joshuaho96

Cool to see that they at least found that EGTs had less spread with the Nismo collector. They didn’t really find low end or midrange gains on the dyno but weirdly enough he makes mention of having to add more fuel to the tune at low RPMs and part throttle. Seems like it does reduce the off-boost lethargy a bit even if the dyno pulls they did didn’t show it very well. Also interesting that deep into the revs they found that they made more power with less boost. Not much but it’s there.

That... doesn't really make any sense? Why wouldn't it have a rev limiter?

I would've just sold the car and called it a day at that point.

Yeah, I'm just glad my stupidity didn't require me to tear apart my entire intake manifold again. I checked every cylinder after that up to the intake valves to see if any debris had made it into the engine but I didn't see any signs of issues. Took it for a test drive and the car seems to run properly now. Idles properly too, before I could tell the ECU couldn't hit the idle target so it just had the AAC valve perpetually closed in an attempt to hold it around 1050. Now it idles at 950 rpm with about 33% AAC duty cycle.

I'm just so glad it came out as a single big wad instead of getting shredded to bits. Genuinely was terrified that I was going to have to pull everything apart again to fix this.

The dirty secret of almost all of these aftermarket injector companies is that they do not do much, if any R&D in most cases, they're just relabeling Bosch, Denso, or Siemens injectors. They might decap the injectors for higher flow and maybe also match them again so that they actually function properly. Check your injector against these, I bet they are a direct match: https://www.bmotorsports.com/shop/product_info.php/products_id/1759 The one exception maybe that I've seen is Nostrum Energy out in Michigan. If you read Glassdoor reviews they have ex-employees that complain pretty mightily about how they want to be a big time OEM supplier like Bosch or Denso but they waste their time on direct to consumer aftermarket parts. AMS Performance uses their HPFPs for the VR30DDTT as they've designed a custom HPFP that flows a lot more than stock and is also E85 compatible.

Yet another bump because I can't edit my posts. Turns out I'm a huge idiot that left a paper towel in there:

How is the idle? Are all injectors flowing properly? I recommend running a cylinder balance test where you disable injection on each cylinder at low idle and see if any one of them seems to have a weak influence on idle RPM, they should all drop the same amount of RPM when you turn it off.

It was pretty curious, I noticed that the ECU basically doesn't do closed loop idle control if it doesn't see the idle switch. So it can idle low or high depending on coolant temp. I suspect it's also to help transient fueling control as well, it can be pretty hard to tune throttle tip-in just off idle.

Major thread bump, for the RB26 (and likely the RB25 I'm guessing) the idle switch is a mysterious thing that somehow resets if you unplug the 5V supply (not the connector with the EV1 style metal clip) while the ECU is on. No need to do weird key tricks, no need to run the engine at all. Just turn the key on to get the ECU running, adjust the TPS to get the requisite 0.45V at idle, tighten it down, then unplug the TPS and plug it back in. It seems an awful lot like the ECU is using TPS voltage for idle detection but if you look at the R33 FSM it details how there's a physical switch in the TPS and that directly goes to the ECU for idle detection. I have no idea how that works.

I fixed it. Turns out with the first heat cycle (or something?) the throttle coat decided to jam the throttle partially open. Solution is to repeatedly let the throttle body slam shut on spring pressure until it smooths out.

Any advice for how to fix an ITB that has had its full close stopper slip? My cylinder 5/6 ITB did this and there's basically no way for me to adjust the ITB linkages properly. Really dreading pulling the engine all apart again to fix this.

You're supposed to have that feed before the throttle body, this looks like a single TB conversion so look for where the TB is and put that hose before it. As others have said DBW is the way to go but it's up to you.

Definitely get a rubber adapter designed for your jack stands if you're going to put the jack stands on the pinch welds. You will regret it if you don't. I haven't had issues with my alibaba special adapters but it's really important to note that I am extremely careful to inspect them after each use to make sure it hasn't been cut through by the pinch weld and I check many times to make sure the jack stand is lined up properly before I slowly lower the car onto the stands. I also use a rubber floor jack puck to avoid marring/crushing the floor jack points. That front crossmember looks really weird to me but checking around online that is in fact the floor jack point. Note that in the factory service manual the floor jack has a lifting block on it, presumably because just using a bare floor jack on it will damage the crossmember.

Do you have torque spec on those things? Generally speaking I recommend only using a short quarter inch ratchet to tighten things like this because the torque spec can be as low as 5 ft-lbs. I don't know how you'd fix the issue pictured, you'd have to ask a machine shop what they could do.

I wouldn't run pods, MAF or not.

I see the adapters you're talking about all the time, it's just a rubber block with a slot cut in it:

https://sl-i.net/FORUM/showthread.php?18087-Subaru-Factory-Service-Manuals-(FSM)-Every-Model-USDM-EU This is for the US models but it can't be that different, as a general rule unless you really know what you're doing you shouldn't deviate from what the FSM states. Super important to note that the 2 post lift and emergency tire jack do not just sit on the pinch weld alone, they also contact the flat portion beside the pinch weld which is the primary load bearing surface so do not just blindly set your car down on jack stands without an appropriate adapter:

Something to keep in mind with doing a blow-through setup is that it will likely increase MAF maintenance requirements, engine blowby with the right particle size can get through the dirt/fluid bypass that the card style hot film MAFs often have and deposit on the sensor, affecting the calibration. I have no personal experience to say whether this is a problem with the modern R35 Hitachi sensor but like anything in engineering you're picking what compromises you're willing to deal with. Putting the sensor closer to the intake plenum improves transient accuracy and more accurate IAT but it all comes at a price.

Good to know, when I was thinking about this issue for a while I had a feeling that was the case but I checked all the connections from the PTU to the coils anyways. My circuits professors are probably shaking their heads right now.

R34s shouldn't, but the logic is the same. The ECU sees the primary side wiring from the ECU output all the way out to the coilpack pins and back to the ground point. If it sees open circuit it sets a code 21. Short circuit you would probably detect pretty rapidly regardless of whether the code is reported or not.

If you want to conduct this test without the work involved in pulling coilpacks then you can use Consult to disable individual cylinders, I use the active test section of NDS1 for this: In theory you could use this to figure out exactly which coil is misfiring under load but at least in the case of an RB26 the coilpacks lack ignition feedback signals so the ECU cannot detect any issues on the secondary windings, so you're going to be spraying quite a lot of unburned fuel into the exhaust. A misfire is really not that hard to figure out though, your engine is going to run quite rough from the imbalance in combustion forces. I say this a lot it seems but NDS is a super useful tool for diagnosing problems with these cars. In the case of your fluctuating idle just reading the ECU logs would give you some clue as to what's going on, such as weirdness in the O2 sensor signal or MAFs.

Yet another bump. I recommend buying new hoses for your brake and clutch master booster because my brake booster hose is rock hard and split at one end. I'm going to live with it for now but the replacement has been ordered. The plenum intrudes and takes every little bit of space that the factory clutch booster setup needs so the Nismo manual for this part details how you need to remove the clamp that goes around the check valve as well as the bracket that holds the hose and check valve straight. You will still need the nut and lock washer that goes to the clutch master cylinder, just the bracket is removed. The rubber insulator on the check valve clamp is also retained on the check valve as far as I can tell, presumably to try and keep it from grinding into the plenum. This thing has been a ton of work and money, so hopefully the long runners at least do something for low RPM response. It's definitely not going to be a huge difference but just looking at a similar kind of experiment I'm optimistic:

I found the same link through google and it doesn't seem ridiculously sketchy at least.

Code 21 means the ECU isn’t seeing the right resistance on one of the coilpack primary side circuits. I got it once when my ignitor unit had a slightly loose connection.