Perscitus

The tranny bolts right up supposedly. PnP, sans the 5EAT wiring harness and TCU disable. Go back a page for a link to NASIOC EZ36D swaps into previous-gen STIs. They used 6MTs and turbo'ed the H6s. AMR has done a good bunch of these swaps too, for DD and race/auto-x customers.

My oh my! Such a tease.... http://shop.amrperformance.com/AMR-Performance-teams-up-with-Raptor-Superchargers_b_11.html

Cold and semi-warm start video of the headers/header-back exhaust: Warm start and some revving @ 2:06. Plus nstg8r's brake flasher mod toward the end of each section (when I applied the brake).

http://jp-carparts.com/subaru/figlist.php?maker=subaru&car_baseid=F23000&type=B14&cartype=005

Some 'blast from the past' (circa 2008 and 2014 update) inspiration for all of our H6 brethren... ItalianLegacy I'm looking mainly at you boy! Then again, who am I fooling? You don't need any further inspiration. The beast is coming along very nicely. http://forums.nasioc.com/forums/showthread.php?t=1437246 http://forums.nasioc.com/forums/showpost.php?p=20928237&postcount=22

A few quick shots just before first snow:

I had considered that but decided to pass... for now. Marginal if any power gain on our NA platform and not enough engine bay heat with it as-is. Maybe next spring or summer. On the flip-side, the P&L Motorsports downpipe my brother picked up a weeks or two ago for the 2015 WRX will be ceramic coated right after fabrication.

Some 'fresh' pics of the baked-in headers and revised exhaust. Headers outlet ID 2.25" -> 2.25" SS section of pipe -> stock cats -> 2.5" SS section of pipe -> 3" SS braided mesh coupler -> 3" pipe with Vibrant resonator -> Magnaflow 3" catback

Found something potentially interesting (especially for Fredrick/Shrek as a 76 alternative): http://www.torcousa.com/torco_product/un_acce.html I might just have to give a 32oz bottle a try.

I've got some proof of that (note ItalianLegacy has plastidipped his beast blue for the video but you get the idea..) [ame=http://www.youtube.com/watch?v=D9tTJVT31U8]Subaru | Gymkhana WRX STI - YouTube[/ame]

ItalianLegacy can likely smoke a current/previous-gen M5 with the #s his beast is putting down. Power to weight ratio is now on his side vs the M5.

Sounds good. When you remove it to inspect for any warp/damage, consider re-installing it with the stock crank bolt. Just an idea. nstg8r, ItalianLegacy and Fredrik used all new/provided parts I re-used the stock crank bolt because I liked the bolt head better (its deeper with better defined grooves) Grab a gatorback belt (I have the part # posted up somewhere) while you're at it too. Its worth every $$$ and a good idea to swap out anyway if you're anywhere near 50K. The gatorback is better at handling any offset between the pulleys, wobble, etc.

Ill shoot a video of mine tomorrow evening so you can compare. Like nstg8r, I dont recall any play, wobble or warping. Did you end up using the stock or provided crank bolt? What about the bolt head cover washer? And when torquing the bolt cover back on did you do the standard x torque pattern? Stock or replacement serpentine/accessory belt? When I looked at ItalianLegacys setup last month I also didnt see any wobble when the sc beast idled or revved.

Very nice! The shift points and gear transitions looked/sounded very good - I can almost 'hear' the difference. Impressive!

- 2010 3.6R Premium - Graphite Gray Metallic - Male - 33 (feeling like acumenhokie) - by day: Market/Credit Risk Business Analyst (aka pricing monkey) by night: grease monkey by school: mechanical/mechatronics engineer - 1st Subaru (sealed the deal on brother's 2015 WRX and helped family pickup an XV) - a few mods here and there

And here is our actual throttle % translated from the OBDII reported TP @ Manifold values. For the EZ36D: Closed throttle reads 11.37 (2015 WRX ~12.55) Wide open throttle reads 80.78 (2015 WRX ~78.82)

I've 'created' two custom PIDs to use in Torque/Torque Pro. What they do is add the STFT and LTFT per bank of cylinders up and report as one %. This is useful both for a quick glance of what's going on, data logging (2 columns instead of 4 that need to be added up later) and to reduce on-screen clutter (1 element for 4 cylinder cars, 2 for our H6s - instead of all 4 earlier). Torque allows you to setup custom PIDs off of original/exposed PIDs by performing simple math operations on them (in this case addition of two known PIDs) so I did just that. Works like a charm! Rule of thumb to see if your combined fuel trims (STFT+LTFT per bank of cylinders) are 'within spec' (lol): While here, check if your Bank 1 and Bank 2 %s differ. You might see a fairly constant fuel trim diff ('offset') of 2 to 5% between Bank 1 and Bank 2 (with Bank 2% < Bank 1%). If they're dead on and respond the same to throttle input - lucky! - I'd love to see some logs! My fuel trims are and have always been ~+2.7% offset, regardless of mods, stage, tunes etc. With the average of Bank 1 fuel trim always about 2.7% richer than Bank 2. I also took the time to scan all the memory addresses on the regular OBDII interface using the 'Torque Scan' plugin - with the engine running and just the ignition switched on. I was unable to find any additional PIDs. This way we know there are no extra PIDs using the regular protocol. All the cool stuff is using the Subaru custom protocols on CAN, and we can't get to it without special tools/licenses to access them. Bummer.

This! And now I see where he's being kept from my apt window. Ugh. My brother who is another NYC-based doctor says this guy should be stripped of his licenses and I couldn't agree more.

These are about 8" long; 6" wide - sitting on my dinner table. The filter on the left that we're going to fit with AMR has a 5.5" outlet that will velocity stack down to 3" ahead of a straight section of high density cross-linked polyethylene. The filter element will live in a matching polyethylene barrel/bucket style airbox with a long section of polyethylene pipe leading down to the ram air ductwork we've setup with nst8gr. Dyno sheet - posted in this thread maybe 3 pages back now. Old dyno sheets are up in my build thread (which I have to update, slacking on that a bit).

And the mods continue.... secret hybrid intake project in the making:

Good question - this would be something AMR or any of our tuners would have to answer. Richer (unless pig rich) is definitely safer and for sc/turbo'ed tunes its recommended and more optimal. For NA, typically 12.8:1 to maybe 12.4:1 is the theoretical sweet spot for power. But how that vibes with the 3.6R (if running richer at WOT gives more power) is ?. I was only going by the measured vs commanded AFR difference. On my car, prior to the header install, the car always ran somewhat richer than what the ECU/tune was commanding. Say commanded AFR of 12.6, actual was in the mid 11s, commanded AFR of 11.9 - actual measured in the low 11s, high 10s even! Car ran fine though and made good power (butt dyno). With the headers, and latest re-tune, the difference between commanded and measure AFRs is just much smaller. So I take that as a good thing. Prior to the re-tune with the headers and new exhaust on, the car ran very rich and as a result lost power. By very rich I assume Rob was seeing 10.x:1 or maybe even 9.x:1. If I had left AMR's shop without a re-tune it would run but the power loss would be there plus my MPGs would likely reflect the rich AFRs.

Oh no, my previous tune and current one were/are fine. It was the intake R&D changes we made in-between tunes that created some transient pulled timing blips as you either got on or off throttle (2-5 data points out of say 200-400K data points). 2nd row, 3rd plot from the left shows before/after Timing Advance data across ~ 8 hours worth of logs: Check out the count of data points below 0 on the left side (very few, but they were there). And then compare to the right side - not a single one With the headers installed and Rob's latest re-tune, he was able to take advantage of not just the headers but the intake changes. Latest tune has 0 instances of negative timing (no matter what I do), in fact even under WOT (where timing can sometimes get close to 0, its now never lower than +4.5/+5 degrees, which means there is advance even under full load, high RPM, WOT.... this is great and means that there is potentially even more room for improvement.

Running great Tim. Here is some feedback I shared with the tuner (Rob @ AMR): · Fuel Trims cleaned up nicely - within 0-10 (usually 0-5%) for Bank1 and -5 to +5% for Bank2. · No more excessively rich measured AFRs (vs. what's commanded) under WOT - nothing south of 12:1. · No more negative timing data points (not even any transient blips). I've logged the trip up to Danbury (before) and trip down back to New York (after) to get a direct comparison of logs. I've also logged some more Sunday and Monday to see the after trends continue. Fuel Trims - much improved (even vs just adding the honeycomb airflow straightener) 15% throttle and up, they’re almost always within the +/-10% bracket, both banks. Looking per bank at both LTFT and SFTF, you can see how the STFT adjustment swings are smaller. Far fewer cases of the larger negative STFT swings. If anything I see a few more positive STFT swings. Even fuel mileage seems up (I’ve gone through half a tank with some spirited driving since leaving Danbury and I’ve clocked 250 miles thus far; that’s ~32.3mpg!) Of course I’ll monitor this more over a few tank-fulls. *We forgot to tweak an inherent fuel trim offset we see between Bank 1 vs Bank 2. Ecutek software usually commands changes to Bank 1 only but there is a way to force a hard adjustment factor to Bank 2 for some compensation. Its decreased a LOT from pre-header tunes to a point that it’s almost gone. But, I wonder if getting rid of it would net more power or just a sign of an injector performance difference in one or more of the cylinders between the banks. AFR - also improved (leaner under load). I don’t see any more dips into 10.x:1 rich territory. 15% throttle and up, the richest it got was 11.93:1, typically in the 12.5:1, very close to what's commanded. Timing Advance – has generally gone up. And no more pulled timing! In fact, I see positive timing advance even under WoT. While crusing, I can typically see +35-40 degrees now. Plenty of +40-45 degree data points. No negative timing values! Under WOT the lowest I see is +4.5/5.0 degrees! Seems like it hit that a few times. Before the headers and re-tune, we had a few random -5.0 degree data points. MAF – no issues. Still rock solid at idle. Since you rescaled the sensor (changed transfer function?) I’m seeing lower g/sec ‚readings‘. With the new scale the highest I’ve hit is ~208g/sec. IAT vs AAT Temps are still good – now that its gotten colder, I get within 2-5F of reported ambient while crusing. After a trip, when the engine heat soaks, the intake only reaches about 110F (used to be 140-160F) and the intake temp drops back down below 70-80F much much faster once you get going. Data log plots attached: This is all >15% Throttle Position to remove low throttle, low load noise. Legend: SATURDAY AM (BEFORE HEADER INSTALL): 1st section, to the immediate left of the red vertical line SATURDAY PM (AFTER HEADER INSTALL): 2nd section, to the immediate right of red vertical line SUNDAY AM/PM (AFTER HEADER INSTALL): 3rd section, last continuous section of data toward the right (180K to 225K on x-axis).

Found some goodies on the AU sites: (pause and play @ 1.07, then skip ahead to 2:57) [ame=http://www.youtube.com/watch?v=lStnaJeNgzE&feature=player_embedded]Subaru Liberty SC 2 - YouTube[/ame] http://forum.liberty.asn.au/viewtopic.php?f=65&t=22082&start=210 http://forum.liberty.asn.au/viewtopic.php?f=65&t=17494

lol, I hate myself too. I should take a reverse mortgage out against the non-existent mortgage I still don't have. We cut the collars off and re-welded new ones in, or rather cut as ItalianLegacy suggested (effectively the second weld line you see on each cat; the elements inside are even further/deeper in so they are not damaged if you do this) and then mated that up to the 2.5" pipes, welded all around.