I've got a bunch of EcuTek logs from some filter testing I did a while back I should be able to look over that data in the near future and post it up for all to see so we can unify the theories and come up with a solid answer backed up by hard data.

I don't disagree with what your saying above about the sample area as that will affect the amount of air "seen" by the sensor. Going to a larger or smaller diameter maf tube would certainly fall into this scenario. I just don't think a change in the filter element material would cause a change in the sample area.

 

Just took a quick look at the logs comparing the stock filter to the works dropin. Same day, no tuning changes, runs about 10 minutes apart, both on 2nd factory reflash.

At almost every data point the grams / second reading is about 5 g/s lower with the works filter even though the boost (as measured by the factory map sensor) is about 1psi higher with the works.

I'll clean up the logs and post the some relevant data points this evening.

 

how can g/s be lower if the boost is higher at the same temp?

 

Because boost feedback correction was turned on and the calculated load is lower with the works filter so the ecu compensates by feeding more duty cycle to the solenoids to get the load back where the ecu wants it. (surely I'm not telling you anything you don't know already, but I probably was not clear that this was on stock ecu boost control).

Again, these back to back runs were done on the second factory reflash with no other changes than the filter.

 

yeah I can understand why there is more boost.

but I don't understand when you say g/s is lower even though boost is higher.

how are you measuring g/s?

 

Because the drop in filter is altering the maf reading, which is the whole topic of discussion for this thread.

Logs are taken via EcuTek logger which logs just about everything. G/s is logged from the ECU along with all the other vitals like calculated load, knock retard, maf voltage, intake temp, charge air temp, etc etc.

I see what your saying now. the actual flow rate is I'm sure higher, but the original question was how do drop in filters make power, by "cheating" the maf or by allowing more flow. My point has always been that it is mainly due to "cheating" the maf, causing the ecu to register less airflow and there for you get leaner afr and little more timing. Now I have the logs to prove it, just need to attach them when I get home from work for everyone to see.

 

so the explanation is that the hf filters cheat the maf into thinking that there is less aif flowing

i have yet to see an explanation on how thats done? magically?

 

Well, I know I'm gonna get flamed for this but here it goes.

I've noticed that many factory hot wire maf sensors have a honeycomb in front of the sensor element. One can only assume the purpose of this is the "straighten" the flow or ensure smooth uniform flow across as much of the maf tube area as possible. Anything that changes the flow characteristics across the sensor element will change the reading.

If I recall correctly the X does not have a honeycomb in the maf tube (don't know for sure but I don't recall seeing one). Add to that the fact that you are changing the flow characteristics by changing the filter element and ... tada.. magic

The other possibility that comes to mind is that by replacing the "restrictive" factory filter with a less restrictive aftermarket filter you reduce any pressure drop in the maf tube, or effectivly increase the air pressure in the maf tube (less pressure drop = more pressure). This however does not make sense based on the log data I have looked at because an increase in pressure in the maf tube would translate to an increase in air density and there fore the maf would read higher than actual due to increase in mas flow.

 

I think we are essentially saying the same thing then. I do not believe the sampling area will change with a filter element.

On your other point above...I'm not sure I can relate to the "magical waves"...but I am interested in looking at my old logs as well now when I did the IC and filter tests last year. I would love to see your logs. If you don't have time to post please email them to me as I would like to look at many different areas including TP, temp and the like. A pressure differential COULD mess with the readings slightly I guess... Not really sure. I have to do some digging.

Good post!

 

You are mistaken .. the honeycomb you speak of is for KARMAN vortex generation .. The older Evos and alot of the mitsubishis run this sensor as standard .. The honecomb is used to generate vortices / eddy currents and the sensor measure these currents to determine the amount of airflow through the intake .. you can trick karman sensors like you explained through changing the airflow pattern .. on the Nissans the SQUARE mesh is to protect the sensor from foreign object damage ..

Hotwire sensor used in the current evo and subarus now measures voltage in the wire required to keep the sensor at a constant temperature (hence hotwire) .. the intake air cools the wire/sensor and the amount of voltage required to heat up the wire to a fixed temperature is the calibrated airflow across the MAS..

The most common way to trick hotwire is increase the size of the intake piping because this changes the volume that the sensor is calibrated for .. No matter how you change the airflow pattern .. the voltage required to cool the wire/sensor is going to remain the same UNLESS the flow increase or decrease ..

 

Gunzo - how would changing the diameter of the piping change the hotwire? As long as you keep the hotwire assembly the same then it shouldn't matter right?

 

yea .. it doesn't change as long as the assembly is the same ..

The thing is the current X MAS sensor is just a plug that can be taken out from the MAS piping and fitted into a larger diameter piping ..

Doing that changes the volume that is measured though .. velocity of the air required to cool the hotwire * by a fixed known area (original MAS piping) = g/s ..
By doing that .. I agree that the sensor is now tricked into measuring a 'lower' volume flow than before ..

 

ahh ok - I thought it was like the MAF from CT9A.. ie whole assembly.

if its just a plug yeah - users will need to recalibrate it

 

Exactly what i was just about to say, great post. With heated wire, or heated Grid type MAF's the PCM varies the amperage give to the heater, in order to keep it at a constant temp. More air flow= more amps needed to keep the wire at the desired temp, which tells the PCM how many Grams/sec are going into the engine.

 

Hot wire sensor still requires laminar flow to operate correctly. And yes, I do understand how hotwire sensors work. And yes I have seen many hot wire sensors that use a honeycomb to straighten the air and induce laminar flow. You are correct about the older Karmen Vortex Sensors, but that is not what I was talking about.

Yes, of course the most common way to "trick" a hot wire sensor is by increasing the size of the sample tube, but that is not the only way to do it.

So how bout you explain away these logs then maybee you can find something I missed.

Ohh, the plots are here:
https://www.evolutionm.net/forums/ev...-vs-works.html