utc_pyro

Are the the JDM ones with the motors? If so they aren't exactly plug and play, but can be made work. You need to reuse your OEM connector and change the pins around. But having power folding mirrors is kind of nice. It's not THAT hard to wire them up to work, and you can add inexpensive modules that will auto fold/unfold them based on if the doors are locked.

NGK 93420 plugs cross reference to LFR6BHX's which are the correct plugs for the car. Those are the Ruthenium version of the SILFR6B8 Iridium plugs used on the '10+ LGT, which are upgrade plugs for our cars. They have a slight difference at the tip that speeds up the flame front by some trivial amount, which can result in less knock tenancy. They were only available on the US market about a year ago, so unless the previous owner imported them from japan they probably were changed recently. For coils, OEM Diamond Electric are best but some of us have had OK luck with Cosmos racing coils for half the cost. If they aren't misfiring, there isn't any point in changing them.

Did you use the harness from the OBXT or your LGT? Your coil connectors could also be failing.

It’s possible the higher voltage cooked them, but the ECU coil dwell table is scaled up to 16V IIRC so should have compensated. When are you seeing the suspect misfires? Idle/cruse or high load/RPM?

Cosmo doesn't actually make the coils form what I can tell, there is some chinese OEM that makes them for them with red cases. They aren't actually even as good as OEM, but they are "good enough" and much much cheaper. So I doubt that they'll be able to test them for you. Full disclosure, I'm running Cosmos coils right now as the whole set was the price of the single replacement I needed. It's not terribly hard to drive the coils our cars uses. The trigger lead is low current, so you can build a driver with an Arduino and TO-92 size mosfet. Hook up varyable spark gap (2x screws in a pvc pipe works) and see if you can jump your (gap x dynamic pressure, ~15mm). You can also turn up ignition dwell if you want to push up spark energy. I did it for a while when I was running NA plugs (shop installed wrong ones) with success. Info here: http://dtec.net.au/Ignition%20Coil%20Dwell%20Calibration.htm

It was relatively cool last night, so I went boost tuning using the Cortex EBC that controls the EWG. I have a BNR16g, and am trying to target a flat ~17psi. The EWG is controlled with a 4-port solenoid and the IWG controlled by the ECU but set to 100% WGDC at full throttle. Anyway, after trying to refine the open loop boost control I finally enabled closed loop just in 2nd gear to see if the PID controller could find the correct curve. Did I mention this thing has per gear control? Anyway it spikes up to 17.5psi, then sags below target to 16PSI. By 5500RPM the EWG goes quiet and WGDC up to the 80% range (about where a MAC 35A/45A flatlines), while still staying at 16PSI of boost. Around 6500 RPM boost becomes controllable again and starts trending upwards with WGDC in the mid 60%'s. So by the upper midrange the EWG is more or less closed, and the MAF is only barley more than the IWG was flowing for the same boost level. Thus I think the IWG is cracking open until VE drops and the EWG can come back into play to limit pressure. The IWG has a V2 Mamba actuator with their 12PSI/0.8 bar spring in it, but at 0% WDGC it wants to boost 10PSI ish. I figured that would be enough based on the 2x spring PSI boost limit figure people toss around, but it looks like not. This is the spring chart from my current V2 actuator. There is a big jump from 12psi to 18psi spring options This is the chart from the the V3 Mamba actuators, which they sell an upgrade kit for. I use the EWG for part throttle control to keep things civilized as the EWG is VTA. It is also used by the meth failsafes to limit boost if they trip out. The green connector is pulled to ground and the ECU goes to IWG spring boost and a conservative fueling map. So I'm in a bit of a pickle. If I swap in the 18PSI spring, the IWG will be useless at part throttle as I don't plan on ever getting to 18PSI of boost. It also wont do much of anything in failsafe mode. So do I leave it as is but try and crank up preload, do I buy the upgrade kit to get a 15 psi spring, do I try the 18 psi spring and hope for the best, or do I say f**kit and find a way to fit a 2-port wastegate on the IWG. Kingwana makes one that can be had for about $110 that will likely fit with the rod I had made, but it used a diaphragm instead of the super robust Mamba piston. I think Turbosmart actuators might also be adaptable to 2-port mode, but even used they are $$$.

https://www.block8head.com/ This is my preference for gauges. It’s at the edge of your LOS, greatly increasing safety. I have one of his Legacy 2-gauge pods, and likely will be picking up a 3-gauge soon.

BIU translates the ECU/TCU information and sends it over to the cluster at 125kbps on different CAN ID’s. Contact G-ROM cluster flashing on Facebook, he has all the cluster ID information.

That’s exactly want mine was doing, but at -12.x%. Leading theory is high ground resistance, creating a fairly consistent voltage to offset.. The maf’s Output signal generator (opamp or digital electronics) will be referenced to it’s own ground and thinks everything is fine. But it’s ground is maybe 1/4 Volt higher than the ECU ground (which the ECU uses as reference), creating a bad reading. Compounding this the power draw of heater circuit in the MAF is proportional to the output voltage and thus the ground voltage offset through the ground resistance will rise causing the final scaled difference to the ECU be close to a fixed percentage. Guess this is why some domestic manufacturers use PWM signals from their maf’s vs the analog signals most Japanese and Euro manufacturers use. As long as the signal is still making it and the MAF input voltage is high through for it to operate these sort of issues don’t matter.

Thank you for this post... I’ve been fighting the same issue for the last few weeks, and actually just ordered a new MAF connector and enough shielded wired to replace thing all the way back to the ECU. This seems faster than redoing everything, may try this first.

It can increase load on the turbo center housing bearings accelerating wear. How much? Enough that OEM’s put BOV/BPV’s in tuned to totally prevent it, but not enough that racing teams go out of their way to make sure there is absolutely none. My Evo3 16G starting making them under certain circumstances when I changed to the Grimmspeed intake. Am I worried enough about it to even bother adjusting my adjustable BOV? Nope.

Not exactly “today” but over the last week: Quieted down the Q300 exhaust in the cabin significantly by adding exhaust tips to it. With the wagon rear bumper the Q300 tips are tucked in by about 50mm. This is enough that some of the sound coming from them gets radiated straight into the cabin, and make it pretty annoying. Previously I had stuck motorcycle exhaust baffles in to try and shut it up, but they only helped a little. These are just past the bumper and thus the pulses aren’t transferred as directly into the cabin. As the two exits are separated by 0.4in you can’t stick unmodified 3” tips to on, so I had to cut a 0.8in slit the inner ones so the can partially overlap. Not sure about the look (bit flashy for my taste) but driving the car is much more comfortable. Today: ordered new MAF connector, a spare MAF, and some connectors for relocating the IAT sensor into the throttle body hose when I rewire everything. I have one there now, but it’s just used for logging. Need to change the ECU input so it will adjust timing based on it.

As you're not flowing any meaningful amount more air, it sounds like flow is choked somewhere. You're at 45lb/min which (depending on who you ask) means you're at the max a 20G compressor is rated for. You're also on a TD05H, which will be starting to choke out at those flow numbers. So your flow numbers are right inline with the average 20G on gas. Most 20G builds you see going much further have TD06SL2 turbines in 10cm housings, or EWG's (like you previously had). Remember, flow is more directly proportional to power output than boost. You'll reach a point that the boost increase doesn't raise airflow enough to compensate for the losses producing that boost create. As for trying to push it harder, due to geometry your wastegate may still be creaking open up top due to high EBP. Get a mighty vac and pressurize the wastegate through it's travel to check the alignment. You can see really odd things if it's not acting as direct as possible on the wastegate actuator. You could also try external helper springs, but you're well into the point of diminishing or even negative returns now.

The ECU grounds land on a bracket mounted to the back of the drivers side TGV. It's kind of a pain to get to the ground point directly, so I just hit the TGV. I haven't checked the grounds on the TGV sensor connectors. So what I'm implying by that is the ECU ground was at a higher than chassis ground. The MAF is (probably) fine with this as it's ground is the same ECU ground, but in seeing this it makes me wonder about all the other random stuff I have hooked up piggy backing off stock sensors but chassis ground. Well I learned something new today about what that is! Measurements were taken with the UNI-T UT210E I keep in the car (it has a DC clamp on meter, super handy for car stuff) with normal test probes. So not super super accurate, but it was enough to think "Hum.... This is higher than expected"... I have a Fluke 117 and some clip on probes that I could use for more accuracy, particularly when I go try to find the MAF wiring fault. Come on, we don't mod and tweak out cars because it ran "well" from the factory . I obviously don't think this is going to add power or anything that those crazy JDM aftermarket grounding kits clame, but if I can reduce the noise in the aftermerket sensor data and improve battery recharge charge time might as well.

Some of the double din size JDM navi bezels are on eBay right now for ~$80. Just buy that and build brackets. You can only fit mechless units, but there are plenty of good ones now.

Or not. Now I’m getting random super lean/rich spikes. Great, now the issues is intermittent Ordered 20ft of milspec shield avionics wire for the MAF signal wire. I think I have enough automotive grade TXL wire for the rest. Probably going to wire in a cutout point to make swapping the IAT sensor to the post IC one simple. Now the question is how far into the harness do i need to cut back and replace. I have enough wire to get all the way to the ECU, but as it’s like 35C and 90% humidity outside I don’t really want to dig that far into it. Also noticed the engine/ECU grounds are a bit sketch with a 0.2V drop between the driver side TGV ground point and the negative terminal on the battery. Probably need to redo that as well, but I’m not sure how to go about doing this without causing major ground loops. Not taht the OEM system doesn't have them, the block is grounded to the body in at least four places. Maybe one of the other engineers on here with more recent experience with sensor wiring/shielding/grounding might have some tips?

We need a little more information: 3-port? WG spring? What is your mass flow in this range? Turbo exhaust housing size? To the first point, presuming you’re using the standard Mac 35A based 3-port (Cobb, Grimmspeed, ect), there isn’t an effective difference between 89% and 100%. See here: http://www.circuitse7en.net/page26.php The stock ECU controls the WGS at ~15HZ. Your spring/preload may be too low for your desired boost level still. Or you’re choking out the turbine housing and there isn’t enough drive energy left to spin the turbine fast enough. Did you pick up mass airflow between the two WGDC settings?

Fixed it! No longer showing 310 g/sec airflow in 1st gear with only 15psi of boost... I changed too many things at once to know for sure which change fixed it, but mechanically the intake and MAF looked fine. I cleaned the pre-filter and MAF to be sure, but it looks like the MAF wire going into the Grimmspeed airbox was pinched. So long term it'll require rewireing, but moving the entry point on the wire and cleaning the contacts seemed to do it.

Presuming I don't find something physically wrong, that would be my guess. I had to readjust the turbo inlet when doing the EWG plumbing which necessitated moving the Grimmspeed intake a little. If the wiring is starting to go, that slight movment may have caused an issue. Covertrussian had similar issues when doing inlet testing where he though the had to rescale the MAF, which 100% should not be needing for changes upstream of it. Subaru kind of cheaps out on there wires that have lots of movment (see wagon hatch issues). So if the MAF ground went high resistance, it should cause the MAFv reading to go up (output is referenced to ground, ground is ~0.1v higher than ecu ground, ECU sees a 0.1v higher signal). Not looking forward to rewiring the MAF harness if that is infact what happened though... Finding the correct shielded signal wire will be a pain.

Yes, I know I added MMT? Best theory I have right now is something up with the MAF sensor to cause a consistent(ish) offset. Sensor fowling, filter getting knocked out of orientation, or electrical. Everything else I can think of would cause it to read lean.

Went to a LKQ for the first time today. Both outbacks they had were complete trashed so I didn’t get much, but I did manage to grab a few clips and things I needed. Over the last three days my car has started to seriously pull fuel. As in I’m -12% across the A, B, and C ranges. Watching trims while driving it gets up to the 20’s or as low as zero net pulled though. Fuel pressure looks fine idling (34 psi), and the meth system isn’t showing flow when it shouldn’t, so I’m stumped on what could cause it to go negative like this. An air leak should cause the trims to go positive. A bad O2 sensor would show up as a delta between the OEM and AEM sensor. Only changes in the last few days were installing the new boost controller on the EWG, and adding some MMT based octane booster for additional knock margin while tuning it.

It's not crap, it's what you SHOULD do first if you want to lift the thing. Using other lift kits you'll just loose suspension travel, and cause bottoming out issues on your rear shocks. Use OEM outback or Kind springs, and new outback KYB struts. If that's not enough, THEN start looking at spacing it up.

You might have a hard time there unless it's a NA, has a blown engine, or maybe an '09 OBXT. There arent many of these left out there. As cheap and appealing as the DW740 swap is, skip it and do something EV14 based. They are so much better injectors and easier to tune. If you're on stock 200K injectors, you may need to have them cleaned. Also on plugs, consider the NGK SILFR6B8 from the later model cars. Same price as our normal ones, theoretically slightly better performance from a cleaner/faster flame front. Marketing material from NKG for the aftermarket equivalent below: http://legacygt.com/forums/attachment.php?attachmentid=248718&d=1493920657

Spark plug temperature correlates more with knock than misfires if you need to go cooler. You need the plugs to get hot enough to self clean, but not so hot the plug auto-ignights the charge. Cooler plugs are to prevent knock from happening if the heating gets too high. If anything your plugs look a bit cool from the charts I’ve seen. Also one defiantly has some issues with deposits. This is ether oil contamination (valve guide leak), a sticking injector, or deposits from the misfires.

I’d like to forward you to this bit of intercooler testing that was done by variyforum members: https://legacygt.com/forums/showthread.php/post-intercooler-iat-testing-268293.html TL;DR: Grimmspeed > Stock >> AVO If the Grimmspeed cost too much for you, one of the Grimmspeed/process west knockoffs would still be better than the Perrin/AVO/Misimoto style.