Greddy Piggyback computer

[dzx]

Does anybody have an opinion on this unit?
http://cgi.ebay.com/ebaymotors/ws/eBayI ... 73582&rd=1

[BAC5.2]

The E-manage is super sweet, but for some reason, our ECU will learn around anything it'll tell.

I wish it would work, because it's the most complete piggy-back system there is.

[TDGKORN]

The E-manage is super sweet, but for some reason, our ECU will learn around anything it'll tell.

I wish it would work, because it's the most complete piggy-back system there is.

Are you sure the e-manage will not work in our cars?
Not doubting you just wondering, because this is the system I am planning on buying in the near future.

Anyone else have experience with the system working or not working on the turbo legacys?

[BAC5.2]

It's been generally established that no piggy back units will work with our cars. I don't know why, and no one can really give a straight answer.

I, personally, think that a likely reason is that the ECU sees 10.xx psi of boost with the FCD in place. The computer reads this, and spits fuel accordingly. Now if you have a piggy back unit telling the ECU to dump more, or less, fuel to the system, it's going to compare that value with the boost reading, and get a conflict of interest.

The ECU will adjust itself to the proper amount of fuel required for 10.xxpsi of boost.

[TDGKORN]

Just for arguments sake wouldn't this be a problem for every car not just ours.
Or is our ecu the super genius of ecu's.

I have heard of the less complex fuel controllers having problems (s-afc), but I was under the impression that the e-manage could work around the ecu's.

[BAC5.2]

Considering the overengineering in other aspects of the car (block, tranny, rear brakes, just about all the important bits), it's not surprising that our ECU has so many different fail-safes and fall-backs.

Most ECU's don't mind being lied to. Ours is fairly adaptave. That's why you run one day, and it'll feel awesome. You run the next day, and it'll feel like shit. Our ECU tries so hard to maintain consistency of tune (keeping multiple parameters in check with eachother), that it learns around any foreign introduction (like that of a false signal from a fuel controller).

I'll give an example for example's sake.

Say your IAC valve is acting up. The ECU jumps to a fail safe mode, and monitors several parameters to protect the engine from harm. It monitors the throttle position, engine speed, and vehicle speed. If these things change to rapidly, the ECU will cut fuel because it infer's that there is something wrong. THAT is why, that even though my fuel cut defender is in proper working order (the ECU only sees 10.xxpsi) when I try to run 15psi, the engine cuts fuel because of my failing IAC (it sees that engine speed and vehicle speed are moving disproportionately to the throttle speed sensor, and the ECU cuts fuel).

[dzx]

So kelly's ecu is probably the best choice even if its a little steep?

[TDGKORN]

BAC5.2,
You make some good points.

I just got an e-mail back from a guy who has set up the e-manage in dozens of cars including subarus(not the turbo legacy). He said since we are OBD 1 the e-manage is less prone to be "learned" around. He is just speaking in general terms so I don't have any specific examples for you.

Maybe I'll have to be the first one on the bbs to try it. Maybe in a few months I'll have the money to buy the rest of system.
Unless people want to sent me some money for research purposes.

[BAC5.2]

I'd like to know how it works out, as $1900 for Kelley's bits (although undoubtedly worth it) is a bit steep for my wallet.

I don't know about the validity of the "OBD1 not learning as well" because it's to harsh a generalization.

No two OBD1 cars are the same. What works in a Honda, might not work in a Toyota. Or Subaru.

It would be nice to have a seperate, programable chip, that receives all signals of what cuts fuel under what conditions, as well as metering what fuel trims are based on.

VRG3 has made headway with his FCD, but actually cracking the code is a bit more difficult.

If the proper wires could be interecepted, then altered (without the need of a laptop constantly running fuel control programs), it would seem that things like fuel control could be controlled (synthetically) by the new bit. The new "chip" would just take input signals that would normally go to the ECU, and alter them and feed them back out.

Unfortunately, since our ECU is so fail-safe, it seems unlikely that it would be happy about this. If all of the sudden, it stops seeing signals of X variable, or if it sees X-variable being disproportionate to Y variable, then it will cause a fuss. If the secondary "chip" can deal with that, that'd be cool, but who knows WHAT would be going on if it didn't. Fuel cut, timing retardation, anything.

The only real solution that would be easy is remapping the stock ECU. The problem with that is finding the pinouts. Since we don't share ECU pinouts with the JDM EJ20G (for which I CAN get the pinout diagrams), finding them is difficult. ESPICALLY since the company who made our ECU was sold off, and files "deleted" in spring cleaning (or some such. VRG3 was telling me a sad tale about how the pinouts are no longer existant for us).

Tracing out the pinout's could probably be done, but that's a LONG and TEDIOUS task.

If we could crack the stock ECU and alter maps and such, then everything would be GREAT. Simply reflashing the stock ECU leaves everything good, and prevents a wiring fiasco when trying to switch to a different standalone engine management system.

Until the stock ECU is cracked, a unit by Autronic or Kelley is the best bet. It's unfortunate that it's a $2000 buy in .

[dzx]

Does rssitiboy have any type of engine management?

[BAC5.2]

S-AFC I think?

Remember, he's not in the US. His engine management is significantly different than ours. Not to mention his engine is COMPLETELY different than ours (EJ20G vs. EJ22T).