legacycentral bbs

Just something I thought up. If we can get the n/a injectors high flowed we still need some way to control the extra fuel.
Here's my crazy idea. I would tune the car at full power wot to a good afr by adjusting the fuel pressure. Then use a adjustable ramp ala FCD to adjust the air meter readings so it will idle well and have a good afr at crusing. Not the most high tech thing to do but with a wbO2 I bet it would could work.

Could use the same idea to run a bigger MAF too. A MAF is a simple voltage ramp output right?

Just a crazy idea to get things going

I'm not sure I understand -- are you basically just describing a simple method of ramping down airflow readings in order to kind of cancel out the effect of the larger injectors?

What does fuel pressure have to do with that?

Let's start by looking at these graphs that I pulled out of my ECU:

http://www.graphics.cornell.edu/~v/pics.cgi/aa694.gif

There are a couple of things to note... One is that the measured airflow plays a large role in the ECU's load computations. Computer load plays a large role in computing ignition advance.

Look at the two spark maps. At low load, the minimum advance is much higher than the maximum advance at high load. Also note that the difference between base and maximum advance is almost negligible at low load. The ECU only actually gets smart about controlling timing at higher loads. So if you make the ECU think there's less air than there really is, you'll probably end up with more ignition advance than you'd ideally want, and you'd have much less flexibility. You could probably get it to work passably by making the mixtures much richer to stave off detonation, like some people do with S-AFCs, though.

Now look at the MAF sensor transfer function (it's roughly quadratic, by the way). The units of the "airflow" axis are 1/256th of a gram per second. The sensor can't actually read any higher than about 200 grams per second, irrespective of any signal conditioning afterwards. Enlarging the sensor, or using a different one, wouldn't have the same limitation though.

Ok you said somewhere else I think that at the limit of the MAF the ecu just throws the injectors to full duty, so adjusting the fuel pressure down so you get a good afr at max duty was the goal.

If going to a bigger MAF and after getting the ecu to read it right, does it even have any mapping above the 200 grams per second to use?

hope that helps

Oh, okay, I see what you're saying about using fuel pressure to tune.

But you really shouldn't run the injectors at 100% duty cycle. They'll overheat and die early. Our stock injectors do have some advantage because they're side-fed so the fuel flowing through them helps keep them cool, but it's still not something you want to design your engine management on...

The ECU can read up to about 5 volts from the MAF sensor, and no more. That 5 volts is about 200 grams per second. So it doesn't really mean anything to discuss mapping above that range.

So if you got a bigger MAF and ramped it down some, then just increased the fuel pressure to make up for the extra air?

And as for timing without the tiny stock turbo and with a good IC I think we could use some advance.

Yeah, a bigger MAF and bigger injectors or a bigger MAF and higher fuel pressure might cancel each other out and give you increased range, modulo the timing concerns.

By the time you spend the money on the bigger maf, higher fuel pressure, injectors, etc.......you could have spent that money to properly tune the car.

Just my opinion. Do it right, do it once.

Hey don't mess with Cheap Ass Motersports
I bet there is a combo of JY parts that with the research would do just what we want for cheap.
I like the once right thing but what if you had told that to the guy who figured out the turbo bleeder valve?

I'd say a bleeder-type boost controller is more "right" than what you're proposing, since it does its job without interfering with anything else.

But I'm very eager to see if you come up with something good.

Well I'm just throwing around some ideas right now. So now that I know you have gotten into the ecu and had a look around do you think you'll be able to make any changes? From what I can tell Subaru left a slot for a chip on thier ecu's untill 97 (I think) when they made ecu's that could be reflashed. Do you know if our ecu has that slot?

The other crazy idea I'm looking into is getting a early WRX or Legacy GT ecu to work on our cars. I need to get some part #'s off the Jspec cars to see how many of the sensors are the same. So far no help from the RSLib board. Even If it will just run there are stand alone ecu's with base maps that will plug right in. I don't want to spend 1k plus on an ecu then how many hours to wire it in and not be able to even get it to start.
And all for the power I could get with the same simple mods we have, on a JDM Legacy-Impreza.

Our ECU has a different slot, since it's a very different computer. But, yes, it has a place to plug in a replacement ROM. I started a project to reverse engineer the ECU. At some point when I'm done with school I'll be going back to that project.

Pre-97 non-US turbo ECUs mostly fit in place of ours... The only wiring you really have to do is to take care of the different ignition setups (coil-on-plug vs. waste spark). I think it can be as simple as adding a few wires and diodes, but I haven't tested it.

A standalone meant for Pre-97 non-US Subaru turbos should wire into our cars in the same way as they do on their intended applications, aside from the ignition issue.

The sensors are pretty much the same as ours... you need a JECS MAF sensor though (the black plastic-housing one).

I don't know how easy it would be to recalibrate the EJ20T ECU to understand that it's on a 2.2-liter engine... it might be easy or it might be impractical. Or it might not matter. I haven't played that much with the JECS ECUs since my car doesn't run on one.

Matt Monson knows more than anyone else here probably about wiring EJ20G ECUs.

With the stock turbo, you aren't going to gain much from tuning the fuel. The stock turbo runs out of puff well before you can put out any power. You are only making max boost until 4000 RPM or so, then it falls on it's face. It's just to small to do any damage in the teens of boost.

FWIW, it took 16psi on the stock turbo to max out the MAF and cause enough of a strain on the stock fuel system to run lean as the fuel pump couldn't keep up. But the amount of time it was at 16psi was almost none. I wouldn't get overboost fuel cut for that reason, unless I was on the highway pulling in a high gear.

With a larger turbo, however (TD05), the MAF maxes out around 10psi.

The stock turbo is not efficient enough to max the MAF out for more than a few seconds at a time. You will get overboost fuel cut, but never once did I notice the MAF reading 5v on the scantool with the stock turbo.

SO, that said.... Bigger injectors wouldn't do much with the stock turbo. You would never really take advantage of them.

One question I have is why you would go with larger injectors, and then tune down fuel pressure to match the air coming in. Why not just run the stock injectors at 90% IDC? What advantage does running a larger injector at 100%IDC at a low fuel pressure have over running the stock system at a normal fuel pressure at 90% IDC?

Currently, I am running the stock system at 100% IDC at 10psi+ on the TDO5-16G with a Walbro 255lph fuel pump. I don't think I would gain ANYTHING by running larger injectors and a smaller fuel pump.

-K- doesn't have the stock turbo. He put a bigger compressor on his stock turbo.

About the fuel pressure thing, that confused me at first too. Read it again... what he meant was he'd put big injectors on, but if they're too big he could dial it down by lowering fuel pressure.

With the stock setup, 90% duty cycle isn't really a choice you have.

ooohhhh

Yea, I wish we could run 90% IDC.

It's OK. I don't think I will be on the stock fuel system for much longer anyway.

vrg3 wrote:Our ECU has a different slot, since it's a very different computer. But, yes, it has a place to plug in a replacement ROM. I started a project to reverse engineer the ECU. At some point when I'm done with school I'll be going back to that project.

Pre-97 non-US turbo ECUs mostly fit in place of ours... The only wiring you really have to do is to take care of the different ignition setups (coil-on-plug vs. waste spark). I think it can be as simple as adding a few wires and diodes, but I haven't tested it.

A standalone meant for Pre-97 non-US Subaru turbos should wire into our cars in the same way as they do on their intended applications, aside from the ignition issue.

The sensors are pretty much the same as ours... you need a JECS MAF sensor though (the black plastic-housing one).

I don't know how easy it would be to recalibrate the EJ20T ECU to understand that it's on a 2.2-liter engine... it might be easy or it might be impractical. Or it might not matter. I haven't played that much with the JECS ECUs since my car doesn't run on one.

Matt Monson knows more than anyone else here probably about wiring EJ20G ECUs.

I don't have anything to add to this really. The early WRX and USDM Legacy Turbo wiring and sensors are pretty close to identical. You would only need to change/add 4 wires to run the coil on plug ignition and install the individual coilpacks onto your heads. But I don't think they would work on the SOHC heads. So, what you really would have to do is rewire your wasted spark ignition to work with the ECU thinking it wasn't wasted. Kind of tricky me thinks...

Matt, are you sure the JDM Legacy turbo engine and ECU could run with the USDM wire set? I have an JDM engine on my 92 Turbo but the ECU can't just plug and play since as you metioned it missing the 4 ignition wire... I have the coil pad too. Now I only run with the USDM ECU to my JDM engine.
Does anyone try or already did the modify with the JDM ECU , as far as I know most of people who did the JDM Legacy turbo engine swap with out using the JDM ECU.
Please help

Ben

I'm not worried about the spark because I know that some of the RSLib guys run a coil pack as an upgrade.
MAF and fuel injectors are my biggest concern. Our injectors are high imp and I think the Liberty uses low imp. This is from what I could get out of a few searches on the RSLib board. Both injectors are 380cc but low imp could be why they can make a lot more power than us.
Matt do you have any JDM part #'s for things like the MAF injectors etc.

I was pretty sure non-US BC/BF injectors were high-impedance... I might be wrong though. But in any case, low impedance injectors can't support "a lot" more power than high impedance injectors. They can support a little more because they open and shut faster, but not a lot.

I'm pretty sure non-US BC/BF turbos use the same MAF sensor our JECS-based non-turbos do.

Well the nice thing about low impedance injectors is you can actualy run them up to the rated flow. Because they are low impedance they don't overheat and act up if run over 85% duty cycle. So 85% to 100% is a good bit. Still need to find out if they are the the same as ours.