CovertRussian's 05 LGT Build Thread

[Rhitter]

What no cursing about getting the OEM heat shields off? Sure is pretty with out all the wrap on it.

[covertrussian]

OEM heat shields have been off for like 6 months man, dropping the stock header was a breeze!

[Flinkly]

OEM heat shields have been off for like 6 months man, dropping the stock header was a breeze!

what!?! heatshields should gain mpg on multiple fronts. why remove? or rattling cause rusted out?

 

and any plans to retain heat on these new ones?

Edited April 24, 2017 by Flinkly
sp

[covertrussian]

what!?! heatshields should gain mpg on multie fronts. why remove? or rattling cause rusted out?

 

and any plans to retain heat on these new ones?

 

I removed them in November due to false knocks (and annoying rattles). I actually got some of my best MPG, 31.33mpg, to the same location, after I removed the stock heat shields (and well did a few other things). Detailed post about it is in my Tuning for Economy thread.

 

It caught me by surprise too, so at least for stock headers I'm not convinced heatshields are a must. For this header, maybe, but I'll test it as is first. I just gotta figure out how to not melt my plastic belly pan.

[covertrussian]

It's been a while since I bought this header... finally got around to replacing my clutch, tires, and fine tuning my UEL tune, and getting some gas mileage runs out of the way, thus I could finally get the ELH header installed.

 

Installation

Started with removing the O2 sensor and Up pipe bolts

http://legacygt.com/forums/attachment.php?attachmentid=253423&stc=1&d=1501645968

 

After that unbolted the header from the head, very easy stuff. Man the valves are super white, at first I was worried about it, but it sounds like it's a normal thing on pump gas.

http://legacygt.com/forums/attachment.php?attachmentid=253426&stc=1&d=1501645968

 

Now my gaskets show signs of exhaust leakage, but the head and header don't

http://legacygt.com/forums/attachment.php?attachmentid=253424&stc=1&d=1501645968

http://legacygt.com/forums/attachment.php?attachmentid=253425&stc=1&d=1501645968

 

Removing the header was super easy (because I had no heatshields). Removing the up pipe on the other hand was more involved with having to unbolt the downpipe and all the turbo bolts.

http://legacygt.com/forums/attachment.php?attachmentid=253427&stc=1&d=1501645968

http://legacygt.com/forums/attachment.php?attachmentid=253428&stc=1&d=1501645968

 

Header and Up Pipe out and enjoying their new home on the garage floor.

http://legacygt.com/forums/attachment.php?attachmentid=253384&d=1501642619

 

Here is the Tomei EL 193105 vs OEM header

http://legacygt.com/forums/attachment.php?attachmentid=253429&stc=1&d=1501645968

 

Next I installed the EGT plug (luckily Tomei provides you with one), and the up pipe studs (using the two nut method seen on the lowest stud)

http://legacygt.com/forums/attachment.php?attachmentid=253430&stc=1&d=1501645968

 

Since I had two parts of the header apart, I reinserted them too far in which means it was too short for my head. Used a rubber mallet to separate them back out. This is how much it had to come out to fit on my heads.

http://legacygt.com/forums/attachment.php?attachmentid=253431&stc=1&d=1501645968

 

With the up pipe being loosely bolted onto the turbo I installed the header. If you pre-tighten the up pipe, you will have a hard time getting it to align with the header.

http://legacygt.com/forums/attachment.php?attachmentid=253433&stc=1&d=1501646016

http://legacygt.com/forums/attachment.php?attachmentid=253432&stc=1&d=1501646016

 

Before:

http://legacygt.com/forums/attachment.php?attachmentid=253422&stc=1&d=1501645968

 

After:

http://legacygt.com/forums/attachment.php?attachmentid=253435&stc=1&d=1501646016

 

The header does stick out past the bumper. I didn't install the belly pan because I didn't want the header to melt it. Might have to just get me a proper aluminum skid plate...

http://legacygt.com/forums/attachment.php?attachmentid=253434&stc=1&d=1501646016

 

First impressions

The engine sounds much smoother, it also seems to idle much calmer, come to think of it haven't felt the typical missfire/bad ignition coil like idle on stop lights. I'll pay more attention tomorrow.

 

About the noise, on stock exhaust it sounds pretty much the same as it did before. I'll have to let someone else drive it to really hear it now.

 

Fuel Economy

To be determined...

 

Power Gains (or Losses)

Before removing the header I went for a last spin and adjusted the timing for max power my stock header could do, which was 254whp/260wtq at 80F. This was done on one of the warmest days so far.

 

I installed the header that night, it rained the next day, and when it finally stopped raining on Sunday the temps were down to 55F with more rain coming. Thus I had to take advantage of the dry weather and try to tune the header out. To make sure I wasn't seen bloated numbers I did 3 base runs on the UEL tune, tuned timing, then went back to UEL tune.

 

UEL tune with 25*F colder temps got me 241whp/257wtq. Kind of a bummer right, $700 and lost power.. I started searching for the MBT timing, I do this by adding/subtracting 2* from the map. Subtracting 2* from the whole map made me loose even more power: 238whp/246wtq. This told me that I need to go up instead of down.

 

Next I added 2* to the UEL base timing, this got me up to 256whp/264wtq, this meant the header raised my MBT and the engine was now fine with more timing (mind you colder 55F weather could have had something to do with this too). To find how much more timing the engine would now like I went much higher on purpose, I added 4* more (6* over UEL timing), this set me back to 232whp/256wtq, as in I'm way past MBT now.

 

Finally I removed 2* from the whole map, which put me 4* over UEL timing now. This got me 242whp/264wtq, with the power/tq curve being identical to the 2* over base map until 5200rpm. When you get closer to MBT your torque curve will plateau before declining. This told me that my original +2* over UEL map was about the most optimal timing for my current setup with the current AVCS.

 

I know that's a lot of text, virtual dynos are coming next, I figured some of you would be interested in how to tune a car for new mods, with this method you will pretty quickly find how much timing your new mods like. Basically you increase timing until you stop gaining power. Do keep in mind your tuning temperature, as it gets colder or hotter the amount of timing you can run will vary though.

 

http://legacygt.com/forums/attachment.php?attachmentid=253436&stc=1&d=1501646016

 

End results is 2whp/4wqt peak gain, dollar per HP is not looking good, and it really bummed me out yesterday. I did gain up to 10whp/13wtq in some parts of the map, but I could attribute that to colder temperatures too.

 

Here are my ideas on why the power gains are so minimal:

1. I'm only running 13psi peak boost, which doesn't create as much back pressure as most stage 2-3 setups.
   
2. This header is about 150rpm slower to spool to full boost.
   
3. Stock tune doesn't run much AVCS after 4k rpm to reduce reversion. I think this is Subary's way of accounting for a crappy flowing header.
   
4. My tune is very fine tuned for my setup (fuel and timing wise), I'm already at MBT for the motor's VE. In this case, even though header increases VE, AVCS is artificially reducing my VE now. This header should provide me more AVCS tuning headroom.
   
5. Most people that install the header wont be as fine tuned, which means the new parts will get them closer to MBT and thus they make power simply by bolting up the new part.
   
6. Another item in my exhaust is the bottleneck, I have a 2.5" divorced downpipe to the stock 2.4" exhaust/mufflers.
   
7. My stock intake & filter or 2.5" inlet could be the bottleneck now too.
   

 

UPDATE - Similar Temp dyno runs - Detailed post

http://legacygt.com/forums/attachment.php?attachmentid=262214&stc=1&d=1519228817

Edited February 21, 2018 by covertrussian

[Flinkly]

you are flowing more freely with elh vs. stock and thus the 2.5" exhaust should be fine, at least in comparison. pulses should be moving single file now. and while not absolute fact, "common knowledge" says stock intake is fine till 350-ish.

 

my bet is overall tune, and AVCS specifically.

 

well, and you are also losing more heat/power from these than before. way more surface area (2 primaries vs "1") and thinner material (vs the ~25% cast oem headers) and no tray.

 

and the slower boost makes sense with the larger volume to fill.

 

can't wait to see more work to fine tune for these.

 

and not to say you don't already know all this, just chatting.

Edited April 25, 2017 by Flinkly

[covertrussian]

I wouldn't think that heat loss is that big of a deal honestly. Without heat shields the stock header would loose a bit of heat too when it was this cold, yet it would gain .

 

As for intake, my previous testing told me that factory intake are great cold air intakes, but suck at flowing still. My quick DIY warm air intake gained 26whp/8wtq without any tune changes. Though back then bullet point #5 applied. So those gains could have been simply from my timing being so miss-tuned. The only reason I'm not running it right now is high IAT's, damned thing needs an airbox otherwise it's knock nation.

 

As for AVCS

Here is the AVCS map that I've been running for a while now.

 

Today, in the pouring rain, I did play with a little with AVCS. I tried out running 10-15* of AVCS on the whole map, instead of letting it tapper off to 0* AVCS.

http://legacygt.com/forums/attachment.php?attachmentid=248258&stc=1&d=1493093332

 

The engine really liked 15* of AVCS below 5k rpm, went from 256wtq to 269wtq. After 5k power just plateaued, guess 15* of AVCS (5* of overlap) is too much for it's liking, it didn't seem to like even AVCS at 10* (0* of overlap). Guess I'll try 5* and if all else fails go back to 0*.

 

Thus my theory that the header will tolerate more overlap down low is starting to look accurate. Testing AVCS in 5* increments would be awesome, but then fine tuning the timing would take for ever, especially on the street.

Edited November 8, 2018 by covertrussian

[BrianEarlSpilner]

and the slower boost makes sense with the larger volume to fill.

 

I've noticed a slight increase in time to full boost after installing elh. Posts on the topic make you believe that the spool is always faster. I figured with the engine pumping a given volume per rpm and boost that the longer runners would take more time to fill.

 

Have you had to increase wgdc to hit target boost since the install?

[covertrussian]

I'm actually running wastegate pressure only (EBC is not hooked up). My 16G has about 12.8psi gate, it has more then enough power and it avoids ECU messing with WGDC's varying my boost levels between runs/mods.

[Flinkly]

honestly, you should just get an STI intake manifold to even out the intake flow. their TMIC is free-er flowing too (shorter path; top-to-bottom rather than left-to-right).

 

then maybe E85 for it's added knock resistance, and move back to the WAI.

 

granted, E85 is a step backwards for MPG, but sure gives alot more room for knock resistance and power. can always credit yourself the ~28% difference in power/mpg between premium and E85.

 

 

Seriously though, always interested in your tinkering.

 

As for MPG between the legacy and outback, the outback never got the underbody aero that the LGT did. Does yours still have the 2 belly pans and front wheel well covers? I also just heard that there is a transmission cover as well, for the area between the undertray and two rear covers (not for USDM market).

Edited April 25, 2017 by Flinkly

[covertrussian]

It's been a while since I looked at the manifolds threads, I thought the plastic manifolds were pretty decent flowing? I still have TGV's, which are a bigger restriction at this point.

 

E85 is out of question for me, no local gas station has it, nor do I have the time or money to deal with it. I'm starting to get closer to 300whp/wtq and I'm actually starting to get concerned about breaking my trans or engine. Thus I think I might artificially (via tune) keep my power at 270 levels.

 

As for aero, I have one belly pan (which is off right now), and two side pans (by jacking points), which are somewhat damaged from running the deer over. I do have the wheel well covers and one rear tire deflector (deer took out the other). No transmission cover or exhaust/diff covers beside that.

[Flinkly]

It's been a while since I looked at the manifolds threads, I thought the plastic manifolds were pretty decent flowing?

 

yeah, mostly giving you a hard time. depending on where you look, the STI manifold has tighter distribution of flow to each cylinder, and flows maybe 6% better. definitely lower hanging fruit as far as cost and work involved.

 

And i'm in the same boat with E85. nowhere to get it.

 

Would you do a TGV delete? and would you remove the center rib? I guess for the early cars it only functions at startup, so not a lot if any efficiency to be had from retaining them.

[covertrussian]

I'm thinking about doing the TGV Deletes yeah, especially since they are there only during cold starts. I would get rid of the center bar and either weld it (might be too hard to find someone here to do that) or use the weldless delete plates.

[covertrussian]

It finally stopped raining and was in the upper 70's today. Which means I could do some dyno logs which would be similar to the last UEL header logs.

 

First thing I noticed between the 55F runs and these is, the engine load is about 10g/s less in 78-80F weather. This changed a lot of variables in timing and fuel and finally made this header shine.

 

Red line shows the Stock UEL header, Blue line is Tomei EL Header with the same exact tune, and Green line is ELH header with 2* of timing added to the whole map:

 

First, when we compare the Red line to the Blue line, first thing you will notice is the massive torque loss (25ft-lbs) under 4.5k rpm. But then towards the redline the header regains 16ft-lbs, which leads to a 16whp gain too.

 

Next, is the green line with 2* of timing added to the whole map. This made a massive recovery in torque across most of the map, except under 3k. At first I was going to blame turbo lag, but at 3k rpm there is only 1psi difference between the two. I'm afraid this is the byproduct of the bigger runners and the current AVCS map (basically 10* until 4.8k, then 0*).

 

Continuing with the green line, the extra 2* added 20whp/wtq at 4-5k rpm, but then by 5.5k we started loosing power, but the blue line with 2* less was excelling. Tells me that from 5.5k on I need to reduce my timing by 2* at those specific load columns (virtual dyno can't show those, so you have to look at logs to get that data).

 

Now the air to fuel ratios are about 0.4 leaner. Some of the power gains can maybe be attributed to that, but I don't think it would be as much as 20whp. Either way I updated the fuel to be richer in those columns for next tests.

 

Update, forgot about the Boost portion: If you look carefully the EL Header is actually running as much as 1psi less after 4,000rpm on my boost control-less setup (wastesgate spring only). I think this can be contributed to the equally spaced pulses. UEL headers will stack pulses more causing a stronger push.

 

AFR Update: I'm now pretty convinced that most of the gains were from much leaner AFR's unfortunately.

Edited June 11, 2018 by covertrussian

[StkmltS]

The way you explain your tune changes (what and why) is really helpful to us less experienced people. Thanks!

 

I'm thinking about doing the TGV Deletes yeah, especially since they are there only during cold starts. I would get rid of the center bar and either weld it (might be too hard to find someone here to do that) or use the weldless delete plates.

 

Another option is to tap the empty holes where the center bar was and plug+seal them with bolts. It's quick, easy, cheap, and you don't have to wait on anyone but yourself. On mine I just left the center holes (between 1/3 and 2/4) empty. A little crossover that close to the valves probably isn't a bad thing.

[covertrussian]

Update, forgot about the Boost portion: If you look carefully the EL Header is actually running as much as 1psi less after 4,000rpm on my boost control-less setup (wastesgate spring only). I think this can be contributed to the equally spaced pulses. UEL headers will stack pulses more causing a stronger push.

 

The way you explain your tune changes (what and why) is really helpful to us less experienced people. Thanks!

 

Glad to hear that it's coming in useful! I find that, taking the extra time to think about it to be able to explain it, forces me to see things that I would have otherwise missed. This happens a TON, so it's as beneficial to me as it is to you guys

 

Another option is to tap the empty holes where the center bar was and plug+seal them with bolts. It's quick, easy, cheap, and you don't have to wait on anyone but yourself. On mine I just left the center holes (between 1/3 and 2/4) empty. A little crossover that close to the valves probably isn't a bad thing.

 

That's an excellent idea! I'm not sure why I didn't think about that before. I have spare set of TGV's that I might just do that to next week while I'm chilling at home.

[Rhitter]

Your experience about slightly later boost threshold mirrors my own. Difference is Cryotune tuned my car for a GS TMIC and Tomei ELH at the same time. There is just a lot more areas to fill with air I think.

 

I am also peaking around 19 psi and the midrange feels so much stronger than the VF52 on the OEM header

[covertrussian]

Did the first gas mileage test, it was a pretty windy day 15-25mph winds, got 28.36mpg. While this is way down, last week on UEL header I got 26mphg with 14mph winds. Both of these tests were 80-90*F. Before rejoicing with a 2mpg gain, 3 weeks ago I got 29.61mpg with up to 15mph winds, temps were 74-88F.

 

I wonder if something is going wrong, maybe spark plugs are due for replacement, or maybe brakes are starting to drag.

[covertrussian]

Pulled the spark plugs to make sure they looked good, while overall they looked fine they were majorly out of gap... Lets just say these were gapped to 0.040 37k miles ago (these are iridiums). At 0.059" I'm amazed that I didn't have spark breakup.

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Engine/Sparkplugs/20170502_123238.jpg~original

 

The spark ate a dish into the ground strap, which is hard to see in a picture:

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Engine/Sparkplugs/20170502_123830.jpg~original

 

One interesting observation, cylinder 1 (left) vs cylinder 2 (right). Notice that Cylinder 2 one is worse looking (I'm betting due to UEL header).

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Engine/Sparkplugs/20170502_152958.jpg~original

 

I had LFR7AIX laying around the garage from when I bought this car, I ended up using 6's since I was not even stage 2 back then. Lets just say I've been past overdue for 7's, though EL header probably makes 6's be sufficient. Gapped these to 0.035.

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Engine/Sparkplugs/20170502_141743.jpg~original

 

 

Was gonna explain the drastic changes (for good and bad), but it's easier to post a dyno graph. Keep in mind these runs are on different days but from same road.

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Engine/Sparkplugs/Spark%20Plug%20Gap%20vs%20Headers.png~original

 

Take aways:

- New plugs caused the map to lean out by .5 AFR, Sending me into 12-12.5AFR range!

- Spool increased back to UEL levels

- Blue line is new plugs before richening up the map, it was knocking up top, this can bee seen in the Total timing graph.

- New plugs gained 20wtq at peak torque (peak boost), which means that spark was being blown out.

- I lost power overall, I probably need to find Optimal timing all over again

 

 

Next I installed the undertray, I think it has plenty of clearance to the header

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Exhaust/Header/Tomei%20193105/20170502_155708.jpg~original

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/Exhaust/Header/Tomei%20193105/20170502_155720.jpg~original

 

 

Finally gas mileage test: With new plugs and undertray, 200lbs less weight, I got 30.27mpg, this is with wind speeds of 12mph (18mph gusts). I think with 0-5mph winds this would have been closer to 31mpg. I'm glad to be back in 30's, but I was hoping for 2-3mpg bump that everyone talks about. I am using 10* of AVCS, so maybe finally the engine will be fine with 15* of AVCS?

 

Since it's been so windy here, I think I might turn focus to doing some aero mods...

Edited January 21, 2018 by covertrussian

[Max Capacity]

I'm confused about why your opening up the plug gap from .028"

 

 

I think your old plugs had opened from age and usage.

[utc_pyro]

Ok this is bothering me, why would regapping your plugs effect fueling?

 

Air entering the system is metered at the MAF way up at the intake. The ECU takes this and calculates how long to open the injectors to create a given air fuel ratio. How does this change due to anything happening after the intake valves close?

 

Also if your spark was being blown out, shouldn't that cause the AFR to READ lean due to more unused oxygen being present in the exhaust?

Edited May 4, 2017 by utc_pyro
typo

[covertrussian]

I'm confused about why your opening up the plug gap from .028"

 

 

I think your old plugs had opened from age and usage.

 

Manual calls for .028-.031", so I'm not too far off. The idea is, the bigger the gap the stronger the spark, is exposed to more air/fuel, and reduced quenching.

 

Now the reason the spark is stronger, with smaller gap it takes x amount of voltage to jump that gap, as you increase the gap the more voltage it takes for electricity to jump that gap, which in return does give you a stronger spark. This in theory helps with fuel economy and power.

 

Now I was surprised that iridium plugs were that worn at 37k miles, but since it was being exposed to a stronger and stronger spark, it accelerated the wear everytime the gap got bigger

 

 

Ok this is bothering me, why would regapping your plugs effect fueling?

 

Air entering the system is metered at the MAF way up at the intake. The ECU takes this can calculates how long to open the injectors to give a given air fuel ratio. How does this change due to anything happening after the intake valves close?

 

Also if your spark was being blown out, shouldn't that cause the AFR to READ lean due to more unused oxygen being present in the exhaust?

 

Good call on this one, it's kind of boggling my mind too, more complete burn = less fuel but less oxygen too. I just checked those two logs, since they are 10*F apart maybe it was in different load ranges (reading different fuel map values to cause leanness).

 

Well the new spark plug log, which is 10*F cooler, is running about the same engine load as other one, in some places .05 g/rev less. This is odd because cooler runs usually run more load too.

 

On the other hand the run with green line with richened fuel, which is 6F cooler ontop of blue line, is running almost .10 g/rev more then blue line. Which definitely threw me into another column, where apparently my timing is not as well tuned . I lost power with colder weather, ironic

[utc_pyro]

Good call on this one, it's kind of boggling my mind too, more complete burn = less fuel but less oxygen too. I just checked those two logs, since they are 10*F apart maybe it was in different load ranges (reading different fuel map values to cause leanness).

 

Well the new spark plug log, which is 10*F cooler, is running about the same engine load as other one, in some places .05 g/rev less. This is odd because cooler runs usually run more load too.

 

On the other hand the run with green line with richened fuel, which is 6F cooler ontop of blue line, is running almost .10 g/rev more then blue line. Which definitely threw me into another column, where apparently my timing is not as well tuned . I lost power with colder weather, ironic

 

Found a study that talks about O2 in the exhaust and engine timing. It appears that advancing the timing more can cause higher peak pressures resulting in both fuel and oxygen getting taped in areas it wont burn. These come out at the end of the cycle after the flame is quenched, and thus can screw with O2 sensor readings. Optimally gaped plugs should make the mixture light off and propagate quicker, so should have the same effect as advancing the timing.

 

Study and the effects of ignition timing on gasoline engine performance and emissions

 

So the question is does this change the true effective AFR, or is it just messing with out ability to measure things?

Edited May 4, 2017 by utc_pyro

[covertrussian]

Excellent find! And yes, even with costly widebands we can still be blind in right situations. There is also the trend of wideband seeing a richer mixture as you get a more complete burn (because you burned more oxygen). So you start removing fuel and boom goes the engine.

 

Luckily Subaru's knock sensor and control is top notch. It's so good that you can actually tune by how much timing it pulled for you. Say you do a run and it feedback pulls 4* of timing, well now you know that your around 3-4* over advanced in that area, if you were 1-2* over advanced it would only pull 2* of timing.

 

With new spark plugs I was getting 2-4* of timing pulled via feedback, and in some cases my IAM dropped too. This is when I knew something wasn't right, so zoomed in on the AFR area and bam she's in 12.4-12.0 range.

 

Even with me richening the mixture way up (by 1 whole point), I'm still only hitting 87% duty cycle on my stock injectors and fuel pump. Makes me wonder if I can hit 300whp on stock fueling simply by keeping my PSI down.

[Max Capacity]

It's a known thing in the turbo world the to much spark gap...and your spark will get blown out. As we raised boost in the race car we tightened up the spark plug gap. That was even with a high output coil. At 35psi we were down to .018"