If you use a larger MAF housing diameter without reprogramming or otherwise changing the signal, the DME won't be calculating air volume correctly. The car will run leaner than the DME thinks it is.

I do tune the cars that I have used this setup in, but I also believe that the stock ECU (no flash or tune) can easily adapt for this and headers.

This really is only slightly different from installing a cone filter setup.

Thoughts?

I have doubts. The DME uses the MAF to measure air speed. It needs to convert that speed to volume flow so it knows how much air the engine is injesting. The conversion calculation has to include the diameter of the MAF housing.

Increase the MAF housing without telling the DME and it thinks the airflow is lower than it really is. Same speed + larger area = more flow. And not a little difference, it's a 23% increase in area! Huge! I would be really surprised if there was some way the DME can compensate to that change.

It means Porsche would have had to allow for housing changes while programming and then figure out a way to double check the sensor. I don't see a system that can make the correction. You know how slow O2 sensors are, they aren't fast enough to make a meaningful correction. They are mostly used for steady state (cruising) corrections. And there is no MAP sensor to verify flow with a pressure measurement.

I bet when you do a tune, the car is running very lean to start, yes?

Thinking about this :)

They are not lean to start, but I know why they are not lean.

Funny part? I run the cable cars with the MAF disconnected pretty regularly. The base map in the ecu is pretty damn good..

Still thinking. Everything you posted makes perfect sense. I understand what we are doing (I'm not using this on anything smaller than a 3.2) but I will in the future.

The base map with the MAF disconnected probably runs fuel metering off the throttle position sensor. So it's probably fine as long as you don't mess with the cylinder filling efficiency of the engine (very free flowing exhaust and intake) and you don't change the size of the throttle body. If you do those, it'll probably go lean again. But the non-MAF map is probably conservative to begin with so it might bring the fuel level to about right.

Correct. TPS metering. The throttle response is not as crisp when I have them disconnected.

I have seen multiple cars adapt upwards of 20hp with no computer change, knowing that we changed the plenum (after MAF changes)

Typing out loud

With safe AFR's?

I'll install one of these on a stock 3.2 engine with headers with a before and after A/F readings. I have a gutted stripped 3.2 race car in the shop that will be easy to install one on. I'm pulling the engine this week anyway!


B

i've been looking into tuning recently and here is what i think i know:

dme looks at (a) maf signal (amount of air going into the engine) and (b) rpm, to determine how much fuel to deliver via the injectors. this is done using a 3D, 3-axis map (x=rpm, y=quantity of air, z=amount of fuel required). these are the maps that get modified by tuners to increase performance.

dme then looks at exhaust o2 sensors to see how things went during combustion. depending on the amount of oxygen in the exhaust, the dme either adds or removed fuel by lengthening or shortening the injector pulse.

this information is put into a short term fuel trim map (stft) as a % modifier of the primary map.

if the stft stays stable (ie, over about 50 kms of driving) the data is transferred to the long term fuel trim map (ltft). this is your dme 'learning'.

if you find that your ltft is staying at a constant value (ie, +5% due to intake modifications) tuners can go to the maf calibration map and calibrate it so that your ltft zeros out. you may wish to do this to make room for other modifications, as max ltft is 25%.

so, if you put a 3.5" diameter maf housing on a car that needs a 3" housing, you are looking at a maf that is reading 36% out (pie are square, right?). dme can't adapt to such a significant change. worse, your engine will be getting more air than it thinks it is and run lean - hard on engines (no cooling). you can do it, but a dme remap is required.

however, it also shows that those who state their intake products require an expensive remap (which they also sell) to realise full benefit are wrong, unless airflow is increased by more than 25% (not likely).

That should be interesting. Hopefully it's interesting, anyway.

Porsche runs these super lean anyway, but yes, more than safe.

King,

you pretty much nailed it. nearly verbatim to what Porsche has released in factory bulletins!

I'm all up for learning :) Now, sharing what we have learned? Not always :)

That's never a problem for me, I always know all the answers :D

Brad, do you think I'm safe to run the stock tune with the 987 box and stock throttle body? I plan to upgrade the TB and plenum at a later date, and probably headers, then tune it to match.

Let me get it installed in a stock Boxster-S with headers. I'll use a stock CaymanS throttle body on the 3.2 plenum and report back with numbers.

It all makes sense. But I wonder if the DME just makes small corrections to one load/speed point on the map at a time or if it tries to smooth the map around the point of correction. The O2 sensor isn't fast enough to relay good data while the engine is dynamically changing, like changing loads or RPM. It needs the engine to be steady for a while to know what the conditions are at the intake relative to the exhaust.

Although, I suppose if they charactorized the lag, maybe they've figured out how to use that data too. Maybe.

BTW, there's a really good book on the principles of engine tuning by Jeff Hartman. Also, the manuals for the programmable EMS computers are mostly available online. They give a good overview as well. But nothing beats doing an engine yourself. I learned more with my Haltech than I could ever have by reading about it.

I managed to get my hands on an actual Bosch .pdf explaining exactly how the 5.X and 7.X work. I should dig it up and throw it on a server for download. It is the most comprehensive break down I have ever seen published (I just can't regurgitate verbatim)


B