CovertRussian's 05 LGT Build Thread

[covertrussian]

Back to intercooler sizes, I think Mishimoto's last formula is the wrong one in the equation (Comments in that blog suggest the same). EDIT: Looks like Mishimoto got this formula from Corky Bell, thus it's most likely accurate. I'll have to do more research on why.

 

Mishimoto's formula:

Internal Flow Area (in²) = Charge-Air Surface Area (in²) ÷ Core Thickness (in)

 

For reference that gave these numbers:

Stock: 32.07

Precision 350 hp: 34.22

 

Since intercooler thickness matters, it should probably be multiplied instead, but that would make it into volume in^3

Internal Flow Area (in²) = Charge-Air Surface Area (in²) x Core Thickness (in)

 

This formula results in:

Stock: 200 in^3

Precision 350 hp: 419 in^3

 

This formula makes the Precision 350hp look 200% better. Which didn't seam right, so I went back to the basics and calculated the actual internal volume:

Stock: 90 in^3

Precision 350 hp: 189 in^3

 

From a sheer volume perspective the intercooler is twice as big. Now twice the volume doesn't necessarily mean twice the HP support, but it does mean that stock is plenty under 350bhp mark that Precision rates theirs at, probably 250whp. Which is around where I'm stuck at too.

Edited February 8, 2018 by covertrussian

[covertrussian]

Alright update to this, I found a local guy with a 07 WRX shell that has engine/trans/rear diff still. He wants to sell the whole car for a reasonable rate delivered to me. Should have a few other goodies in it that I should be able to part out too.

 

Running through the parts compatibility list:

07 WRX's (TY754VW7AA-TX)

 

Axles:

Still use traditional speed sensors and shorter axles then GT's. The front spline count matches, and the axle design looks the same.

Rear spline count should match but can't find definite answer just yet.

 

Gears:

One downside to 07 WRX is it does have slightly thinner gears, like the 03-05's, I for some reason was hoping that 06-07's were the same as LGT's, but that seems to be 08+

 

This makes me wonder if I should just swap Final Drive gears and rear diffs instead.

 

Drive shaft:

Unknown, LGT one should be fine though.

 

Rear Diff:

Unverified compatibility with LGT axles

 

Differentials:

Same: Open front, Viscous center, Viscous rear

 

Anything else that I'm missing?

 

Besides that, this would give me an excuse to finally install the Blast plates that I have picking up dust .

 

 

Has anyone swapped in a 07+ LGT, 08+ WRX or 2006-2007 WRX transmission into the 05/06's? While doing some research and math on gear ratios between cars, I noticed that 07+ LGT's and 08+ WRX's have a 3.900 final gear and that 06-07 WRX's have 3.700 final gear, while having the same 5th gear as our cars (0.738). Our final gears are 4.111.

 

Why? Because it would greatly reduce the highway RPM. Here is an example with 225/45/17 (24.97") tires.

2005 LGT (4.111): 2,859rpm @ 70mph

2007 LGT (3.900): 2,712rpm @ 70mph

2006 WRX (3.700): 2,573rpm @ 70mph

2005 OBX (4.444): 3,090rpm @ 70mph

 

For comparison sake, here is the 6 speed's with same tire:

2007 LGT: 2,598rpm @ 70mph

2007 STI: 2,778rpm @ 70mph

 

And just for Flinkly OBX on stock 225/55/17 (26.7") tires:

2005 OBX (4.444): 2,890rpm @ 70mph

 

This means that the 2006-2007 WRX rev's lower at 70mph then even the 6 speed Spec B's.

 

Anyone looking to sell their WRX trans?

Edited February 6, 2018 by covertrussian

[Sgt.Gator]

Use the 07 WRX as a donor to build a FF 818R.....

[Sgt.Gator]

Back to intercooler sizes, I think Mishimoto's last formula is the wrong one in the equation (Comments in that blog suggest the same)

 

Mishimoto's formula:

Internal Flow Area (in²) = Charge-Air Surface Area (in²) ÷ Core Thickness (in)

 

For reference that gave these numbers:

Stock: 32.07

Precision 350 hp: 34.22

 

Since intercooler thickness matters, it should probably be multiplied instead, but that would make it into volume in^3

Internal Flow Area (in²) = Charge-Air Surface Area (in²) x Core Thickness (in)

 

This formula results in:

Stock: 200 in^3

Precision 350 hp: 419 in^3

 

This formula makes the Precision 350hp look 200% better. Which didn't seam right, so I went back to the basics and calculated the actual internal volume:

Stock: 90 in^3

Precision 350 hp: 189 in^3

 

From a sheer volume perspective the intercooler is twice as big. Now twice the volume doesn't necessarily mean twice the HP support, but it does mean that stock is plenty under 350bhp mark that Precision rates theirs at, probably 250whp. Which is around where I'm stuck at too.

 

See this link too:

https://www.linkedin.com/pulse/20141110020646-32218462-intercooler-sizing-for-turbocharged-engines-foundation/

[covertrussian]

Earlier in the thread utc_pyro mentioned that I should be able to run a GM IAT sensor in the EGT input. I had a spare GM IAT sensor from my project Nissan:

http://legacygt.com/forums/attachment.php?attachmentid=261797&stc=1&d=1518325482

 

 

Installation

Since this is a quick proof of concept, I didn't want to over complicate it so went with the quick and dirty route. I took my spare intercooler, drilled a small hole in a flat spot of the cold side of the intercooler, scuffed it up and then used QuickSteel putty to secure the IAT sensor to it.

http://legacygt.com/forums/attachment.php?attachmentid=261791&stc=1&d=1518325146

http://legacygt.com/forums/attachment.php?attachmentid=261792&stc=1&d=1518325146

 

Intercooler installed and connected the IAT sensor to the EGT plug

http://legacygt.com/forums/attachment.php?attachmentid=261793&stc=1&d=1518325146

 

 

Tune Updates

Next the tune needs a new sensor scale for EGT sensor. At first I tried Cobb's GM IAT scale, but using the regular scale made the ECU not display any useful values. Talking to utc_pyro, he said that EGT sensor input is 8 bit, thus the scales need to be multiplied by 10. Update: ECU doesn't read below 392F, which means even multiplied by 10, it wont read below 39.2F, I am testing multiplying by 13 right now.

http://legacygt.com/forums/attachment.php?attachmentid=261794&stc=1&d=1518325146

 

For easier pastability into ECU Flash here is the actual values that you can then multiply as needed.

Temperature
248
248
235
215
199
186
176
168
161
152
146
140
133
127
120
113
106
100
94
88
82
76
70
64
56
47
33
17
-9
-40

 

Next you can modify the RomRaider Logger definition to divide the displayed value by 10 to make it easier on the eyes, otherwise you have to do the math in your head (though seeing 600F might be funny on the eyes at first).

 

<parameter id="P79" name="Exhaust Gas Temperature" desc="P79-Exhaust gas temperature reading." ecubyteindex="40" ecubit="1" target="1">
...
<conversion units="IAT F" expr="(32+9*(x+40))/10" format="0.0" gauge_min="-100" gauge_max="300" gauge_step="0.1" />
...

 

 

Run Data

 

More in depth testing at different temps can be found in this post.

 

Finally to some run data, here is a 2nd gear pull, this is a very first pull of the session, thus intercooler has not been heat soaked yet. Atmospheric temp was 47F:

http://legacygt.com/forums/attachment.php?attachmentid=261795&stc=1&d=1518326475

 

Post IC temp starts at 52.7*F, by redline it is 64.4*F. This is 11.7*F rise, this doesn't seem too bad, but keep in mind it was the first pull, before that car was just driven in city/backroads.

 

I then made 4 more pulls, this road is a 55mph one, I basically made 4 circles with about half a mile between pulls. Here is what the 4th pull looks like at the same 47F ambient:

http://legacygt.com/forums/attachment.php?attachmentid=261798&stc=1&d=1518328283

 

Post IC temp starts at 59*F and by redline it's 68.9*F, that's only a 9.9*F rise ! So far the stock intercooler is not doing too bad though, only 5*F higher then the first pull, but still 22F above ambient.

 

Next up is seeing how well it does with continues back to back pulls, this is with 44*F Ambient temperature:

http://legacygt.com/forums/attachment.php?attachmentid=261796&stc=1&d=1518326475

 

Post IC temp starts at 50F, end of first pull is 68F, second pull 78F, third pull 83.3F, fourth pull 84.2F, after that they stayed around 80F. Looks like 40*F over ambient is the highest it would go.

 

All of these logs were done with my cold air intake, which gets air from fenderwell. Even the intake couldn't keep cool: started at 46F and was at 52F for most of the pulls (this just shows that manifold/uppipe heats fenderwell a good bit).

 

 

Conclusion

It's crazy to watch the temps jump up so quickly, usually it's a little delayed too, as in the intercooler hits peak temp after the pull is done. I'm also not 100% convinced that stock intercooler is the culprit for me (at least at under 60F ambient), especially since it didn't go up that majorly even after 4 circle pulls (more of how I test and tune instead of back to back punching it).

 

While the stock intercooler jumps way up in temp, it also cools down fairly quickly, a denser unit would rise in temp slower, but also cool slower. I think this is one of those, FMIC or go home mods for me.

 

Finally, big thanks to utc_pyro for suggesting this and helping me out with getting it to actually read the data!

 

Continued in this post with 33F testing.

Edited March 3, 2018 by covertrussian

[covertrussian]

Quick update that turned into a book... With the temps dropping to 33F wanted to see how much post intercooler temps would rise at near freezing ambient temps.

 

First thing I noticed is, ECU wont display EGT's below 392F (39.2F), that means even multiplying the EGT table by 10 is not sufficient to get resolution below 39.2F. I multiplied it by 20, which seemed to do the trick (though high EGT CEL kicked on and had it disable it ).

 

With that out of the way, I did two pulls with about .3 miles between them on the same 55mph road. Both had exactly the same IAT's and even post IC IAT's, I didn't do any more pulls since someone got pulled over and I didn't want to attract attention to myself.

 

Ambient: 33F

http://legacygt.com/forums/attachment.php?attachmentid=261954&stc=1&d=1518668340

 

Post IC IAT's start at 35.8F and by redline are at 51.5F. That's a 15.7F rise, that's more then it rose at 47F (odd!). What's interesting is, temp rise above ambient was still around about 20F. More precisely: 18.5F rise at 33F where as before at 47F ambient first pull was 17.4F and by fourth pull it was 22F above ambient. Now I'm anxious to do the same kind of testing at 60-80F, just to see if it's always ~20F above ambient peak, or if it exponentially increases.

 

Also, while 20f temp rise is a lot, it's still relatively small in HP speak, 2 BHP on NA cars, should be more on turbo cars, but hard to say. Though, Out of curiosity I went down the rabbit hole to find out how much power is actually to be made with a 14F cooler temp...

 

Here is the 47F and 33F pulls side by side, only corrected for barometric pressure and fuel weight (ignore the self-esteem spike at the end of the 47F pull)

http://legacygt.com/forums/attachment.php?attachmentid=261875&stc=1&d=1518535548

 

Basically not much, if I correct the temp, then 33F run will loose about 10whp.

 

What about the other factors like boost and timing you say?

 

33F runs a little more boost:

http://legacygt.com/forums/attachment.php?attachmentid=261871&stc=1&d=1518535548

 

33F cools the MAF wire a good bit more, causing the engine load to rise by 0.10g/rev:

http://legacygt.com/forums/attachment.php?attachmentid=261872&stc=1&d=1518535548

 

Finally the Total Ignition timing comparison, 33F runs less timing due to higher g/rev, but then IAT timing comp adds it in and makes the total timing be a hair more.

http://legacygt.com/forums/attachment.php?attachmentid=261874&stc=1&d=1518535548

 

Now here is a new metric. I recently created a new logger gauge, fuel consumption in pounds per hour, I made it to calculate my BSFC (but man it's a pain since Virtual dyno can't export HP numbers to CSV). Colder weather requires a good bit more fuel to keep AFR's happy.

http://legacygt.com/forums/attachment.php?attachmentid=261873&stc=1&d=1518535548

Edited February 15, 2018 by covertrussian

[covertrussian]

Created a thread to have all of the intercooler IAT testing in one spot and have others post their results too: Post Intercooler IAT Testing

 

Next, I wanted to do the same kind of testing against my G20, which has an FMIC. Engine Bay pic, notice that intake is a "warm air intake" and aluminum piping is exposed to hot engine bay air.

http://legacygt.com/forums/attachment.php?attachmentid=261953&stc=1&d=1518667782

 

First pull of the night: Post intercooler temps start at 59F and hit 68F by redline, which is 9F of rise and 10F over ambient. What's interesting is on decel it instantly goes down to 64F.

http://legacygt.com/forums/attachment.php?attachmentid=261950&stc=1&d=1518668316

 

 

Next the 2nd gear stress test. These were done after three 2nd gear pulls, IAT's wouldn't go down to ambient no matter how much I drove, they were stuck at least 10F above ambient after those three initial pulls.

 

IAT start at 69F, 1st pull: 80F, 2nd: 86F, 3rd: 91F, 4th: 95F, 5th: 96F, 6th: 100F, 7th: 100F, 8th: 102F. That's 33F rise from start to 8th pull, and 44F rise over ambient.

http://legacygt.com/forums/attachment.php?attachmentid=261951&stc=1&d=1518668316

 

Next, I dug up an autocross log from one of the laps during a hot 86F event. Idling I would get 134F, but during the lap it actually dropped down to 120-122F, then back to 134F while waiting for next turn. Even during that hot day, IAT's were only 36F higher then ambient.

http://legacygt.com/forums/attachment.php?attachmentid=261952&stc=1&d=1518668316

 

 

FMIC vs TMIC Thoughts:

• Even with a proper FMIC setup and less boost, I still had 10F rise in temp by redline at 58F.
  
• The 2nd gear stress test still pushed the temps into 100's even with a non-engine bay mounted intercooler.
  
• Legacy's TMIC drops temps MUCH faster and can get back down to ambient temps, no matter how much I drove around G20's temps would still be 10F above ambient. This could be because the Legacy has a proper cold air intake, while G20 has a warm air intake. Also G20's IAT sensor in an aluminum pipe inches from throttle body, so it might heat soak.
  
• Maybe TMIC is not so bad....
  

 

utc_pyro gave me his logs and I graphed them out too, which shows AVO TMIC actually being inferior to the stock TMIC (we have pretty similar overall setups too).

http://legacygt.com/forums/attachment.php?attachmentid=261919&d=1518631987

 

I'm now pretty convinced that the stock Intercooler is not the issue that's preventing me from gaining any more power. Maybe a big 16g is the only mod that our cars need? No need for fancy headers, Intakes, Exhausts. Just slap the turbo on a stock setup and you'll make the peak power with a proper tune

[covertrussian]

Weather finally hit a rather warm 75F today, seized the opportunity to test the intercooler temps on stock intake and on my custom intake.

 

Doing a single 2nd gear pull here are the results:

 

Custom CAI listed First / Stock Intake Second

Ambient Temp: 74F / 76F

Pre-Turbo IAT: 81F / 81F

Post IC Start IAT: 80.3F / 80.7F

Post IC Redline IAT: 90.2F / 91.8F

Post IC IAT Rise: 9.9F / 11.1F

Post IC IAT Over Ambient: 16.2F / 15.8F

 

Stock Intake single pull graph:

http://legacygt.com/forums/attachment.php?attachmentid=262195&stc=1&d=1519184410

 

 

Custom CAI listed First / Stock Intake Second

Ambient Temp: 74F / 75F

Pre-Turbo IAT: 81F / 82F

Post IC Start IAT: 79.9F / 81.4F

Post IC Highest IAT: 116.8F / 118.8F

Post IC IAT Rise: 36.9F / 37.4F

Post IC IAT Over Ambient: 42.8F / 43.8F

 

Stock Intake Graph:

Stock Intake: Start 81.4F, 1st 98.0F, 2nd 107.7F, 3rd 112.5F, 4th 111.2F, 5th 113.2F, 6th 117.4F, 7th 118.8F

http://legacygt.com/forums/attachment.php?attachmentid=262196&stc=1&d=1519184410

 

 

Overall pretty pleased with how well stock intercooler works and sheds heat, it would go from 116F to 100F in seconds, then would drop to 90F in about 30second and to 80F in another 30 seconds. After doing a similar stress test on my Bar & Plate FMIC, which is heavier and more efficient but took ages to drop in temps, I gotta say my hate for Tube & Fin intercoolers along with my hate for TMIC's is diminishing!

 

Seeing how slow bar & plates shed heat, makes me wonder if they should be used in TMIC application at all. They still will heat soak, but take much longer to drop the temps. This makes me want to retrofit and test test an STI one TMIC or another nice tube & fin TMIC.

Edited February 21, 2018 by covertrussian

[utc_pyro]

it looks like your CAI did intact help your post IC temperatures. The IC will mask a lot of the gain percentage though, and that shows up here.

 

Interesting that you’re seeing zero power gains, as both your boost pressure was higher and post IC temperatures lower.

 

For efficient intercoolers, the newer STI ones are considered the best you can get from the raw efficiency and pressure drop standpoint. You just have to change a lot of parts and build a custom splitter to get that efficiency. STI style turbo compressor, intake manifold, throttle body, BOV, BOV hose, and MAP sensor at the very least.

Edited February 21, 2018 by utc_pyro

[covertrussian]

it looks like your CAI did intact help your post IC temperatures. The IC will mask a lot of the gain percentage though, and that shows up here.

 

I would say they performed identically, temp went up by 2F between CAI test and stock intake, which accounts for slighter higher end results

 

Interesting that you’re seeing zero power gains, as both your boost pressure was higher and post IC temperatures lower.

 

I actually had a decent power drop today, even after accounting for temp and barometric pressure, I think I finally found the Achilles heal of my car, it's temps over 50F

 

Here are 2nd gear pulls from same road: 30F, 50F, and 76F. These are not temperature corrected, only weight and barometric pressure.

http://legacygt.com/forums/attachment.php?attachmentid=262203&stc=1&d=1519190289

 

Now onto the ambient temp correction, this is usually the great equalizer. In this case 30F pull has much more torque. Also notice the slower spool, though that might be from me lunching it slightly differently.

http://legacygt.com/forums/attachment.php?attachmentid=262204&stc=1&d=1519190289

 

Finally here is the Engine load vs IAT graph, notice how engine load stays about same between 30F and 50F, ignore the big dip at 3k-4k, that's because I started running less AVCS in those ranges, back to back testing found no power gain form having AVCS there, just more knock.

http://legacygt.com/forums/attachment.php?attachmentid=262202&stc=1&d=1519190289

 

Since engine load is way down for 76F pull, the timing it ran was also different. Adding 2* to the whole map made it loose 6whp by redline and removing 2* made it loose a 16whp by redline, so my timing is pretty optimized for the temp/engine load range.

 

 

For efficient intercoolers, the newer STI ones are considered the best you can get from the raw efficiency and pressure drop standpoint. You just have to change a lot of parts and build a custom splitter to get that efficiency. STI style turbo compressor, intake manifold, throttle body, BOV, BOV hose, and MAP sensor at the very least.

 

Maybe I should have gotten an STI style turbo afterall. I do have a spare VF39... As for manifold, I have a couple 2.5i manifolds and an 06 WRX manifold now

Edited February 21, 2018 by covertrussian

[covertrussian]

With all the tuning and mod testing my rotors became "warped" (uneven brake pad deposits). Rebedding them in worked for about two weeks, but when I had to hard brake again the death shakes were back.

 

Despaired with how long the Centric's lasted (3.5 years and 36.6k miles), thus I started looking at OEM Subaru ones. While checking part numbers I noticed that 2005's and 2006+ had different part numbers, so I ordered a set of each (in reality my Outback needed rotors too, and uses the same 2006+ part number).

 

More detailed comparison can be found here. I'll post here abridged version of the differences, I wont bore you with the install pics, everyone should know how to do them by now

 

2005 Rotor (26300AG001) vs 2006+ Rotor (26300AG02A)

http://legacygt.com/forums/attachment.php?attachmentid=262350&d=1519532794

 

2005's came in a nice air tight bag while 2006+ were just in the box, hence the rust...

http://legacygt.com/forums/attachment.php?attachmentid=262351&d=1519532794

 

2005's have slightly different vent holes too

http://legacygt.com/forums/attachment.php?attachmentid=262369&stc=1&d=1519533690

 

Centric's for comparison:

http://legacygt.com/forums/attachment.php?attachmentid=262368&d=1519533258

 

Here is where they really differ

2005: 21.4lbs (don't mind the paint chop, digital camera hates LCD display)

http://legacygt.com/forums/attachment.php?attachmentid=262356&d=1519532794

 

2006+: 22.8lbs

http://legacygt.com/forums/attachment.php?attachmentid=262365&d=1519533258

 

Centric's are also almost 1lb heavier at 22.2lbs

http://legacygt.com/forums/attachment.php?attachmentid=262367&d=1519533258

 

Both of them are 29.7mm thick, which means Subaru found another spot to shave weight in. This is where I started comparing all other things. Where they really differ is the thickness of the front surface vs rear surface

 

2005: Front 8.3mm, Rear 8.5mm

http://legacygt.com/forums/attachment.php?attachmentid=262360&d=1519533258

 

2006+: Front 9.6mm, Rear 9.8mm

http://legacygt.com/forums/attachment.php?attachmentid=262364&d=1519533258

 

Overall I think the 2005 rotors are nicer quality (and not rusty!). Plus they are much lighter, not only do they save you 2.8lbs in unsprung weight , they save you that much in rotational mass. I think these might be a great upgrade for those looking to shave every last bit of unsprung weight. Aluminum control arms cost much more and save you less in unsprung weight.

Edited June 4, 2018 by covertrussian

[kzr750r1]

Thanks for the details as always. Have been considering a LGT break upgrade or a 4 pot upgrade for the outback. This adds more to consider making the shift.

[covertrussian]

Thanks for the details as always. Have been considering a LGT break upgrade or a 4 pot upgrade for the outback. This adds more to consider making the shift.

 

If you decide on the 4 pot, I'm gonna have a set from a 2006 WRX for sale soon

[StkmltS]

I thought I fixed that caliper display for you... still showing metric.

[akarras5]

pretty cool!

[covertrussian]

I thought I fixed that caliper display for you... still showing metric.

 

It's old, beat up and stubborn like that, I'm waiting for it to die before I starting using my nicer metal one, which even has capitalist fractions!

[Guest]

For efficient intercoolers, the newer STI ones are considered the best you can get from the raw efficiency and pressure drop standpoint. You just have to change a lot of parts and build a custom splitter to get that efficiency. STI style turbo compressor, intake manifold, throttle body, BOV, BOV hose, and MAP sensor at the very least.

 

I went with 08 intake manifold+tmic swap long time ago, but never bothered with custom splitter. Wonder how much I am hurting the IC efficiency...

[covertrussian]

More Vacuum Leak Fixing

 

I noticed that my A learning was back up to ~10% again. I tugged on the blue tee hose and even though it's clamped it popped off easily.

http://legacygt.com/forums/attachment.php?attachmentid=262630&stc=1&d=1520060628

 

Dug up another Nissan hose from the spare parts box and had to stretch it, thanks Subaru for your love of varying inner diameters in one hose. After lubing & forcing the hose on and clamping it, I was giving it a very good tug and it was pretty on there.

http://legacygt.com/forums/attachment.php?attachmentid=262631&stc=1&d=1520060628

 

Sadly that didn't fix the A learning being at ~10%. I sprayed every accessible hose or connection and I still couldn't locate the vacuum leak. I'm begging to wonder if it's fueling related now.

 

 

Pre-Intercooler Temp Sensor

 

While the intercooler was off for vacuum line fixing, figured I would add a pre-intercooler temp sensor to see efficient the stock intercooler is.

http://legacygt.com/forums/attachment.php?attachmentid=262632&stc=1&d=1520061567

http://legacygt.com/forums/attachment.php?attachmentid=262633&stc=1&d=1520061567

 

Here is both of the sensors installed, I can alternative which one it's connected to depending on what data I want to log.

http://legacygt.com/forums/attachment.php?attachmentid=262634&stc=1&d=1520061567

 

 

Stock Intercooler Temp & Efficiency Testing

Now here comes the fun part, actual data! These tests were done with the stock intake installed.

 

First will start off with a stress test, at 40F ambient, pre-intercooler temps hit 180-205F temps!

http://legacygt.com/forums/attachment.php?attachmentid=262628&stc=1&d=1520060281

 

Next is a solo 2nd gear pull with bunch of data graphed to show relations between each other. Keep in mind that since I couldn't log pre and post IC temps at the same time, I did two separate pulls and combined the data together, it's not gonna be 100% accurate but close enough.

http://legacygt.com/forums/attachment.php?attachmentid=262629&stc=1&d=1520060281

 

With that data logged I can now take it a step further and calculate the intercooler efficiency.

 

Formula: Intercooler Efficiency = (Pre IC Temp - Post IC Temp) / (Pre IC Temp - Ambient)

 

Ambient temp was at 39F, but notice the MAF IAT, even though it was cold and windy stock intake refused to go down to ambient temps like my cold air intake would.

RPM	PSI	IAT	Pre-IC	Post-IC	Diff Efficiency
2090	1.29	48	75.8	49.5	26.3	71%
2163	1.54	48	75.8	49.5	26.3	71%
2201	1.8	48	75.8	49.5	26.3	71%
2232	2.09	48	76.5	49.5	27	72%
2332	2.33	48	76.5	49.5	27	72%
2310	2.64	48	77.2	49.5	27.7	73%
2402	2.94	48	77.2	49.5	27.7	73%
2430	3.27	48	77.2	49.5	27.7	73%
2500	3.61	48	77.9	49.5	28.4	73%
2544	3.96	48	78.6	49.5	29.1	73%
2629	4.38	48	78.6	49.5	29.1	73%
2674	4.84	48	79.3	49.5	29.8	74%
2710	5.35	48	80	49.5	30.5	74%
2798	5.99	48	81.4	49.5	31.9	75%
2850	6.83	48	82.8	49.5	33.3	76%
2927	7.53	48	82.8	49.5	33.3	76%
2967	8.32	48	84.8	49.5	35.3	77%
3064	9.07	48	86.9	49.5	37.4	78%
3111	10	48	89.7	49.5	40.2	79%
3184	11.2	48	89.7	49.5	40.2	79%
3298	12.32	48	92.5	50.2	42.3	79%
3337	13.22	48	96.6	50.2	46.4	81%
3473	13.34	48	100.8	50.2	50.6	82%
3492	13.41	48	100.8	50.2	50.6	82%
3609	13.35	48	104.9	50.2	54.7	83%
3638	13.33	48	109.1	50.9	58.2	83%
3780	13.33	48	113.9	50.9	63	84%
3859	13.29	48	113.9	50.9	63	84%
3966	13.33	48	118.1	50.9	67.2	85%
3994	13.27	48	122.9	51.6	71.3	85%
4107	13.38	48	127.1	51.6	75.5	86%
4235	13.23	48	130.5	52.3	78.2	85%
4300	13.33	48	130.5	52.3	78.2	85%
4418	13.27	48	134	52.3	81.7	86%
4454	13.21	48	137.5	52.3	85.2	86%
4592	13.12	48	140.9	53	87.9	86%
4644	13	48	140.9	53	87.9	86%
4758	12.95	48	143.7	53	90.7	87%
4872	12.89	48	146.5	53.7	92.8	86%
4950	12.82	48	148.5	53.7	94.8	87%
5050	12.7	48	148.5	54.4	94.1	86%
5074	12.6	48	150.6	54.4	96.2	86%
5213	12.59	48	153.4	55.1	98.3	86%
5257	12.5	48	156.8	55.8	101	86%
5425	12.41	48	156.8	55.8	101	86%
5465	12.33	48	159.6	55.8	103.8	86%
5593	12.28	48	162.4	56.5	105.9	86%
5655	12.15	48	164.5	57.2	107.3	85%
5668	12.07	48	166.5	57.2	109.3	86%
5760	11.92	48	166.5	57.8	108.7	85%
5913	11.94	48	168.6	58.5	110.1	85%
5983	12.04	48	170.7	59.2	111.5	85%
5992	12.24	48	172.8	59.2	113.6	85%
6070	12.3	48	172.8	59.9	112.9	84%
6140	12.22	48	174.8	60.5	114.3	84%
6293	12.09	48	176.9	62	114.9	83%
6367	11.99	48	179	62	117	84%

 

Well there you have it, the stock intercooler is up to 87% efficient at removing charge heat on these solo pulls, that's flipping amazing. Now cold weather probably did help, but even at 77F I measured only 11*F post intercooler temp rise, which tells me it should be just as efficient still.

 

What about intercooler efficiency during a stress test?

Sadly since I can't log both pre and post IC temps, and I didn't think about doing a stress test with post-IC temps I'm extrapolating the post IC temps from my 44*F stress test.

 

Ambient: 40F

RPM	PSI	Pre-IC	Post-IC	Diff	Efficiency
2318	3.14	157.5	81.5	76	65%
2370	3.46	157.5	81.5	76	65%
2400	3.85	156.8	81.5	75.3	64%
2483	4.21	155.5	80.6	74.9	65%
2548	4.62	154.8	80.6	74.2	65%
2620	5.06	154.8	79.7	75.1	65%
2657	5.58	154.1	79.7	74.4	65%
2703	6.13	153.4	79.7	73.7	65%
2789	6.82	152.7	79.7	73	65%
2862	7.86	152.7	78.8	73.9	66%
2936	8.65	152.7	78.8	73.9	66%
2972	9.67	153.4	78.8	74.6	66%
3072	10.61	154.1	78.8	75.3	66%
3134	11.83	155.5	77.9	77.6	67%
3266	12.95	155.5	77.9	77.6	67%
3329	13.61	157.5	77.9	79.6	68%
3376	13.68	159.6	77.9	81.7	68%
3480	13.61	161.7	77.9	83.8	69%
3568	13.61	161.7	77.9	83.8	69%
3683	13.54	163.1	77.9	85.2	69%
3748	13.53	165.2	77.9	87.3	70%
3870	13.57	166.5	77.9	88.6	70%
3896	13.51	166.5	77.9	88.6	70%
4049	13.48	167.9	77	90.9	71%
4124	13.44	169.3	77	92.3	71%
4195	13.46	171.4	77	94.4	72%
4270	13.44	172.8	77	95.8	72%
4374	13.45	172.8	77	95.8	72%
4502	13.41	174.2	77	97.2	72%
4570	13.31	175.5	77	98.5	73%
4706	13.33	176.9	77	99.9	73%
4727	13.22	176.9	77	99.9	73%
4841	13.22	178.3	77	101.3	73%
4912	13.1	180.4	77	103.4	74%
5016	13.07	181.8	77.9	103.9	73%
5130	12.94	181.8	77.9	103.9	73%
5230	12.87	183.2	77.9	105.3	74%
5313	12.76	184.5	77.9	106.6	74%
5355	12.74	185.9	77.9	108	74%
5430	12.61	187.3	77.9	109.4	74%
5530	12.5	187.3	77.9	109.4	74%
5598	12.4	189.4	78.8	110.6	74%
5752	12.31	190.8	78.8	112	74%
5831	12.23	192.2	78.8	113.4	75%
5837	12.23	192.2	78.8	113.4	75%
5930	12.37	193.5	78.8	114.7	75%
6074	12.55	194.9	79.7	115.2	74%
6150	12.49	196.3	80.6	115.7	74%
6170	12.37	196.3	80.6	115.7	74%
6233	12.21	198.4	80.6	117.8	74%
6399	12.17	199.8	80.6	119.2	75%
6460	12.14	201.2	80.6	120.6	75%

 

These are slightly underrated since I used 40F as ambient instead of 44F (which costs about 3%). Even with a heat soaking stress test stock intercooler maintained up to 75% efficiency, that's still very impressive!

 

 

In conclusion, I believe these tests rule out the stock TMIC a the bottle neck in my setup, especially with the cooler running big 16g. I'm also now convinced that these cars only need 3 mods to make decent reliable and cheap power:

1. Stage 2 downpipe

2. Big 16G or similar @ 13psi

3. Tune

 

For about $1k you can safely and reliably increase your power by 100bhp, everything else is dress up and for noise.

Edited March 3, 2018 by covertrussian

[utc_pyro]

More Vacuum Leak Fixing

...

 

I sprayed every accessible hose or connection and I still couldn't locate the vacuum leak. I'm begging to wonder if it's fueling related now.

 

Did you check for internal leak in the break booster? These cars are getting to that age....

 

 

Pre-Intercooler Temp Sensor

 

.....

Keep in mind that since I couldn't log pre and post IC temps at the same time, I did two separate pulls and combined the data together, it's not gonna be 100% accurate but close enough.

 

 

If you want to measure both at the same time, move your post IC sensor to your MAF sensor input and pre-IC to the EGT input. It might actually be a good idea to do this, as you could hijack some of the stock cat protection logic to reduce turbine drive energy if it starts overdriving the turbo.

 

You'll need to disable MAF-IAT compensation, and change the IAT-timing tables. The former aren't used in the '07+ roms (even though in theory you should), the later would actually make your tune safer as it would compensate for heat soak before knock actually happens.

 

Edit: Also when are you going to get those TGV's out so you can check pre-IC pressure with that MAP sensor I sent you? With that you could calculate your compressor efficiency and really nail down what it's doing.

Edited March 3, 2018 by utc_pyro

[covertrussian]

Did you check for internal leak in the break booster? These cars are getting to that age....

 

I haven't, how would you check that? Brakes seem to work fine, actually the best at least since I've bought the car, thanks to new calipers, rotors, and pads

 

 

 

If you want to measure both at the same time, move your post IC sensor to your MAF sensor input and pre-IC to the EGT input. It might actually be a good idea to do this, as you could hijack some of the stock cat protection logic to reduce turbine drive energy if it starts overdriving the turbo.

 

You'll need to disable MAF-IAT compensation, and change the IAT-timing tables. The former aren't used in the '07+ roms (even though in theory you should), the later would actually make your tune safer as it would compensate for heat soak before knock actually happens.

 

That might be my end goal, for now I'm just data gathering but I don't think I'm gonna trust my quicksteel putty to last long enough to have to worry about long terming these.

 

Edit: Also when are you going to get those TGV's out so you can check pre-IC pressure with that MAP sensor I sent you? With that you could calculate your compressor efficiency and really nail down what it's doing.

 

I've been busy trying to get good mpg base tests, I was close but then I found that my gas pump underfuels if I fill up under 4.5 gallons. This inaccuracy goes through the roof if I use only 2 gallons. I can't figure out why, but it applies to my Infiniti too (Which got 44mpg after 80 miles, when normal would be 30mpg at that time). I've searched google left and right and can't seem to find anyone else having the same issue with gas pumps...

 

Back to TGV's, the plan is to do them during my paternity leave this month, where I'll have some day free time hopefully! But weather hasn't been nice so far, so might have to wait a little longer.

 

Also I'm thinking about using the TGV inputs for IAT temp sensors, but would have to figure out a lookup table somehow then.

Edited March 4, 2018 by covertrussian

[utc_pyro]

I haven't, how would you check that? Brakes seem to work fine, actually the best at least since I've bought the car, thanks to new calipers, rotors, and pads

 

I think you’d pull and plug the vacuum line going to it and check fuel trims. Or use a hand vacuum pump to do kind of a reverse leak down test.

 

Back to TGV's, the plan is to do them during my paternity leave this month, where I'll have some day free time hopefully! But weather hasn't been nice so far, so might have to wait a little longer.

 

Also I'm thinking about using the TGV inputs for IAT temp sensors, but would have to figure out a lookup table somehow then.

 

Congrats on the kid. To put your IAT sensor into the TGV’s you just need a pull up resistor between the signal and 5v+ wire. A 940-950ohm resistor will make the scaling the same as the IAT, EGT, and fuel temp sensors.

[covertrussian]

Edit: Also when are you going to get those TGV's out so you can check pre-IC pressure with that MAP sensor I sent you? With that you could calculate your compressor efficiency and really nail down what it's doing.

 

Decided to do a quick intercooler pressure drop check. I teed a spare LGT MAP sensor into the wastegate line and just moved the pigtail from sensor to sensor after doing a few pulls.

 

http://legacygt.com/forums/attachment.php?attachmentid=262742&stc=1&d=1520225489

 

Now that's weird, turbo outlet pressure is seeing less pressure then manifold. Interestingly enough, at 0psi on boost gauge, turbo outlet was at 4psi. This just shows how much of a restriction the intercooler is at even lighter loads. It's odd that is showing less at full boost, maybe one of the lines I used has a leak in it?

 

 

I think you’d pull and plug the vacuum line going to it and check fuel trims. Or use a hand vacuum pump to do kind of a reverse leak down test.

 

Not sure why I didn't think of that myself . Tried unplugging it, plugged it up with my finger, corrections stayed the same, so not the master cylinder.

 

Congrats on the kid. To put your IAT sensor into the TGV’s you just need a pull up resistor between the signal and 5v+ wire. A 940-950ohm resistor will make the scaling the same as the IAT, EGT, and fuel temp sensors.

 

Thanks and thanks for the info!

[utc_pyro]

Now that's weird, turbo outlet pressure is seeing less pressure then manifold.

 

Potentially Bernoulli effect at the WG pressure reference port? That's a pretty high linear flow area, but one wouldn't think it's that high of a pressure drop. Where is Flinkly? He’s an ME and shoul be able to answer this.

 

You sure the spare sensor is on good health?

 

Interestingly enough, at 0psi on boost gauge, turbo outlet was at 4psi. This just shows how much of a restriction the intercooler is at even lighter loads. It's odd that is showing less at full boost, maybe one of the lines I used has a leak in it?

 

Pressure drop across your throttle body when the turbo is at high enough speed to create boost. At part throttle sometimes my BOV even opens due to this.

 

In searching for MPG gains, this is an area you could improve. The energy going into compressing this air is being wasted.

Edited March 5, 2018 by utc_pyro

[covertrussian]

Potentially Bernoulli effect at the WG pressure reference port? That's a pretty high linear flow area, but one wouldn't think it's that high of a pressure drop. Where is Flinkly? He’s an ME and shoul be able to answer this.

 

You sure the spare sensor is on good health?

 

Which also means that wastegate would actually boost less at a less turbulent flow source. Maybe moving the WG feed to right after intercooler is not such a good idea .

 

Not 100% sure on sensor, I didn't want to be cutting zip ties on the side of the road, but will try it out in a few days.

 

 

Pressure drop across your throttle body when the turbo is at high enough speed to create boost. At part throttle sometimes my BOV even opens due to this.

 

In searching for MPG gains, this is an area you could improve. The energy going into compressing this air is being wasted.

 

Ah that makes sense, so placing the post intercooler map sensor would be more accurate before the throttle body at part throttle then.

 

As for BOV, I have a spare Greddy BOV, I need to make/buy a stock to Greddy adapter and then I can test this out for sure.

 

Scenario 1: It will increase MPG because air wont just circulate back to intake, which will increase the amount of air being forced into the engine reducing pumping losses thus increasing part throttle power too.

 

Scenario 2: It will decrease MPG, because now engine has to breathe through the restrictive intercooler and turbo increasing pumping losses, where as before engine could get easy air from BOV/inlet.

 

Scenario 3: Won't make a difference, the forces will cancel each other out.

 

I'm adding it to my list of things to test out eventually

[covertrussian]

Intake Depression (Vacuum/Pressure Drop) Testing

 

Intake Depression is basically vacuum that happens between the air filter and the turbo. This depression increases the amount pressure that the turbo has to work against, forcing it to spin faster (as if you are running more boost), which increases your pressure ratio.

 

Read more about intake depression and pressure ratios:

A Technical Discussion of Intakes and Turbocharging

COBB Tuning - 2015 STi Stage1+Intake Testing

Pressure Ratio by Garrett

 

 

To make smart changes, you first need to know how restrictive your current system is. You can plumb a manometer into different parts of the turbo inlet/intake and determine how much pressure drop there is from that point. You can then deduce which part of the intake track is causing the most amount of pressure drop (restriction).

 

Read up about intake pressure testing:

Ultimate DIY Automotive Modification Tool-Kit, Part 1

We Have a Record! (Notice how the Subaru intake is the most restrictive of them all!)

Intelligent Modification: Measuring Intake Restriction (Part I)

Intelligent Modification: Measuring The Intake Restriction on Project Lexus (Part II)

 

 

Testing

With the basic concept out of the way, I busted out my manometer (that I bought for HVAC pressure testing, nice to have a tool be useful in multiple fields).

 

Since I'm curious about intakes and not the whole inlet, I plumbed a line right after the intake ends in the inlet. The location matters in this case, when I plumbed it a little back and to the outside radius of the inlet, my pressure readings were lower. This is due to the Coanda Effect making the outside radius of the inlet have a lower pressure zone.

http://legacygt.com/forums/attachment.php?attachmentid=262822&stc=1&d=1520404515

 

Stock Intake & Newer Style Subaru Filter

First I did a baseline pull on the stock intake and stock filter. This registered a peak of 44.55 inches of water column (in H2O), in more familiar measurement that's -3.27 inHg (inches of Mercury).

http://legacygt.com/forums/attachment.php?attachmentid=262823&stc=1&d=1520404515

 

To put it another way, there was so much vacuum between the inlet (not even the whole turbo) and the outside, that it would raise a tube full of water 44 inches above the ground! Looks like our "non-restrictive" Legacy intakes beat even the Liberty's -30 inH2O record.

 

K&N Drop In Filter

Next I wanted to see how well the K&N drop in filter would flow. A whole 0.63 inH2O less at 43.92 inH20, yeah don't expect to see any performance gains from any drop in filter, the stock airbox is just too restrictive, AutoSpeed also echo's this: Don't Bother Changing the Factory Filter

http://legacygt.com/forums/attachment.php?attachmentid=262825&stc=1&d=1520405548

 

 

Removing the Milk Jug

To see if the stock airbox is just unable to bring enough air in from the outside, I removed the milk jug. This actually worked and reduced the depression by 8.84 inH2O, but still totaling a high 35.71 in H2O.

http://legacygt.com/forums/attachment.php?attachmentid=262824&d=1520404026

 

While this helped a good bit, there are still flow restrictions between the inlet and the outside. This could be due to the cordion style S bend between the MAF and inlet. I could find out, but it would require me to drill a couple holes in my stock intake box, I'm not too keen on doing that, but I could be persuaded though...

 

 

At this point I gave up on the stock intake and installed my latest iteration of the cold air intake. I do believe the stock intake could be improved by replacing the cordion S bend, but I didn't want to spend any money on a stock intake when I had a custom cold air intake built already.

 

Custom 3" Cold Air Intake

This intake has one 45* bend instead of two 90* bends, all smooth piping, and a flow stack with a K&N 6" Flat Top filter:

http://legacygt.com/forums/attachment.php?attachmentid=262835&stc=1&d=1520408761

 

Flow stacks help reduce pressure drop majorly in the HVAC world, in theory they should help us too. They also help with making the airflow be more stable for more accurate MAF readings.

http://legacygt.com/forums/attachment.php?attachmentid=262821&stc=1&d=1520400749

 

This intake reduced the depression/vacuum by a huge 23.21 inH20, totaling 21.34 inH20, which is half less then the stock intake!

http://legacygt.com/forums/attachment.php?attachmentid=262826&d=1520406108

 

 

CAI Pod Filter Testing

Out of curiosity I did test a bunch of other filters from the same inlet location:

 

http://legacygt.com/forums/attachment.php?attachmentid=262836&stc=1&d=1520409353

 

inH2O - Filter

21.34 - K&N Flat Top

22.08 - Amsoil 6" Flat Top (EAAU6065-EA)

21.27 - S&B 6" with Inverted Cone Top

 

3" Filters not requiring the dedicated Flow Stack:

23.56 - Amsoil 3" with Inverted Cone Top and built in Flow stack (EAAU3050-EA)

22.53 - eBay 3" with Partial Inverted Cone

 

To my surprise the dedicated flowstack didn't reduce pressure drop all that much, not nearly as much as in HVAC world, probably because the filter it self is a bigger restriction. But it does help stabilize the airflow before it hits the MAF, which I believe leads to more accurate MAF readings, which is why a lot of stock cars use flowstacks on stock MAF housings.

 

Before you give the ebay filter any special treatment keep in mind that it did better then the other 3" filter because they barely filter anything usually.

 

Final note, Amsoil filters are going to be more restrictive due to their dry flow design & superior filtering ability. I think it's fine to gain 0.74in of pressure drop to avoid using oiled filters that ruin MAF's.

 

MAF Pressure Drop

Knowing how much pressure drop the MAF produces is important, since it will tell you how much restriction you will just have to live with.

 

I didn't want to drill my bend or MAF housing, so I cheated and used a spare 3" pipe and tapped a nipple into it then just moved that pipe around.

 

Post MAF I saw 12.17 inH2O worth of pressure drop.

http://legacygt.com/forums/attachment.php?attachmentid=262827&d=1520406108

 

Pre-MAF I saw 9.34 inH2o, which means our MAF's pressure drop is only 2.83 inH2O.

http://legacygt.com/forums/attachment.php?attachmentid=262828&d=1520406108

 

That means the filter it self contributes to 9.34 inH20 worth of pressure drop, this can be verified by doing a run without a filter, though I'm not sure I'm that ballsy just yet.

 

Conclusion

Stock intake is VERY restrictive! You can improve on the stock airbox design by removing the milk jug and replacing the cordion after MAF S bend, though keep in mind these changes might require a retune still. These changes might be good enough for most, but for those that are seeking every bit of improvement, an aftermarket intake becomes a must.

 

Since I had the spare parts, it was cheaper to build the custom Cold Air Intake then buying an aftermarket S hose. I wouldn't mind testing a proper after MAF silicon hose though.

Edited March 8, 2018 by covertrussian