T-leg Piggyback ECU R&D

[93forestpearl]

Huh? PP6? What are you guys refering to?

[bmxpunk]

the Rallitek perfet power smt 6

[Aerotech]

What happened to the legacEBC project? Anyone have info?

[free5ty1e]

I'm back - partially at least. The LegacEBC is not dead - I have just recently been able to begin working on it again. It's gotten much better, but is still not good enough for production.

Right now, lab space is extremely limited. This residence is temporary until March or April, when the snow season ends in northern Nevada and I can begin construction there. Then I plan to really get into detailed engine management with my two Scoobs, and of course offer discounted products to members of Legacy Central.

In the meantime, I'll do the best I can. I will post more updated info and some photos when I get a chance.

[IronMonkeyL255]

Any updates?

I'm beginning to look into Engine Management, and if I can get one built for our cars, it'd make a whole lot simpler.

[free5ty1e]

I suppose I could post some updated information here, although there is little to report; right now top priority is packing for the move to NV and tying up all the loose ends here in FL.

The old prototypes were functional and neat to play with and watch (lots of readouts to look at and cycle through), but I've opted for a full re-design for a few reasons. One is that I now have more than one set of ECU connectors to work with and actually get the plug-n-play installation happening. Another is that unless I went to the ECU connectors (instead of at the sensors, which is what I did last time), the signals were rather unstable and susceptible to large amounts of noise. This resulted in a final boost level that oscillated just around the set target and was not stable enough for a production unit.

The basic design is going to a very modular and upgradable layout, with one control board at the ECU and a display unit for the dash (the 90 degree board/LCD worked out nicely and is very easy to mount). The control board will have slots for modules such as Boost Control, Engine Management, and Remote Start/Locks/Alarm/Turbo Timer so far, with room for future expansion. Each individual unit (control board, display board, modules) will be outfitted with a quick-programming connector to allow firmware updates with a PSoC Mini-programmer (very inexpensive burner, $35 from Cypress.com IIRC).

The Boost Control module will be first, and then I will begin work on the Engine Management module. This will, in effect, create the full LegaCU. The Remote Start/Locks/Alarm/Turbo Timer module is something I've wanted to do for a while now too; it will be a fully programmable active-disabling alarm system, with a high-range wireless remote system for starting, engine kill, door locks, alarm functions, and a remote trunk unlock (once I figure out a way to implement one on the Legacy's manual trunk).

More to come once I have a shop! I can't wait....

Edit: It will be at least a month or two before I'm out in NV and then I have to put up a building... it might be a few months before I've actually got anything to sell. I wish I could have gotten started out there sooner... things don't always go as planned.

[stipro]

Well I still am very interested. Cant wait for the real deal . Good to see you will be back in business!!

[IronMonkeyL255]

For the trunk, It looks like it would be as simple as installing an actuator to yank the rod (god that sounds dirty...) for the trunk latch. There is plenty of room to mount it, IMHO.

[free5ty1e]

That's what I thought too, either put one by the trunk latch itself or come up with something to make use of the existing cable system... can't wait to be able to work on this all again.

[gt2.5turbo]

ya i def would like to pick one of these up keep us updated!!!

[evolutionmovement]

How about active aero with the ECU triggering them at a certain speed? Probably better manually, though.

[free5ty1e]

Active aero...? You mean a Porsche Carrerra-ish high-speed wing? That'd be hilarious. Have it raise at something ridiculously low like 50MPH so people on the highway can be astounded and confused

I bet I could make that happen if there's enough interest.

Oh and of course, the control board would have an RS232 (and probably a USB) port for laptop interfacing (internal log download, external logging, tuning/config)... and the Subaru Select Monitor as well so all the neat functions of Vikash's scan tool will be available at any time. The user will be able to connect the Test and the Read Memory connectors at will through the menus.

[Legacy777]

I'm going to throw my two cents in here regarding this project and dealing with previous design projects.

Let's learn to crawl before we learn to break dance.

Get the basic boost controller working, and then enginemangement. After that...work on the like tweaks and extras.

I could care less about all the other crap. I just want something that controls the A/F ratio, timing, and boost.

I too am at the point I need something, and need something that works well....if we can make something that works with the stock ECU....even better.

[free5ty1e]

Of course; I'm only going to develop the boost controller portion first, that's all I've been working on when I've had the chance to.

Didn't I say that? Boost controller, then engine management, then the bells and whistles... I'm not going to waste time on the remote trunk release when I've finally got the ability to do so; I'm going to replace this damn MBC I've been stuck with that never delivers solid results.

However, by thinking out the potential functionality that I ultimately want to achieve, I am able to lay out a far more effective base structure for this string of products. Believe me, it's much easier to develop on a solid base platform - one can focus on the functionality and not how to workaround some limitation in the design.

I have been developing embedded products for a number of years now; This project is my baby, I want to put my experience to work to create an optimum environment for the ultimate piggyback car product, one step at a time.

It sure is frustrating without a shop though, and even more so now that everything must be packed for the move - more delays. Once I've got a shop again though.....

[Josh Long]

Wow that was quite a read.

If I lived anywhere near you I would come help you move so you could get working on this sooner.

Add me to the list of buyers.

Josh

[206er]

will there be any way to run this on NA cars for a badass motor with boost when finished? sorry for the stupid question but all this electronics stuff is pretty difficult for me.

[free5ty1e]

Sure... will be able to control any variety of turbocharged setups. If you're not using a turbo ECU it might not be as easy to install and you won't get as many features in the end but the electronic boost control will be universal.

Should only be a week or two now before we're able to start the cross-country journey. If we make it there in one piece, a shop will be forthcoming shortly afterwards.

[Josh Long]

I have been thinking about this quite a bit since last night.

I would suggest not really worrying about the knock sensor feed back. If you have a realtime timing image that is coming out of the ECU you can see when it is being retarded for knock. Once you get into solid boost your timing should be around 23ish degrees, if you can log timing even after your manipulation you should be able to view any occurance of knock by pulled timing, being less than this value, whatever it is for the turbo leg ECU.

So if you can log timing, engine speed, wide band O2, and injector duty cycle, you will be on the way to having an awesome tuneing platform.

Another awesome feature would be a user defined boost based switched signal. This would allow you to switch water injection or nitrous.

Check this page out for some great info on tuning dased on DSM's. I have used this guide with great results on my Galant VR4, with a datalogger and an SAFC. I believe a big shout out goes to Kyle Tarry for this information.

http://users.wpi.edu/~ktarry/dsmtech/tuning.html

Thanks

Josh

[IronMonkeyL255]

+1 for the user defined boost-based switched signal.

If not, here are some schematics for one.

edited for speeling

[free5ty1e]

Thanks guys, keep the suggestions coming. Water injection / nitrous controller is a good idea, and a natural for the LegacEBC --- it will be a no-brainer to implement on top of the base system. I'll have a menu for selecting the boost trip points, and even whether the outputs should be active high or active low (individually, even).

Here's a question, how many controllable switched outputs would be useful? I've got plenty of I/O available to me... so just two? Would there be any situations where three or four would help?

And I already planned on full logging capabilities - both onboard and laptop (for either downloading onboard logs or realtime logging)... that piggyback tuning guide will be very helpful in setting this all up right.

[Josh Long]

The only other thing that sounds really useful would be making the wide band O2 have a faux narrow band output, allowing you to only run one o2 sensor.

I think you guys covered a launch control setup right? In the DSM world they use a clutch switch to activate the multi rev limit.

[free5ty1e]

Simulated narrowband output and wideband support are definately in the plans.

Launch control setup? Please to elaborate We've discussed anti-lag before, and it's something I will probably implement in the end but not actually use myself just yet...but I think you're referring to a different feature.

[gt2.5turbo]

i believe he is refering to a studder box setup which would be sick

[Josh Long]

Exactly,

check out http://www.dsmchips.com/ this has most of the useful features that people like when upgrading.

Taken from Jeffs above site:

The 3-step gives you 3 different rev limits, depending upon the postion of the clutch pedal, and whether or not the car is moving. If the clutch is NOT pushed in, you will always get the standard, highest rev limit. If the car is NOT moving when the clutch is pushed in, you will get the lowest, launch rpm, to build boost off the line. Once the car is moving, and the clutch is pushed in, you get a third rev limit, sometimes called the "no-lift-to-shift" or "flat-shifting" rev limiter. This rev limit allows you to keep the gas pedal pressed to the floor in between shifts, so the turbo stays spooled. The use of this feature is recommended ONLY for guys with REALLY good transmissions. But, PLEASE NOTE that even if you order this feature (6000 rpm recommended) you do NOT have to use it. The car will shift and run normally, as long as you do not keep the gas pedal flat to the floor between shifts.

Thanks for taking on this project, and I hope the move goes smoothly.

Josh

[THAWA]

So yeah, info about the connector stuff. Someone doing a 1JZ swap made a little how-to and more info on the ECU connectors: http://www.autosoftsys.com/supra/ecuext/

Very nice.