924Board.org

[-nick]

makes sense, kind of... seems counter-intuitive to put the throttle body on the side. the distance the air has to travel is maximized between cylinders with it on the side rather than minimized if it were in the center.

but yours is identical to a 951 intake, and Porsche engineers seem to know what they're doing...

keep up the good work (nice welding job) and keep us informed.

regards,
-nick

[Cbass]

[Geddy T]

I've never worked with programming an ECU myself (although I was going to until an engine I was working on injecting was stolen out of my garage), but I have hung out at the shop (Western Washington University Vehicle Research Institute) and watched our formula team program theirs. It, too, was a Haltech E6K. Oh-and with our university research "discount" it ran upwards of $1700 USD with the wiring harness and hall effect sensor and magnets. I had one priced and was getting ready to give the go-ahead when my engine was yanked.
Anyway, what I'm getting at here is that our boys raved about the Haltech because it was so programmable, they could get anything they wanted out of it. But this is a group of guys with unlimited free access to one of three most sophisticated engine dynos and data acquisition/analysis instrumentation rooms on the west coast. Thirteen guys spent several hours a day for more than a few weeks to tune that engine to their liking.
Maybe I'm missing something here and they were just going overkill, but my understanding is that these ECUs come out of the box blank and need to be programmed to a specific engine. So you give it an estimated map and tweak from there. How do you guys plan on doing this? Are you going to rent dyno time? Are you going to plug in a laptop and guess and check until it runs correctly? I remember there were really complicated bugs to work out like start-up cycles and things like that which I couldn't imagine doing without computer readouts and a dyno.
The benefits of a sequential EFI system are in the fine-tuning--squeezing out five horses here or getting one mile per gallon better there. Otherwise I would think it wouldn't offer enough improvement over the CIS to warrant the cost (unless you're in Austrailia, apparently). I abandoned all hopes of an SEFI project once I got out of school and lost access to that engine dyno. But if you guys know a way around it, I'm getting excited again. Let me know.

[Cbass]

I'm talking about building an ECU as well. It gets programmed from the ground up.

The likeliest plan is to copy 944 chip data and tweak from there. Then, it comes with a preprogrammed setting for a 924, or 944.

I have no idea how to build a DIS though, so this should be fun

[Geddy T]

Well that's what I'm talking about. When you say "tweak," are you talking about adjusting based on how it drives--drive, adjust, drive, adjust, drive, adjust...?

[Benski]

You can actually get pretty good tuning results with an EGT (exhaust gas temperature sensor) and a wide-band lambda sensor (basically an expensive and very accurate oxygen sensor). They are expensive but much cheaper than a full blown dyno!

[Geddy T]

Ahh, fuel delivery can be done, yes. With an E6K, though, you're probably also talking ignition mapping. That's what I would think would be a lot more difficult. Thanks for the insight with the wide band O2.

[-nick]

Geddy T-
yes, the final runs have to be made on a dyno to really get the most out of the system. -and dyno time is not cheap for sure!

Cbass, unless you're running a totally stock engine (which if you're switching to efi that's doubtful) then the final tune will be different for each engine. you're not going to be able to just copy and edit a 944 ecu. not to mention you're not going to be able to read the maps directly anyway. so you're not going to have much luck selling a system with a generic map.

i'm guessing it wouldn't take too much to get a better tune over the stock CIS, but a lot to really see the full potential of a programming ecu.

-nick

[ricomartinez]

Have any of you heard of this:
http://www.bgsoflex.com/megasquirt.html

I'm not sure if it has been mentioned here before, but check this out. They also have already made large "kits". Earlier this year they contracted out to have someone supply 100+ packages with all the parts to make the ECU, costing the buyers about $100. I didn't find out in time to get on the list, so now I'm on the waiting list until they do it again.

I occasionally do a search to see how builders are doing on projects. Seems to take a few hours to build. Hell, for $100 if that's as far as I get, I'll be happy for the experience.

I figure if I do this, I'll just go to the junkyard and get a harness from a 4-cyl EFI engine. I hear those aren't too pricey now. Hopefully there will be enough collective experience to know what pieces to use. After building, the programming doesn't seem "too complex". Sure, it might be a life of trying this-and-that, but that's what these cars are about.

[Cbass]

The idea is to use 944 data as a starting point for a 924. The program will be for a 924 with stock internals. There will be control for fuel and ignition, fuel cutoff, abs/traction control, electronic boost, and anything else I can squeeze in.

The ABS will only be on electronically controlled ABS cars... like 911TT brakes

This Megasquirt looks like a good starting point. If I can build a 100 of these or so, I can give some pretty good discounts to people here. I can make a GOOD profit margin selling these things for $400 a piece.

[Geddy T]

Still, I'm with Nick. ECUs are not the playing field for guess and check. That's why companies with a dyno and a few engines can make hundreds of thousands of dollars selling performance chips. They've got the resources to test and test and test until they improve upon the system. These chips aren't doubling an engine's output. I've seen some that only add ten to fifteen horses. And that's WITH dyno runs to tweak out the most possible power.

[Cbass]

Geddy, I am talking about an investment of more than a few thousand dollars. If I can market the product, than I will take the risk. Dyno time will not be an issue, the only reason I plan on using an existing map to start with is because it is similar to the requirements of a 924 with DIS and EFI.

[Geddy T]

If that's the case, I agree with you wholeheartedly. A 944 would be a fine starting point. In fact, you could probably even theorize a pretty good starting map from nothing. Keep me posted on how things go with this project. If you get the website up, put up some pics and progress. I'm with you. In fact, in another year, I'll be back in school with access to our dyno again. I might be of service.

[Geddy T]

Talk to Alex Roy. He might have some insight into this also as it looks from his website like he's planning to program an ECU soon as well.

[Cbass]

Yes, but he's programming his for a custom turbo application. I mean a fuel map for a 924 with the distributor, but you'd have to buy your own ignition controller or something. If you had a DIS, you could control the timing with the ECU too.

The idea is to stand on the shoulders of giants(eg steal like a bastard). I recognize once I have some decent products out there, people are just going to rip them off anyways. It's all fair, just be careful about patents.