What is necessary for proper engine management?

hackish · Apr 13, 2008, 09:27 PM

I've fought this turbo issue for like 4 months now and have finally arrived at a solution. First I will outline the problems with the ralliart fuel system as it comes from the factory.

First it uses a drive by wire throttle. If the ECU sees too much air and little throttle it will immediately assume the throttle is stuck open and trip limp mode. So if you have a properly sized turbo that makes boost at part throttle this is a problem.

Second it uses an oddball MAF sensor that most piggyback manufacturers do not support. There is a workaround circuit that does work but ultimately there is no way to adjust MAF signals enough to put in the necessary amount of fuel for a turbo setup. Some people have installed 5th injector setups but they run some cylinders rich and others lean which means eventually you'll be looking for another engine.

Third, the ECU only goes out of closed loop mode at WOT. Again, if you have a turbo that makes boost before WOT it will keep subtracting fuel until it pegs at -25% and will then throw a fuel system too lean code. Depending on how you drive at this point you're probably running way too lean and if you're lucky the engine stays together. This is one of the major problems plaguing the 5th injector setups.

I have determined now that the only way to correctly fuel the engine is to install larger fuel injectors with a proper fuel management computer that can modify the injector pulsewidths and trick the factory ECU into thinking the engine is running inside of it's restrictive parameters.

I used the AEM FIC as part of the solution to accomplish this but it still suffered from the closed loop problem. I've worked closely with the AEM engineers and they have concluded that their unit is simply not compatible. You can't just fake the O2 signal because the ECU determines that your O2 sensors work by adding a little fuel and watching for the O2 sensor to respond. Being stubborn I've modified the AEM unit - taking it apart and making a number of internal modifications and developing my own microprocessor controlled piggyback to fix some of the shortcomings of their piggyback.

Finally I have a kit that works. It is in the final stages of testing now but so far it's keeping the long term and short term fuel trims in check. It's keeping the air fuel ratios in check. It's not throwing any cat inefficiency codes like other kits do. It's not throwing any fuel system lean codes like others.

So a kit I am considering would consist of 4 larger fuel injectors with plugs and adapters so it will bolt on to your fuel rail. A complete wideband/EGT gauge. A complete and custom modified piggyback ECU that will drive the larger injectors. A custom developed tune that works with all the ECU parameters.

This is a lot of hardware and it has taken a lot of work to develop.

The question I have is there anyone out there with a properly insturmented car that runs the proper AFRs under boost and has stable long term and short term fuel trims with no check engine or limp mode conditions? At most I've seen a car that appeared to run properly for about a week before deteriorating into a poorly trimmed mess.

Although I can acknowledge that this particular car is a little more "difficult" because it's making a significant amount of power and has a turbo that spools to full boost around 2500 RPM I'm interested in knowing what others have seen.

-Michael

Silverbolth · Apr 13, 2008, 11:40 PM

Quote:

Originally Posted by hackish

I used the AEM FIC as part of the solution to accomplish this but it still suffered from the closed loop problem. I've worked closely with the AEM engineers and they have concluded that their unit is simply not compatible. You can't just fake the O2 signal because the ECU determines that your O2 sensors work by adding a little fuel and watching for the O2 sensor to respond. Being stubborn I've modified the AEM unit - taking it apart and making a number of internal modifications and developing my own microprocessor controlled piggyback to fix some of the shortcomings of their piggyback.

Damn... if you were near me, I would buy you a

... way to go

k3lv · Apr 14, 2008, 01:26 AM

^ You are near Hackish. I believe Hackish resides in Ontario, Canada.

I would be interested in a better management system along with oversized injector setup compared to the 5th injector. However it would have to flow with the rest of the RRM kit

hackish · Apr 14, 2008, 07:16 AM

I believe the RRM kit uses a 5th injector. You'd just have to remove that injector and their piggybacks and install my solution.

On another point I thought I'd list off the piggies I've tried and tried to work with without success.

Unichip - unable to control the larger injectors, O2 problems, incompatible MAF.

Split Second - unable to control the larger injectors, unable to fix the O2 problem.

AEM - O2 problem - there is a workaround but it conflicts with the MAF workaround I found

Haltech Interceptor - unable to control the larger injectors, unable to fix the O2 problem, incompatible MAF

-Michael

**otter** · Apr 14, 2008, 07:48 AM

This wouldn't be an ideal solution (expensive) but what about going stand-alone and switching to another MAF that's easier to deal with, or eliminating it all together?

hackish · Apr 14, 2008, 08:19 AM

Standalone plus replace the DBW throttle is an option but they're tough to get a perfect startup in all conditions - who wants a car that won't start in the cold. It can take many hours to get a stable idle in all conditions. Often it requires a complete custom harness so forget about going back to stock if you don't like it. Standalones are best suited to race cars in my opinion.

Another MAF is still going to need signal translation if it's going to be running on the stock computer. The trouble with the MAF is more related to the restrictions within the computer.

-Michael

**otter** · Apr 14, 2008, 08:46 AM

FWIW, some of these MAF issues may be cleared up in the near future. I believe all the 08+ Lancers/Evos now use the same MAF that we have.

hackish · Apr 14, 2008, 08:58 AM

The problem is not really the MAF. Electrically the fix is fairly easy. The MAF will easily support 300hp of air.

Instead the problem lies in the range away from normal that the factory ECU permits. As soon as you go beyond that range it ****s the bed and throws a MAF or throttle actuator error because it's getting too much air.

-Michael

RALLIART15 · Apr 14, 2008, 09:09 AM

You guys are still over my head but I'm trying.

Did you modify a existing piggy or build one completely from scratch?

samj601 · Apr 14, 2008, 09:32 AM

^he tried modifying the aem fic to work with our cars but still has some problems with the maf

hackish · Apr 14, 2008, 10:49 AM

Quote:

Originally Posted by RALLIART15

You guys are still over my head but I'm trying.

Did you modify a existing piggy or build one completely from scratch?

I have modified the FIC but it doesn't have sufficient features to complete the engine trickery so I have designed a prototype of a daughtercard for the FIC - it fits inside and just adds an additional plug to the unit.

-Michael

BlackRaven · Apr 14, 2008, 11:05 AM

Michael, I'm currently having a custom turbo kit made for my car w/ the WRX 440 injectors, using the FIC, boost to be under 10psi, probably in the 7psi neighborhood.

Two things:
1) In your experience, if I don't have somthing like your add-on piece, what problems am I going to run into using the FIC? Will I just need to unplug the battery every week to reset the ECU, and re-up a fuel map (per your comment, "At most I've seen a car that appeared to run properly for about a week before deteriorating into a poorly trimmed mess.")?
2) I will gladly be a guiney pig if you want to try your unit in another car.

Edit: Heck, I'll even pay! PM me if you want to set something up.

hackish · Apr 14, 2008, 12:25 PM

The biggest problem you will be faced with is the closed loop issue. Some people have tried to circumvent this by simulating a narrowband O2 sensor as the engine is running stoich - depending on how you drive this may last up to a week before the fuel trims become significant. If you're on boost a lot it may only take 15 minutes before your engine is running dangerously lean. I have heard from a tuner out west who has tried a number of times to re-tune 5th injector setups and he has said that every car he sees is running dangerously lean. It is very very difficult to get this thing right.

-Michael

RalliartN · Apr 14, 2008, 04:18 PM

Michael, do you know if aftermarket supercharger kits have the same fueling issue as the turbo kits already out there for the Mitsu, and if so, would your new injector/ecu package be a fix for that?

btw, looking forward to the turbo and other power packages you guys are developing for the Ralliart. Your work is fascinating.

OttRalliart05 · Apr 14, 2008, 07:19 PM

then I would like to know how does the 5th set up lean out ROCK stated in a different post that the does not lean out how did they get by the MAF TPS MAP and the O2 issues???......

Quote:

Originally Posted by hackish

The biggest problem you will be faced with is the closed loop issue. Some people have tried to circumvent this by simulating a narrowband O2 sensor as the engine is running stoich - depending on how you drive this may last up to a week before the fuel trims become significant. If you're on boost a lot it may only take 15 minutes before your engine is running dangerously lean. I have heard from a tuner out west who has tried a number of times to re-tune 5th injector setups and he has said that every car he sees is running dangerously lean. It is very very difficult to get this thing right.

-Michael