CovertRussian's 05 LGT Build Thread

[covertrussian]

I'll preface with, I'm on stock and don't personally need one yet, I'm running out of injectors on cold days.

 

DW65c is used a lot, it's a drop in option with no modifications. Walbro is another, but requires modifications of the fuel pump housing. I haven't really done the needed research for keeping the DW or Walbro happy, so can't really say what's needed.

 

For those that are curious, I'm hitting 300whp uncorrected in freezing temps, that's 100% injector duty cycle with ~10.8-11.1 AFR's.

[freaksavior]

Love your thread man. Just a couple cents to add. Don't forget the AEM is also a drop in and rated for e85

 

edit: dw65c is also e85 rated.

[covertrussian]

Thanks, I keep on forgetting about the AEM unit .

[Wasted Potential]

Not banging out a 110%+ IDC ?

[covertrussian]

I probably hit that when it was too rich, and yes it's physically possible, IDC is funny like that.

[utc_pyro]

Mine was hitting ~108% on days below 50F with the stock injectors. Didn't seem to be leaning out (more than it was otherwise) but it was obviously not the best way to keep things. Two sets of injectors, a new pressure regulating rig, two fuel pumps, and a $$$ set of top feed hardware and it's still not perfect.

 

So Covertrussian, you ready to find out how much of a pain these cars are once you change out injectors?

Edited January 24, 2018 by utc_pyro

[covertrussian]

That's actually why I'm not touching them, though top feeds do sound easier to deal with overall.

 

Since I don't want to have to deal with a #YNASB or transmission, at this point I'm not modding for power, mainly for efficiency.

[Flinkly]

...ready to find out how much of a pain these cars are once you change out injectors? [emoji38]

 

why didn't you post this years ago!?! lol

 

just don't give up. misery loves company, and like covert, i appreciate your work.

[covertrussian]

Now I'm curious, what kind of issues are you guys seeing?

[utc_pyro]

The stock tune has all sorts of weird crap going on to compensate for the nonleaner nature of the stock setup. You'd think "changing the injectors I can just change the scaling to match this flowsheet and be good to go!". But some dude in japan is laughing at you because the G/sec airflow reading and G/microsecond fuel scalier aren't real. They are some made up numbers that are kind of close to that, but they screwed with it a little to fix some lean spot or stumble.

 

Go look at your MAF load compensation table for instance. Then go look at an '07+ car. Same MAF sensor and airbox, but completely diffrent table. Why? Because they were using it to compensate for the POS nature of the sidefeeds. They also add a bit of fuel in the per-injector comp tables around idle.

 

BUTTT, our ECU's have low pulse width injector compensation tables JUST FOR THIS, and that was ENABLED. Why didn't they use that instead of the other comp tables and messing with the latency? Who the f*ck knows*!

 

Then you get into the fact the per-injector compensation tables are probably now off a little bit but you have no way to measure that because you don't have 6 widebands in your car **.

 

Oh, and the tip-in/out that noone REALLY knows what it's doing. It's probably simple in the software Subaru uses to tune the thing, but just looking at the code a lot of it makes you scratch your head.

 

And the wall wetting compensations you need to screw with because the injector spray patter is diffrent now.

 

And you're trying to fix all of these things with a flakey aftermarket wideband O2 sensor ~5ft of tubing from the heads. That's the only feed back you have!

 

Did I mention our fuel pressure also isn't remotely stable? And that if you start upgrading fueling bits to stabilize that you end up outflowing some of the bits in the tank. And the stuff in the tank is actually this really cool built in swirl pot and keeping that functional when modifying it to handle a bigger pump is king of an engineering nightmare?

 

So now you probably see why I'm shipping off a stock intake tomorrow to be flow benched. I need some real world numbers attached to this ball I keep massaging.

 

* Except that dude in Japan that did the OEM calibration. F*ck you dude. Ok not really, it would be fun to pick your brain for a few hours over ramen.

 

** If you're willing to try that (or a four channel EGT setup) I'm willing to toss in some cash on it. Preferably with your Tomei headers and EV14 injectors so the data would match my car.

 

Edit: If you're going to try this DW65C pump, EV14 injectors, rails with pulse damper (OEM or aftermarket). Bigger pumps out flow the fuel bucket bits, sidefeeds are super annoying, and fuel pressure fluctuations can cause all sorts of weird fun.

Edited January 24, 2018 by utc_pyro

[Flinkly]

i don't think i coukd have said it better.

 

i'm an engineer and can see both sides. should be able to swap in new injectors, change two values and go. but alas, someone just fudged tables to get it out the door...

 

it is already a complicated system, and is worse cause the tables are used to do things they shouldn't be.

[Wasted Potential]

* Except that dude in Japan that did the OEM calibration. F*ck you dude. Ok not really, it would be fun to pick your brain for a few hours over ramen.

 

I'd do it over Udon noodles, that ramen is like disintegrating cardboard!

 

 

I am no where nearly as knowledged as you folks, especially in the specifics of these ECU's and tables/tuning. I however debated the worthiness of pump/injectors for ~30-40whp for $1400-$1600. I passed due to anticipated complications.

Edited January 24, 2018 by Wasted Potential

[jager]

I personally wouldn't do a coolant to TB bypass, especially if you live where it gets cold and humid. Your throttle can freeze in place, which is obviously not good

 

 

 

I'm kind of over being pissed at them, it is what it is at this point. It just sucks to have to manually download and reupload everything (their bulk downdload doesn't work).

 

Now on interesting note, about 50% of the working pics are still on Photobucket in this thread. They seem to be taking down images that have too high of view count basically, but ones that don't get as many views seem to stay up

 

Yeah I read that about the TB bypass. I've got humidity in spades here in NC. However I drove it through august and sept. with no issues and it was really humid at points. It doesn't get too cold here, though we have had some single digit temps lately. No issues yet.

 

On the PB thing. I totally understand their decision. Bandwidth is relatively cheap these days, however it can add up quick and get really expensive. It just sucks that it screwed up so many good tutorials across the web. Wish they could have found another way.

[covertrussian]

I wouldn't blame the sidefeed injectors for say, I'm currently running them on my Nissan just fine, it's actually a common mod since most of our SR20's came with sidefeeds only. Mine actually came with top feeds, but I converted to sidefeeds because intake manifolds on top feed ones are garbage.

 

The stock tune has all sorts of weird crap going on to compensate for the nonleaner nature of the stock setup. You'd think "changing the injectors I can just change the scaling to match this flowsheet and be good to go!". But some dude in japan is laughing at you because the G/sec airflow reading and G/microsecond fuel scalier aren't real. They are some made up numbers that are kind of close to that, but they screwed with it a little to fix some lean spot or stumble.

 

I've discovered the same with my header and exhaust adventures. Like you mentioned, even comparing MAF scales between cars will vary from year to year, when they have the same MAF's & intakes.

 

Then you get into the fact the per-injector compensation tables are probably now off a little bit but you have no way to measure that because you don't have 6 widebands in your car **.

 

** If you're willing to try that (or a four channel EGT setup) I'm willing to toss in some cash on it. Preferably with your Tomei headers and EV14 injectors so the data would match my car.

 

Sorry, already sold the header to a local friend, who sold the car last weekend . As to why I sold it? Not because I missed the rumble, it's because our heads flow like crap thus header have a tiny benefit. This is why I haven't posted my header vs header data yet, I got tired of staring at it trying to find benefits . Sure it had less knock at partial load, but that only led me to run timing that was way past MBT.

 

 

And the wall wetting compensations you need to screw with because the injector spray patter is diffrent now.

 

Aha, so anyone that replaces their stock injectors with the reman ones is probably gonna be screwed with this one too. Stock have 12 pintle holes while most remans have only 4... I actually have a set of each, I was gonna test going from 12 to 4 on my G20 (since it's easier to deal with injectors on that car).

 

 

So now you probably see why I'm shipping off a stock intake tomorrow to be flow benched. I need some real world numbers attached to this ball I keep massaging.

 

Definitely curious on the data! Sadly I'm not sure if we would be able to put it to good use. Since we have to mess with the MAF table to account for any fueling changes. We basically need a constant multiplier, for overarching fuel adjustments, like replacing the intake you increase that constant instead of fudging with MAF scales.

 

In the pre-2000 Nissan tuning world, we have the MAF scale, Injector Scale, and K constant. Then engine load is TP, which is calculated by a bunch of fields and finally the K constant. Nissan doesn't even bother with the gram/sec/rev deal. When you replace a MAF for a higher HP one, you bring over the MAF scale, then you play with K Constant until your fuel is ~14.7, then you adjust the open loop portion of the fuel maps to be richer/leaner as needed.

 

To my surprise with this system, replacing intakes does nothing to engine load/maf scales. It just leans the mixture our because your flowing more air! Unlike on Subaru, it's leaning out because ECU is injecting less fuel because ECU things your flowing less air.

[utc_pyro]

I wouldn't blame the sidefeed injectors for say, I'm currently running them on my Nissan just fine, it's actually a common mod since most of our SR20's came with sidefeeds only. Mine actually came with top feeds, but I converted to sidefeeds because intake manifolds on top feed ones are garbage.

 

I don't necessarily blame the side feeds, but I do blame whoever tuned all the cars with them. You can see the same odd compensation on other cars (Baja, FXT, STI, JDM/EDM LGT) with the same side feeds though. I have a theory that their latency may be sensitive to absolute fuel pressure (not just pressure differential) and thus why they do the odd compensations.

 

Aha, so anyone that replaces their stock injectors with the reman ones is probably gonna be screwed with this one too. Stock have 12 pintle holes while most remans have only 4... I actually have a set of each, I was gonna test going from 12 to 4 on my G20 (since it's easier to deal with injectors on that car).

 

I have some 12 pintle hole DW injectors sitting around if you're ready to go down the rabbit hole! I'll even through in the scalings I was running before deciding screw it time for EV14's.

 

Sorry, already sold the header to a local friend, who sold the car last weekend . As to why I sold it? Not because I missed the rumble, it's because our heads flow like crap thus header have a tiny benefit. This is why I haven't posted my header vs header data yet, I got tired of staring at it trying to find benefits . Sure it had less knock at partial load, but that only led me to run timing that was way past MBT.

 

And now you have me wanting built higher flow heads

Edited January 24, 2018 by utc_pyro
typo

[Flinkly]

i was going to wait for a few days, but yeah, when are you going to source some spare heads to have ported for the testing, covert?

 

should start putting a paypal donation link in your signature

[covertrussian]

I'd do it over Udon noodles, that ramen is like disintegrating cardboard!

 

 

I am no where nearly as knowledged as you folks, especially in the specifics of these ECU's and tables/tuning. I however debated the worthiness of pump/injectors for ~30-40whp for $1400-$1600. I passed due to anticipated complications.

 

Ahahaha that's excellent .

 

Also good on you for figuring that out, a 250whp Legacy is still pretty damn fast. I'm actually quite content with that power level.

 

I don't necessarily blame the side feeds, but I do blame whoever tuned all the cars with them. You can see the same odd compensation on other cars (Baja, FXT, STI, JDM/EDM LGT) with the same side feeds though. I have a theory that their latency may be sensitive to absolute fuel pressure (not just pressure differential) and thus why they do the odd compensations.

 

I'll have to do some more digging, I've been blissfully ignorant about it, part of the reason I haven't hurried to replace my injectors or fuel pump.

 

I have some 12 pintle hole DW injectors sitting around if you're ready to go down the rabbit hole! I'll even through in the scalings I was running before deciding screw it time for EV14's.

 

Nope not ready for that hole yet, but it does make me wonder if installing my 4 pintle hole injectors will change fueling much. Too bad the injector clips absolutely suck for our cars. They are $7 a pop and I don't think I'll be able to reuse mine again.

 

And now you have me wanting built higher flow heads

 

i was going to wait for a few days, but yeah, when are you going to source some spare heads to have ported for the testing, covert?

 

should start putting a paypal donation link in your signature

 

I'm not sure there is much we can actually do, the big issue is the rear two cylinders have a sharp 90* wall that they hit on the way out. Without recasting the whole thing I'm not sure there is much we can do.

 

To give you guys an example, here is the Legacy GT and 16G vs my G20 with a 2.0L SR20DE (roller rocker) with a Garret GT28r. Notice how much power the SR20 is making at much less boost, now G20 is FWD so drivetrain loss is about 10% less, so do keep that in mind, and keep in mind that it's smaller displacement.

http://legacygt.com/forums/attachment.php?attachmentid=261250&stc=1&d=1516853818

 

Mods are pretty simple, 9.5:1 CR, Intake Manifold from earlier year SR20 (though Subaru 4-1 is still better design), 3" intake, slightly ported stock exhaust mani, Gt28r ball bearing turbo, 3" exhaust with STI muffler.

 

G20 has been putting out about 230whp for the longest time, but then when I went to do the 3" legacy exhaust I was forced to rebuild the G20 one too, which mainly needed a new cat. Exhaust back pressure was 3.1psi with old exhaust, and after rebuilt 2psi. First thing I noticed was spark blow out at 0.034" gap on redone exhaust, so I gapped them down to 0.030" and car gained 10whp.

 

On Legacy I reduced back pressure by 3psi, and gained 0whp . Same goes for other mods, EL vs UEL header? No consistent difference after retune, stock UEL vs Grimmspeed ported UEL and xpipe? No measurable and consistent difference . This is telling me that the issue is in the head, TGV's probably don't help too.

 

I'm testing intakes again right now, I wired in my manometer, and have some cool data to post here soon. But as usual intake cams so far seem pretty tiny, no other car is this hesitant to gain power with better flowing mods

Edited January 25, 2018 by covertrussian

[utc_pyro]

Nope not ready for that hole yet, but it does make me wonder if installing my 4 pintle hole injectors will change fueling much. Too bad the injector clips absolutely suck for our cars. They are $7 a pop and I don't think I'll be able to reuse mine again.

 

Ramp up the peer pressure and put them on sale on the classifieds bundled with something else you want. Got it

 

I'm not sure there is much we can actually do, the big issue is the rear two cylinders have a sharp 90* wall that they hit on the way out. Without recasting the whole thing I'm not sure there is much we can do.

 

According to the NAISOC head flow database, the dog leg is only down ~5% in flow. There aren't any big compensations for this in the stock pre-injector comp tables as well, so Subaru obviously didn't think it was a huge deal. It's looking like people are get pretty big flowbench increases with just PnP'ing and changing out the intake valves, not sure how this correlates with real world performance though.

 

This is telling me that the issue is in the head, TGV's probably don't help too.

 

Of all the cheap easy modes, you still haven't done this one?! Come on man, think of the MPG gains

 

In seriousness though, a guy on my kickball team deleted his with a dremel, four screws, and some JB Weld. Everyone is chomping at the bit to see your A-B testing of TGV deletes alone.

[covertrussian]

Ramp up the peer pressure and put them on sale on the classifieds bundled with something else you want. Got it

 

Woop I meant injector retainers, my pigtail clips are still good.

 

 

According to the NAISOC head flow database, the dog leg is only down ~5% in flow. There aren't any big compensations for this in the stock pre-injector comp tables as well, so Subaru obviously didn't think it was a huge deal. It's looking like people are get pretty big flowbench increases with just PnP'ing and changing out the intake valves, not sure how this correlates with real world performance though.

 

That's interesting I actually expected a bigger impact...

 

STOCK B25 STi head

Lift          Intake       Straight        Dogleg
.100           81             84               78
.200          166            166             155
.300          231            188             174
.400          248            195             178

 

Took me some time but I found the flow data for my head, sadly this one is slightly modified. source

 

ported low port head OEM lower I/M only also ported. OEM DET valves. 87mm bore. exhaust exit radiused port exit block.

 

intake test at 28in H2O, temp 83deg F

 

 

lift cfm

.060 56.9

.080 75.8

.120 110.1

.160 139.7

.200 168.1

.260 201.8

.320 233.3

.400 252.5

.440 257.4

.500 261.9

 

 

 

exhaust test at 28in H2O, temp 96deg F

 

lift cfm

.060 49.5

.080 65.9

.120 96.6

.160 121.1

.200 140.5

.260 160.0

.320 174.2

.380 182.7

.440 189.6

.500 195.2

 

this head made around 630bhp

 

Basically marginal difference on the head flow's, which is surprising me.

 

As for cam profiles on the G20, they are super mild even compared to other SR20's, Nissan built this motor for emissions and economy.

Intake Duration: 232*, Lift: 9.4mm

Exhaust Duration: 240*, Lift: 8.8mm

Overlap: 8*

 

EJ255 LGT cam specs:

Intake Duration: 240*, Lift: 9.60mm

Exhaust Duration: 240*, Lift: 9.78mm

Overlap: 0-35*

 

 

 

Of all the cheap easy modes, you still haven't done this one?! Come on man, think of the MPG gains

 

In seriousness though, a guy on my kickball team deleted his with a dremel, four screws, and some JB Weld. Everyone is chomping at the bit to see your A-B testing of TGV deletes alone.

 

I already built them and did a how to!

 

I'm just trying to test all of the easy to add or remove mods first. Since I always wish I did an extra few logs with the old setup. Once TGV's go in, they are not coming out

Edited January 25, 2018 by covertrussian

[utc_pyro]

Woop I meant injector retainers, my pigtail clips are still good.

 

Noting the bate to add to the FS post

 

I already built them and did a how to!

 

I'm just trying to test all of the easy to add or remove mods first. Since I always wish I did an extra few logs with the old setup. Once TGV's go in, they are not coming out

 

 

Hum, better thing to peer pressure you into. Your TGV’s have NTP ports, ever though about direct port meth injection? Washer fluid is ~30% meth, and would act like fuel. More power and no need to mess with the injectors yet

[Flinkly]

 

I'm just trying to test all of the easy to add or remove mods first. Since I always wish I did an extra few logs with the old setup. Once TGV's go in, they are not coming out [emoji38]

 

i hear you there. have so many questions about how my car was before i put in new injectors. too late to ask those questions!

[covertrussian]

Woop I messed with Virtual Dyno Trim window which caused the graphs to look hilly (before). Here is the proper graph:

http://legacygt.com/forums/attachment.php?attachmentid=261250&stc=1&d=1516853818

 

Oh @utc_pyro, the reason the G20 holds boost to redline better is simply because of the wastegate air feed location.

 

Legacy's wastegate feed comes right form the turbo outlet right and goes straight into the wastegate (no boost controllers). G20's feed comes from under the throttle body right before the intake manifold. Since G20 has an FMIC, it has a much higher pressure drop so getting wastegate feed directly form turbo equals to a bigger pressure drop by redline.

 

Since Legacy has OEM TMIC has shorter path and it's also tube and fin, thus it's pressure drop is not as bad, but still is 2psi!

 

I've actually been thinking about changing the wastegate feed and posting it up as a free power mod (2psi = 15-20whp gain). But I don't want to be tapping the plastic for a new nipple for this..

Edited January 25, 2018 by covertrussian

[utc_pyro]

Oh @utc_pyro, the reason the G20 holds boost to redline better is simply because of the wastegate air feed location.

 

Legacy's wastegate feed comes right form the turbo outlet right and goes straight into the wastegate (no boost controllers). G20's feed comes from under the throttle body right before the intake manifold. Since G20 has an FMIC, it has a much higher pressure drop so getting wastegate feed directly form turbo equals to a bigger pressure drop by redline.

 

Since Legacy has OEM TMIC has shorter path and it's also tube and fin, thus it's pressure drop is not as bad, but still is 2psi!

 

I've actually been thinking about changing the wastegate feed and posting it up as a free power mod (2psi = 15-20whp gain). But I don't want to be tapping the plastic for a new nipple for this..

 

 

You mentioned that before actually. Next time I have my intercooler out I’m going to drill and tap it again to move my MBC reference line there based on your suggestion. AVO List my intercooler as having 0.8psi drop at 400CHP, and I’ll probably be there after turning the boost up. Doing this will also increase the signal hitting the EBCS at the turbo outlet spot, giving it faster response for killing boost.

 

Note that the AVO intercooler is optimized for flow and not cooling efficiency, this it’s lower pressure drop. The other day it was ~40f and I saw ~100f post IC IAT temps during a short pull.

[covertrussian]

I wonder if it would help much for MBC or EBC users, you would have to adjust your WGDC tables again too. If I can find a good port to tap, I might do the same in the near future.

 

I would be curious to see a bar and tube intercooler efficiency in comparison. I personally prefer a slight flow restriction to better cooling capacity, but stock intercooler has been surprisingly decent, though I haven't tried measuring internal post IC temps.

[utc_pyro]

It’s a hybrid setup, so most of the 100% throttle area is set to 99% WGDC with the MBC taking over. I get a pretty bad taper above 5500 RPM so I’m hoping this cuts some of that down.

 

 

Sent from my iPhone using Tapatalk