-05 (Coupe)
STRATUS 01-05 (Coupe)

This MAF is the most practical swap for a non-turbo Lancer, as the MAF values from the 4G64 ECU are easily adapted to the Lancer's ECU with EcuFlash. This MAF is only rated by Mitsu for 210 liters/second, and has ~.78 psi pressure drop because of the larger airhorn blocking the inlet.

-------------------------------------------------------------------------------------------------------

- 482 -

Typically found in newer Mitsubishi vehicles with a 3.0L, 3.5L or 3.8L V6 engine. This MAF flows almost identical to the 399 Evo MAF, ~61lbs/min maximum. This MAF has a pressure drop of ~.54 psi at 240 liters/second. This MAF can be swapped into a non-turbo Lancer, but the values used for the Lancer ECU will be trial and error as this MAF flows almost the same as the Evo MAF and is found on larger V6 Mitsu engines. Turbo'd Lancers will have a better transition to this MAF as they can use either the Eclipse GTS or Evo MAF values as a starting point, depending on the individual setup.

-------------------------------------------------------------------------------------------------------

- 399 -

Found on the 4G63 Evo VIII/IX engine. It is the highest flowing MAF available, flowing a maximum of ~62lbs/min with a pressure drop of ~.526 psi at 240 liters/second. The 399 MAF is pretty much identical to the 482 MAF in both size and airflow. This MAF can be swapped into a non-turbo Lancer, but the values used for the Lancer ECU will be trial and error. Turbo'd Lancers will have a better transition to this MAF as they can use the Evo MAF values as a starting point.
 
#11 ·
Wow some truly great info in here, which has opened my eyes to something I've done.

Assuming all v6 models used the same MAF, I pulled a 501 at the junk yard to replace mine when I tried to drive through a small Sandy-induced lake, and it's been working just fine, probably because I only have a few bolt-ons. From the looks of it, this thing might be choking me down a little bit.

So, I'll be searching for another at the local yard soon, but 3rd gens period are hard to come by there. I may be able to find a Montero though.

In the meantime, it may not change anything but next month I plan on sending my ecu off to Cody for a tune, and if I can't find the right maf is this something I'm going to have to make him aware of so he can include it in the flash? Or is it imperative I find the right maf before getting the tune?
 
#12 ·
You can also modify the 501 MAF by pulling out the air horn, couple that with the Evo MAF adapter or simply modify the air horn and you can get a 482/399 MAF air flow. This info was posted over on GalantTuners.com as well. I have been running a modified 501 MAF now for the past year
 
#16 ·
So I won't really be SOL unless I go boosted or big block? I plan on trying to find one anyway but want to make sure that if I can't, Cody can tune the right smoothing and scaling. I don't know dick about tuning, so I'm going to learn on my $300 Camaro before I go trying to change anything myself on the car I make payments on.
 
#14 ·
Could you clean the sensor element with a mild solution of dish soap and distilled water blown through with a 'wash wand' venturi blower, rinsed with distilled water, then allowed to dry completely over the course of a day or two or in a warm dry room?
 
#15 ·
The electronics are not sealed from the inside, when water gets inside it gets into the circuit board and kills it. This is why people who suck up water always end up replacing their MAF's rather than just drying them out. Besides that dish soap dries with a residue.

Plenty of people will say they've had luck cleaning the MAFs but it's all placebo effect. It's not a hot wire MAF, it doesn't bake dust onto the elements and so there is never really any reason to clean it anyway. Unless something actually gets into the MAF and sticks to the elements it doesn't change much. For stock applications people have gotten away with solvent cleaning because the light changes in metering are soaked up with some grey area in the tune. The 00 Fed Spec GT aside the factory tunes are very conservative even given premium fuel.
 
#17 ·
My understanding was that sucking water kills the MAF when running because the circuit is powered and things short, but off the car with no power and using distilled water (which is nearly non conductive) shouldn't short anything since there's no power to short.

I ask because sometimes at work I encounter a car with KV that's ingested a little oil through the intake system due to being grossly overfilled or just a tired crappy engine, I know well enough not to use a solvent cleaner but wanted to know if something more mild would not harm them?

I'm aware of residue, hence the rinse.
 
#18 ·
There are capacitors inside the MAF that hold a charge even when the car is off. Introduce water to them and you get your short. If you manage to avoid getting water into the circuitry then I imagine you'd be fine though I haven't been able to successfully clean a mitsubishi KV MAF through my experimentation with my extras. Other companies might be more prepared for cleanout.

My original 501 got oil in it when I had a blocked rear PCV hose. It didn't work well in the first place so I upgraded to the 482 and tried cleaning it out. To this day the thing still seeps oil. Every time I pull it out of the drawer there is more to wipe out.
 
#20 ·
a long time ago i used MAF sensor cleaner (the gray can crap). my end result was the car ran like crap for a couple days then back to normal. no gain in power or anything like and after that never did it again. i was just happy i didn't have to replace it.
 
#24 ·
Justin, you should crack open a 501 and a 482 to compare the boards and sensors. He could be right that the sensor/board is the same between them with a difference in calibration due to a different diameter housing.
 
#26 ·
If that's the case then I don't see why one couldn't get batshit crazy with putting the sensor in any sized tube with a mesh flow straightener and then tune their ROM to read correctly. That's pretty much what us Toyota folk do with our hotwire. Just compensate the velocity of the air changing due to different diameter tubing and the sensor sitting more or less in the center. Hell, you can play with leaning the system by spacing the Toyota hot wire in the housing. No reason you can't fuck around with the KV in a similar way is there?
 
#32 ·