Eric_Oz_S2

http://cgi.ebay.com/ebaymotors/Porsc...#ht_1071wt_941

Anyone know if these can be made to work on an S2 with appropriate signal conversion. I could quite easily develop the appropriate transfer function using an interceptor between the MAF and ECU.

The purpose is to be able to run with a supercharger.

Mark944na86

The DME looks for a signal between 0-5V. The trick, of course, is getting the correct voltage to match the value the DME expects for that air flow using the AFM. But if you can do an arbitrary translation using an interceptor, what could be the problem? Sorry if I'm being dense.

Eric_Oz_S2

As it is for a 4.5l turbo engine, it would be rated for a much higher airflow. Not sure if the range/sensitivity would be appropriate for the S2.

fast951

When supercharging the 944S2, there are obstacles in the ECU that you must get around way before you get to the VAF to MAF conversion.
Using a MAF then converting the signal to look like the stock VAF will gain you nothing (except for the free flow becuase of the larger ID of the MAF). However you still have the limitations of the factory VAF.
If you are boosting the S2, and doing all of the ECU work yourself, you will have better luck running with the VAF to start with. Once you get around the road blocks put in place by the Bosch/Porsche engineers , you'll have a better luck converting to a MAF. Of course this means rewriting the code to use the MAF.

Eric_Oz_S2

My main concern with the stock AFM is trying to find a location for it because of its size. A MAF is more compact and easier to fit in with the rest of the plumbing. If anyone can provide advice of where to fit the AFM with a supercharger installation I'm all ears.

Mark944na86

I can't see this being a problem -- it just means you'll be using ~2/3 of the range compared to the 4.5L application (assuming the same level of boost, of course.) Since you will be running closed loop, I can't imagine this slight degradation in precision being a problem. Much better to have some additional headroom rather than the other way around, of course. Even the 951 AFM maxes out at about 300HP, apparently.

The hardest part would be getting the translation right in your "interceptor". I'd be interested in your approach that makes this "easy". I assume you'll be programming a microcontroller? What platform were you considering?

I've done some embedded programming over the years on Motorola, Intel and MIPS, so I'd be interested to hear what you have in mind.

Personally, I've been considering a Link Atom set-up for a forced induction set-up further down the line, and use the built in MAP sensor. Even more compact than MAF! (and cheaper, too). MAF is old hat.

Eric_Oz_S2

I would run the MAF upstream of the AFM and record voltages simultaneously from the 2 meters. This can either be recorded live while driving (using a passenger) or through an oscilloscope. It should then be possible to fit a curve using Excel.

A simple voltage interceptor http://www.jaycar.com.au/productView...T&SUBCATID=347
is all that is required to convert MAF voltage to "AFM Voltage".

I still like the Ostrich 2.0 for tuning. Not sure how that compares to the Link Atom, but I suspect it would require less wiring in.

I'm going to run next year in E class, so I'm looking at developing a system over the next year for installation towards end of next year. Still not sure what class I want to get myslelf into. 230-240hp would be top of E (very low boost - perhaps a first stage for testing - not sure how a low boost would work with parasitic drag. This raises an interesting question - could you control boost in a supercharger so you have higher boost at lower revs, and lower boost at higher revs? Perhaps a restrictor plate in the inlet?). 260-270hp would be top of D class, 300-310hp would be top of C class.

Mark944na86

Yes. that looks like it will work. Having a little purpose built kit certainly simplify things. Have you read the magazine article for the project? It would interesting to find out what the "hand controller for programming" actually is, and how much programmability it has.

Chalk and cheese, really -- from what I gather the Ostrich emulates the eprom with the A/F maps, allowing you to tune by writing to the Ostrich, and then burning or flashing a chip with the final maps. The Link is a full ECU that replaces the stock DME entirely. The thing I like about it the fact that it is designed to use the stock sensors. It's also reasonably inexpensive (about $1000, although the "knock block" which it seems you need to buy to use your knock sensors is another $600 or so -- ugh). Built in MAP sensor. Auto tuning to target A/F ratios with the software. Multidimensional maps for sophisticated tuning if desired (add intake manifold temps, for example). And so on. Just sounds like more fun.

Smaller roots type blower with a BOV set at low boost? I was looking at rebuilt Eaton M90s on ebay... they can be had for not much $$$. Although a fair bit of custom fabrication required, of course.