fahr_side

This guy also has the connectors...

The original motor mount pads are still there on the later block and the 10~ front one just bolts on. I've seen a '10 block in an '06 WRX.

Thanks. Would be good to know. The EDM models are definitely missing that.

I really can't, as I'm not familiar with US rates. As I see it the CP ratio is best at stage 2. A downpipe and a good tune is a really good deal. After that, 18G + meth, or 18G on corn, stock intercooler and exhaust, good tune, keep it to 18psi and love you long time. Past that... better have an understanding spouse because the hp/$ ratio gets silly. Dude, you're the original big turbo pioneer, how do we repay you?

Thanks guys, the 'likes' are much appreciated. I'm here all week, try the fish! Much of what I am getting at in the schemes laid out here is how various items in the hardware landscape play together and how important balance is. You open up the exhaust flow with the cams and then it becomes useful to increase flow on the cold side. With the stock exhaust cam in place there isn't any point in using a >18G compressor. The question will then be, will the little P18 exhaust housing then become the choke point? My gut feeling says the 18G would still be the best compromise between spool and flow given the same exhaust side, but that you'd get the best out of the 18G vs. most of it. I've tuned plenty of late WRX models but none with the GS cooler. From what I see it's nice part and all but as I mentioned, the BR/BM is so limited on ambient airflow most of the gains there are from increased mass to dump heat into, not from the intercooler being that much more efficient. Please read this whole link for a better look at how intercooling works, especially in an environment where direct heat exchanging efficiency isn't that great. It's pretty sad really that Subaru passed up the opportunity to do a factory FMIC while the turbo was right there behind the bumper.

I find the blow-by only gets bad enough to need to use an AOS past 400whp, as long as we're talking healthy engines.

The oil pan on the 2010 has an extra bolt but the '05 will bolt up anyway. Need to look at the turbo oil return as well.

So after some requests I’m laying out what I’ve found in the 5th gen engine and ancillaries that make it unique, and what extra challenges there are in the mod path vs. the well-known 4th gen route. I think all this information has been posted in one form or another, but not in one place. What’s different in the 5th gen engine? It’s actually more like the EJ257 in the GRB / GVH STi than the previous model Legacy GT. Dual AVCS system Shorter duration cam with 9mm lift vs. 10mm Oil scavenge pump at rear of passenger side head, driven by intake cam E25 heads with larger 57cc chambers On AT models the trans fluid cooler is not integrated into the radiator, the heat-exchanger is a separate unit Oil to water cooler above oil filter deleted on EDM models. New short, EL header Unique to this platform low-mount VF-54 turbo Unique to this platform charge pipe to common-with-new-WRX TMIC Of course there are other things like the oil pan shape, number of bolts holding it on, etc. etc. but from a performance perspective this is the important stuff. There is also a great thread here listing many parts from other models which fit the 5th gen. From a purely performance perspective here’s a list of the weaknesses in the stock hardware on this platform: Despite the addition of dual AVCS the exhaust cam lift is 1mm shorter than either previous single AVCS EJ255 units or any EJ257 units Like the older VF-40 and VF-46 units, the turbine and compressor wheels are very small Like the older VF-40 and VF-46, the turbine housing is a P18 with a throat area around 7.5cm2 Unlike the older VF-40 and VF-46 you cannot just replace it with a VF-52 or any number of aftermarket bolt-up turbos going up to TD06-20G sizes. The hood scoop is very short indeed so despite the intercooler having more surface area than the 4th gen it’s starved of air flow to cool it. So, here’s a list of the roadblocks on the road to power as I see them: 1. Stock cat. We usually refer to the cat pipe as a J-pipe now, rather than downpipe, due to the geometry. The stock cat is very restrictive. Pushing boost past about 16.5 or 17psi creates very high EGBP (exhaust gas back pressure) which we can evaluate via exhaust gas temperature. High backpressure means the cylinders don’t get scavenged of exhaust gas properly which negates the benefits of squeezing more air into them. Stage 1 tunes top around the 230whp / 305wtq mark, mostly due to the restriction in exhaust flow and limited boost that can be run with reliability. 2. Stock BPV. The stock bypass valves leaks pretty badly, sometimes from boost pressures as low as just 11 or 12psi, sometimes not until 14psi. Leaks in the BPV make the turbo do all the work of compressing the charge air just to have it escape and expand again, cycle around to the intake path and get compressed again. Leaks make the AFR trend rich which is also bad for power. Leaks make the turbo work harder than it would otherwise to make the same boost pressure at the manifold, which is bad for turbo efficiency and builds more heat, and also wears the turbo more than it would otherwise. While the stock BPV does not put a hp cap on the car it certainly holds it back and makes whatever power you are getting very inefficient and potentially unreliable. If you’re raising boost above stock, you want to upgrade the BPV. Hat-tip to BigBopper for his scientific testing on leakage. 3. Stock TMIC. This isn’t as bad as you might think and there’s some truth in the idea that since the flow of cooling air to the TMIC is so limited, gains from upgrading the intercooler are somewhat limited. I can usually run a little more timing on cars with an upgraded TMIC but what’s more obvious is the extra time it takes to heat-soak them, which is probably simply a function of their increased mass. I feel 300~320whp to be about the limit for a stock cooler on gasoline while on E85 or with meth injection it’s okay to push this into the 330~350whp range. Just don’t expect to be able to boost away to your heart’s content without the intercooler become heat-soaked, and intake temps steadily rising. Anything you do to reduce charge temps will be rewarded, whether via water spray on the cooler core, water/meth injection, a bigger heatsink (intercooler) or whatever. The bigger cores on the aftermarket TMICs I’ve tested did offer less flow resistance and that did allow for less WGDC to run at the same boost pressures. That has to make life a little easier for the turbo. 4. Stock VF-54 turbo. I’m not comfortable pushing these past 10 or 11psi on pump gas at redline because the turbines just get choked anywhere beyond this kind of shaft speed. Note that this is not much more than what’s targeted in the stock mapping. Subaru basically fitted the smallest turbo they possibly could in this application to hit the power target they were looking for. Improvements in peak power on stage 2 cars come mostly from leaning out the very fat stock fuel curve, the reduction in EGBP from running catless or with high-flow metal cats, and then being able to run more timing because of the lower EGBP and EGT. Power limit is about 235whp on pump gas and torque is limited to about 330lbft on the stock headgaskets because pushing beyond 18psi is going to lift a head and pop the head gasket before very long. The other issue with the stock turbo is the wastegate. As many have found out, the flapper is sized ‘just’ large enough for the job on the stock hardware and tune but for example fitting a high-flow exhaust lowers backpressure enough to allow boost creep. It’s especially obvious in colder weather and with catless J-pipes. 5. Exhaust cams. As mentioned, these are weak compared to other EJ25x turbo models and definitely hold power down below what you’d see on say a 4th gen or WRX with the same turbo and supporting mods. In testing with various turbos what we found was that beyond a certain level we could not get any increase in intake air flow because the motor was simply choked on exhaust flow. That corresponded with what our 18G turbo could flow. Fitting a 20G compressor did not improve air flow, it just made the car laggier. The cams also make it pointless raising the rev limit any higher and the valve springs are also weaker than on other models making it potentially dangerous to do so. 6. Header. The stock design is not bad. It’s very short and it is pretty much equal-length. There are pretty sharp bends as it enters the turbo flange but that’s hard to avoid. There’s nowhere to weld a stub pipe on for EWG. Perhaps the worst feature is the turbo flange, which doesn’t mate to any other turbo than the stock VF-54. I have seen some results from an aftermarket header that had slightly larger primaries than stock while being of a similar design. They had higher boost threshold and more lag, but no power gain from that. It’s my opinion that if you’re going to the extent and expense of changing out the header it only makes sense if you’re going to a different turbo flange that allows more turbo choices. The other reason for going aftermarket would be if you’re plagued by boost creep issues on the stock turbo and don’t want to reinstall a stock catback exhaust to ameliorate that problem. An EWG will fix this but note it's a challenge to plumb the gates back into the midpipe, so you better like the noise. 7. Catback exhaust. The Y-pipe construction is pretty craptastic, featuring crushed pipes jammed into clumsy collectors. It’s not very large in diameter. Having said that the stock catback flows enough gas for 335whp and having tuned cars with a variety of catbacks we did not see any gains over the stock catback at that level. 8. Fuel system. The stock injectors will support up to about 300whp on pump gas. On E85 or with larger turbos you will be fuel-limited and will have to taper boost accordingly. With meth injection you can run up to about 335whp at 90 to 95% IDC. 9. Oil cooling. As noted, Subaru deleted the usual oil-to-water cooler on the EDM models. Even at Stage 2 I like to see an external oil-to-air cooler used on these. 10. Intake. We have not seen any sign that the intake was restrictive up to 335whp. I’ve tuned one Stage 2 car with a Blitz intake kit that did better than others but it may just be an anomaly. I’ve tuned cars with other intakes and saw no improvement at all, even on cars with 18G and 20G turbos installed. 11. Shortblock. With stock head bolts you are limited to about 18psi on pump gas before the heads lift and the gaskets fail. I’m also not keen on running the stock pistons past that point on pump gas as they are just as prone to cracking the ring lands as any other Subaru hypereutectic piston. Sensible options. By sensible I mean modification packages that meet useful power targets at reasonable cost. 1. Stage 1. The big stumbling block here is the cat pipe. Upgrade the BPV and get a good tune. If you can get E85 it’s well worth it, though you will probably need to upgrade both the fuel pump and injectors. I’d suggest 750 or 850cc would be enough for now but if you were thinking to go bigger later, go to 1000s. Any TMIC, exhaust, oil cooling etc. upgrades aren’t going to get you more power but may be a good investment for the future, and help with reliability. 230whp/305wtq is typical on gas. 2. Stage 2. Blockage point here is the turbo itself. As above, but add a high-flow catted, or catless J-pipe. Now you can raise boost to 18psi peak on pump gas and run leaner AFR and more timing. Note boost still needs to taper at high revs due to the size of the turbo. Stock injectors are adequate for pump gas but I’d back them up with a better fuel pump like a DW200 or a 300 if you think you may go bigger later. On E85 you will want the bigger pump plus 750 or 850cc injectors and again you may go to 1000s to future-proof yourself. 235whp/330wtq is typical. At this point I think a FMIC would definitely help. As mentioned above, the air flow to the stock TMIC is pretty limited and in longer sessions at WOT and high revs it’s going to get heatsoaked. Oil cooler recommended for EDM models. 3. Stage 3 on hybrid turbo. The limits here on peak torque are again the head gaskets because you should not push past 18psi boost. On power I found the limit to be the exhaust cams and the turbo exhaust housing. You can fit a 20G turbo but it cannot flow to it’s potential because exhaust flow is so limited. It makes no sense to cram more air into the cylinders if you can’t scavenge it. All you get is high EGBP, high EGT and knock, meaning you have to run less timing, which negates any power gains you’d have gotten from flowing more air. Best match for the stock motor is a TD05H-18G hybrid in the VF-54 housings. Note the small P18 exhaust housing does not take well to having a TD06L turbine stuffed into it, the geometry is horrible. No larger housings are available, or any other housings at all, for the unique VF-54 flange. Typical on pump gas is about 300whp/320wtq and 335whp/345wtq with meth injection. Jackal8788 did better on E85 with the turbo we provided for him. The stock injectors are enough if running with meth. You’d want 750 or 850cc injectors on straight pump gas or 1000s on E85. Intercooling is getting to be a problem at this level, or more to the point, the lack of cooling air to the intercooler, and a FMIC would be a huge improvement. Meth injection, IC water spray or E85 all help in this regard. 4. Stage 3 hybrid turbo + STi exhaust cams. 2008~ STi exhaust cams and valve springs fit straight into the 5th gen heads. With the hybrid turbos you will get a bump in power and torque, as S2baru found with his GSC Stg1 cams. I haven’t done a setup like this so far so I’d rather not guess at numbers. It is suspected that adding more exhaust flow will make boost creep worse, so sufferers should think about that. Since swapping the cams requires pulling the engine I think it would be foolish to skip pulling the heads and installing studs at this point. Though we haven't tried it, I think boost creep could get to be a problem with the extra exhaust flow and still the stock flapper. Definitely port the wastegate if staying internal. Despite the extra flow from the cams you are going to run into a choke point which is the little 7.5cm2 exhaust housing from the VF-54. With the cams upgraded you'll now get gains by deleting the TGVs. 5. Stage 3 + STi cams with non-stock flange turbo. This is how I know the exhaust cams are the power-killer. I’ve been around several cars with custom headers and GT3076R turbos with EWGs and not one of them could keep up with our 18G cars, they were just a lot laggier and louder. Of course if you’re shooting for the 350whp+ range you should be installing forged pistons and head studs at the least. A FMIC is also mandatory, as is an intake and a full 3" exhaust. You can see how I think this should be done here.

No, a high-flow catted J-pipe is worth every penny. You really don't want to run more than about 16psi with the stock cats in the way.

Yup What happens on an n/a motor doesn't apply here. The turbine and turbine housing are far more restrictive than the header. Not a chance. Will it hurt? Can I have my brain back afterwards? This Easier than making a hood with a WRX-sized scoop, which is a part I wish someone would make. I asked Varis but they flat-out refused. Not because the market isn't there, but because no one is racing this model. Funny how a maker of ricer body kits cares that much about racing.

I've never been able to measure a gain from a CBE on the 5th gen, even when running 18G turbos. Nothing that stood out from the noise anyway. The stock system works surprisingly well and IMO all you need to change is the J-pipe.

Remove the pump and relay. Think about it... the engine pulls a vacuum while engine is running...

A 52 or 16G will get you 300whp on decent gas. I prefer the 16G as it holds boost better at high revs but the numbers will come out about the same. Add an EL header for a bonus 20whp. Forget twinscroll on the EJ25x, it doesn't do that much. If you're running side-feed injectors take the opportunity to upgrade to top-feeds. The stock 520 side-feeds may come up a little short and I hate the modded ones. I recommend the Bosch-made DW750 top-feeds. The stock intercooler is fine at 300whp if you don't push boost too high and pop it. Some people band them to stop that but mine held 18psi for years. New stock shortblock? Pull the pistons and gap the rings properly before putting the heads on. Toss the stock bolts and use head studs for peace of mind.

Yes, but I find that while the stock bolts and gaskets will hold up to 20+ psi (on pump gas) for a while on a newer motor, they will not on an older one. I have lots of cars of all ages running happily on 18psi so that's what I consider a safe limit for long term use. Of course how the car is driven will have an effect on longevity too.

No, I've seen the gaskets go above 18psi on all sorts of turbos including stock. In this case size doesn't seem to matter. Newer motors go longer at 20psi or so but the older ones tend to have problems soon after bumping it up that high. I think you get away with more on corn as the cylinder pressures are probably lower at any given boost pressure vs. gas.

More than 18psi is going to pop your head gaskets at that mileage. Boost does not make more power, airflow does. In your position I'd probably go to either a 16G or VF-52 rather than toss money at a new / newer / rebuilt VF-40. In that case you will need to re-tune, so you might as well upgrade the injectors at the same time. You can run either turbo on the stockers but you may have to taper boost off a bit at high revs. If you're going to rebuild your engine later with forged pistons and head studs then definitely upgrade the injectors so you can get the full potential out the turbo on the built motor. On these turbos you will definitely want to fit that Hexmods VB as plenty of owners have broken the 5EAT with the extra torque. You should not need an AOS at these power levels unless you already have a good bit of blow-by and if that's the case you may not get that extra 30,000 miles before you need to tear it down if you upgrade the turbo.

Nope. They also can be 'unlocked' via SHbootmode tools. This is a method by which the whole ROM image is overwritten, including all those sectors containing the data for communication protocols which Cobb and ECUtek use to lock the ECU. The whole process is in the public domain and the parts needed for the hardware module are dirt cheap. The only thing you cannot do is read off the ROM image. It can only be overwritten.

If you take before and after logs you can see how long it takes to complete a gear shift from the time stamps. It's not a simple matter of just multiplying the R table values by 10 or 20% and calling it good. The CT table wants to take the shape of your actual torque curve and the values need to be raised in steps until you get the shift you're looking for.

Funny, I was just working on an '06 JDM GT that had stumbling / stalling issues. A2UH000A. Yes they do share the same injectors.

Curious, since most JDM models are EJ20 models. Which ROM?

Holy mile-long header Batman!

Ah, those just have the TGVs closed for cold start eh.

Not necessary. Just copy the values from the non-cruise table to the cruise table.

My approach is to run 5th gens with built motors and/or meth injection at a full load AFR between 11.25 and 11.50:1 At that point I insist on having WBO2 data to be 100% sure of the final AFR. Same with any car fitted with an aftermarket intake, WBO2 mandatory. On the stock block, stage 1 or 2 builds I rely on the very extensive and careful MAF scaling I've done on identical cars fitted with WBO2 sensors to run at 11:00 to 11:20:1 with confidence when only having the stock O2 data to refer to. I also use the stock O2 sensor's heater element resistance to warn of high EGT. I note 5th gen owners are a bit more fussy about integrating gauges into the car's interior than others, so they're less accepting of the usual mounting solutions. Remember most tuners are used to working on cars with the O2 in the header vs. the post-turbo location, so they are used to that data being useless under boost. Of course even on the 5th gen the stock O2 accuracy drops off as you approach its rich limit, making the data less reliable. This is where the time invested into MAF scaling comes in. Its a question of what you're used to and feel comfortable with. I'd never turn down extra data if it were available or discourage anyone from acquiring extra data. Those with gauges installed have the chance to spot a failing fuel pump or FPR before anything bad happens.

Depends on the ROM. Most go to 11.02, and there's no good reason to run them richer than that.