Perscitus

Nope, both still used. The ones facing up like you see in the pics are the secondary set (regular narrow band o2s). The primary widebands come off to a side just at the very front of each cat. You can see their associated bongs in the third pic in the first row of attachments (look at the inlet to the cat that's further away in the pic).

Custom header-back exhaust component pics. You can see the stock 1/1.5"ID piece that was removed and the 2.5" ID replacement in the last few pics. AutoMaster Al re-used all stock mounting points (hence the bracket), cats and heat shields. He also added a section of flexible braided tubing (typically used on FI exhausts). I think it's actively contributing to reducing exhaust noise under acceleration and chassis noise when going over road imperfections/potholes/bumps.

Turn up the bass! The tone and sound is 50% of the awesomeness of ItalianLegacy's Raptor SC + Header + Invidia setup. Inside the cabin its quiet, even under WOT (actually quieter than the catback + resonator only setup). The car purrs and the exhaust noise has moved back, just as ItalianLegacy's did when he added the headers.

On the topic of cats, Al AutoMaster, who fabricated my setup said the only thing we can do is either go full catless or much thicker, just as short cats with 2-2.5" IDs ODs. Hes thinking older diesel or gas truck cats. In essence, hes suggesting short and chubby cats with plenty of ughmm girth. Bigger as in longer is not better for our application. We just need as much surface area as we can get and a relatively short cat so it can reach operating temp quickly. Transferring the bongs should be no issue, especially if we just keep the wideband sensors and ditch the secondary regular o2s. Both can be tuned out if needed. Im likely staying with the stock cats. They looked very good at 55K. No clogs, no damage, clean

yup, 1080p wot pulls on the dyno uploading to youtube tonight.

I will have moarh pics and videos up by the end of the day. Uploaded.net and YouTube are giving me some tech grief for some odd reason. Here is a DynoDynamics plot from last night. Both with the headers installed (before and after tune). Power is up almost everywhere post tune, but not a great deal - maybe 10-15% as Raptor predicted, especially post 4K RPM - look how it opens up! This is expected behavior with headers, top-end power gain. Rob was also able to smooth out the funky torque dip between 3.3-3.5K RPM that our engines exhibit on all dyno runs. Notice how it was still present with the headers and without a tune (violet plot) and almost flat with a re-tune (dark blue)? But the real icing on the cake is the torque! Its marvelous and much higher than pre-header install. When Al got the stock headers off the car, he said they belong on a 1.6-1.8L Honda motors, not a flat 6. He suspects Subaru used this design to keep the power output down on our cars for all their legislative/marketing/target reasons. Word of caution! AMR Performance Rob needed to re-do a lot of the tune once we got the car back on the rollers! Without a re-tune the car had lost power (5-15hp) vs. previous tune, so effectively, as a bolt-on, without any tuning changes (except perhaps from bone stock, no tune) - adding the headers without a tune results in a power loss. Mainly an issue with running very rich and un-optimal timing. Rob was amazed at how much the headers changed the way he needed to tune the motor. More than he typically sees with turbo-backs on FIs or header installs on NAs. Once he made all the adjustments, the car picked up the expected gains you see below. Depending on the shootout mode used this is showing ~ 296bhp (221kW) / 340ft-lbf (461Nm)

The headers are 2.25" at the flange. Then we have 2.5" section of SS mandrel bent pipes running all the way back to the Magnaflow 3" catback with the stock cats welded back in (2.5" ID/OD) in stock location, a flexible section of mesh pipe and the Vibrant resonator moved even further upstream (sounds great this way, 0 drone!!!). AutoMaster reused all the stock mounting locations for the mid-pipe. Al is a welding genius.

Raptor headers, custom header-back exhaust (still using stock cats but with 2.5/3.0" ID piping throughout) installed and tuned. Thanks to AutoMaster Fabrication for the design, welding and bolt-up. Thanks to AMR Performance for the tune. Pics and videos uploading. Gains? Approx. 40HP and 100ft-lbf over stock.

Yup, no vibrational dampening happening in the EZ36D pulley. The H6 is internally balanced (enough) to allow this.

I like it. There is no power gain BUT the engine does climb up and down the RPM range easier and faster. Relatively easy to install, should take no more than 30 minutes. Best on a cold engine. I believe the price was somewhere around $200-240 shipped.

From what I saw and heard last month - it sounds godly! The headers are not loud at all, in fact they move the apparent exhaust tone further back, toward the mufflers. Not raspy at all. I think we'll see something rather special soon. I'll leave it to ItalianLegacy for the reveal.

Steven/Fredrik - are you guys talking about actual knock (events) or timing advance being pulled AS A RESULT of knock? Steven - when you mention something like '-7' that sounds like timing advance being pulled into the negative (timing retardation (lol) of 7 degrees). This can be a result of bad tank of fuel, or an aggressive tune vs operating conditions (combo of: temps/air flow/fuel quality, target AFRs, fuel trims). For example, our cars will default to pull about 1-2 degrees of timing for every 5-10F above 100F IAT. They will also pull timing much more with a tune targeting 93 octane fuel, and a tank of 87. + Usually not an issue, especially if timing advance stays 'in the positive'. Ideally while cruising on the highway it should be +30 to +40. At idle or low RPM, somewhere between +10 and +15.

Holy giant filter! Nice. Most likely no. Interestingly enough I just swapped it out and its helped to stabilize my AFRs, short/long term fuel trims, raise the average timing advance. However, your milage might vary. I will look to get a tune with this filter once the Raptor headers go on. They've been sitting in a box from Oz somewhere in CT for far too long!

Hey Steven, yes I can confirm the filter I recently switched to increases airflow over the Takeda supplied dry cone filter. I tried a Donaldson PowerCore filtering element from Volant. Up top and @ WOT, it outflows the Takeda element by 200-300cfms (25-37g/sec). Torque logs and live data tell me that the MAF sensor is supposedly seeing as much as 240-250g/sec now vs 220-230g/sec I've seen before. However this filter doesn't work well in all applications and intake designs. It can cause issues for both FI and NA (especially FI) applications when the MAF housing is located too close to the inlet. Under and over reporting air as well as outflowing the stock MAF 300g/sec limit. This can be scaled and interpolated on older Subarus like ours, but not the 2015+ cars yet. I believe Fredrik is running a longer (7"?) AEM Dryflow with a custom pipe in place of the airbox while DJMajickMan has a monster filter (9"+) sitting in his passenger wheel well, getting cold air from the lower grill to WRX bumper vent blow-by.

I will have to check this for you, maybe on the dyno because torque seems to only give me g/sec or cfm. I've switched back from the cfm to g/sec recently and similarly to stevenc1703, I see 230g/sec, 240g/sec. I believe I've seen as high as 250g/sec. With the same filter elemenet on a 2015 WRX, and a target boost of 19psi - we pegged the 2015 WRX MAF @ 300g/sec @ 4K RPM and above. Switched back to a standard cone filter and see about 200-210g/sec (@16-18psi, 4.9-5K RPM, WoT, 50mph). I believe the highest logged on the dyno was 250-260g/sec.

Nice setup, I wonder if those are the direct sensor outputs. For example, two AFRs would lead me to believe one is 'commanded' (what the ECU/tune requests and the other 'measured' (as approximated from the wideband and MAF/IAT/MAP readings) as a result of the instantaneous fuel trim changes. Good to see g/sec and C for a change! Sounds a bit strange, the idler pulley you pointed out (and the belt tensioner pulley) both have bearings which allows some play and wobble. You can replace the pulleys or just the bearings... plus while @ it, grab a Gatorback serpentine/accessory belt! Best $$$ I've spent on the car short of the RSB. http://www.subaruoutback.org/forums/attachments/problems-maintenance/21279d1319842077-h6-serpentine-belt-pulley-tensioner-replacement-moya_h6belt_diagram.png http://forum.liberty.asn.au/viewtopic.php?f=16&t=26323

Its possible yes, nstg8r and McDowell would likely know for sure. I wouldn't be surprised if the MAF readings are 'incorrect' (unscaled) in raw form as the sensor itself sees the airflow. But the rest of the sensor outputs I usually look at should be correct: ambient, intake, coolant temps, short/long term fuel trims, wide-band voltages, regular o2 voltages, timing advance. Supposedly yes - although they are not sent using the same protocol as the vanilla OBDII ISO 512 baud stream. The only issue is no one has or is willing to share the details needed to configure them.

No, you got it stevenc1703. With Ecutek you have a few options: a. local tuner tune - no hardware needed, pay for license+tune/dyno time only (use tuners Ecutek suite and cable, dyno tune from scratch or using a base tune with same/similar mods as... base) b. e-tune - hardware optional, pay for license + e-tunes, potentially cables unless you can use and send back (buy or 'borrow' the ecutek cables to dump the stock rom, flash/re-flash e-tune revisions) c. tuning suite kit - hardware and software, pay for license + cables, Ecutek software, RaceROM, tuning suite (grants access to DB of many base maps across a few makes, RaceRom where applicable, allows for customer tunes) Personally I've gone with a, both with Visconti and now AMR. I don't have any hardware, cable or kits and never had to pay for them. I only paid for the license and the dyno tune sessions. Subsequent re-tunes are usually much cheaper (typically within the price range of a dyno pull session). The screenshots above are from Torque Pro, just reading the generic (non-Subaru ISO) OBD II PIDs. I'm logging this data using OBDLink MX and phone.

Nice! My MAF readings off of Torque (converted to cfm) show a supposed peak @ ~1700cfm (802g/sec). WoT/100% load, 55-75mph. Typically in the 400-800-1000 cfm ranges (189-378-472g/sec). MAF (and a few other related sensor read-outs) have been re-scaled with the most recent Ecutek tune, post intake install. Could this be skewing the readings? If your turbo'ed EZ36D is pumping 420g/sec, then there is no way my NA motor is reaching 472g/sec, let alone 802g/sec.

Here are the OBDII sensors my OBDLink MX hands down to latest Torque Pro, and the sensors I typically log, in case anyone wants to mirror the setup:

Hey Giovanni, it might be the TCM stepping in at a particular ATF temp. Even with an ATF cooler, with the amount of torque you're putting down, it can be reaching temps or seeing valve body pressures that put it into some type of limp mode. I doubt it would be your engine oil temp causing this behavior.

Thanks Rob, sounds like we're seeing the same readings. How quickly did it drop south of 100F and at what speed? If you were able to stay below 100F on a day like today (or tomorrow) doing 20mph or at part throttle 40-50mph cruising, then that's different from what I'm seeing from my IAT.

Logged ~30 min trip back home yesterday afternoon. Ambient temp was ~ 88F, humid and cloudy. Mostly city driving, some short stretches of parkway. Medium to heavy traffic. MAF/IAT cleaning seems to have helped sensor sensitivity, but still seeing relatively warm air passing through the intake tract. At least its warmer than I'd expect. On average, the temperature difference between ambient and intake for the whole trip was 21.4F. Giving us an average intake air charge temperature of 109F on a 89F day. Lowest reading was 95F and it typically managed to come down to <100F above 55mph. Highest reading this time came in at 130-132F. Mostly before warm start for this run. I guess that's reasonable. I hope I can get the average temp diff to come down by about 10F with the reflective wrap. And more importantly, help the IAT stay closer to ambient more consistently. Side note - Under WoT, full load and about 3.3-3.5K RPM, I'm seeing MAF rates of 875-950cfm! The highest I've seen was 1,200-1,300cfm, different days/logs.

Ecutek tuned by AMR PS - Latest update from Rob is that he's found and decoded a whole bunch of new tables in our ROMs, including some that directly and indirectly affect the TCM. He's sent them over to Ecutek Dev HQ in the UK, asking them again that they define them for us and add to RaceRom features for us. With RaceRom support he would also be able to re-scale IATs, MAPs, etc for us, give up map switching, and potentially some form of launch control. The quest to control the 5EAT better was inspired by watching your Laguna Seca runs Fredrik. Rob watched them over and over again and he knows he could get the transmission to play nicer when we need it most.

Yup. I will report back when the wrap comes in and I wrap the intake tube, airbox, maf housing. Thinking of running some custom piping to the hks funnel too and out by the radiator on the left. Get more cooler air up there.