Hindsight's build thread

[Frank818]

Man that car is clean.

[Hindsight]

Thanks Frank

So I have come to find out that on the later model donors (like 2007 and possibly 2006), there is a barometric pressure sensor in one of the secondary EGR valves. The ECU uses this to establish current atmospheric pressure as a calculation input. If it's not there, the car may not run properly. Might be a source of one or more of the issues I've been chasing lately because my donor had an air-pump delete. I did some reading and found out that the "right" way to do the delete is to remove the pressure sensor from the EGR valve housing, mount it somewhere, and keep the connector plugged into it. However, I dieted everything to do with the air pump system from my harness. Not a big deal - I have the wiring diagram and could very easily add the wires back in, however, the air pressure sensor is combined with the electronic EGR valve in one unit and there are five wires going to it.

What I can't seem to find is: which of the five wires are needed for the air pressure sensor to work? I could re-wire the whole setup but it uses a relay and I would prefer not dealing with all that if I don't have to. If I had to take a wild guess, I would guess pins 1-3 are for the pressure sensor and pins 4 and 6 activate the valve. I'll keep searching but if anyone happens to know it off-hand, I'd appreciate it. If not, when I get the valve/sensor in (ordered a used one off eBay), I'll just take a shot and see what happens.

[Mechie3]

Its 07 up. I sent a.link to someone regarding this. I'll find it tomorrow. It makes the ecu trip up as it has no base line atmospheric pressure.

[Hindsight]

Thanks Craig. I am guessing pin 1 is power, 3 is ground, and 2 is signal since pins 1 and 3 are also connected to a bunch of other sensors, and pin 2 goes to the ECU directly so it's probably the signal. All the links I've read about deleting the air pump stuff don't get into the wires because most people just leave the connector plugged in and never monkey with the wiring itself.

Edit: I did see a post on Nasioc from you about it from quite a long while back. Your screen name pops up every so often when I'm searching stuff over there.

[Mechie3]

Yeah, I guess they don't call out the actual wire. Here it the post I was thinking of.

http://forums.nasioc.com/forums/show...&postcount=275

I've got a few writeups to my name on there. Use to be much more active back when I owned and autocrossed a subaru.

[Scargo]

...http://forums.nasioc.com/forums/show...&postcount=275

I've got a few writeups to my name on there. Use to be much more active back when I owned and autocrossed a subaru.

I remember that! Not it that thread it seems, but somewhere in those forums I went through what I did and now my harness is stripped of all that in the STi. You inspired me!

[Hindsight]

Whew, long productive weekend.

The barometric pressure sensor is in the car and working. Shows up good in the logs now. It didn't fix the issue I'm having with the engine occasionally dying when I push the clutch in (it's like the drive by wire throttle plate just isn't doing enough to keep the RPMs up at idle), however, I will say that I have been resetting the ECU a lot as I have been working on various things and the issue seems to get better after the ECU has had a chance to learn and adjust, but we'll see what happens as I put some more miles on it.

I got the bypass valve all installed. Due to the rear firewall I added, it's impossible to remove the turbo inlet tube without removing the intake manifold (which is always a lot of fun), so I pulled the intake, removed the turbo intake pipe and welded on a 1" SS tube at an angle. Repainted it black and re-installed, then connected a 1" silicone heater hose between the new 1" barb I added and the bypass valve.





I had two leaks pop up in my AWIC front heat exchanger. If you bought one from Frozen Boost take note:
1. The overflow barb has like 1/8" worth of threads, maybe 3/16" at most and it doesn't seal well. I had put it in with hondabond some time ago and it still leaked. So today I unscrewed the barb and welded the hole for it shut on the filler neck.
2. The rubber o-ring they use on the drain plug fitting is not compatible with ethylene glycol so it disintegrated on me. I have ordered some buna-n o-rings from Amazon: http://www.amazon.com/gp/product/B00...ilpage_o01_s00


Added wiring for the windshield wiper and connected it to the switch. Still need to modify the windshield wiper transmission to convert it to single wiper but will do that once the body is on.

Oh yeah, I can't remember who but someone posted these cool line clamps from McMaster (http://www.mcmaster.com/#2216t12/=x1r97n) with the suggestion of using them for brake lines. They allow you to clip as many brake lines in side by side as you want. While they were a great idea and looked awesome, they just didn't hold up for me. They all cracked and broke with time and especially if any coolant got on them. Very brittle and heat sensitive stuff. I just replaced them with the padded line clamps instead - not as pretty but more reliable and stable.

Lastly, I tested out my race logic traction control setup by logging the wheel speed sensor signals. Something isn't right because the signal from the rear two wheels is rock solid while the fronts are totally haywire once I get over about 8mph. I've tried testing it with the shield wire grounded and un-grounded and it is the same either way. Really scratching my head on this one because the rear ABS wires are like 8x longer than the front two; the rears go from the rear, all the way to the front, then all the way back to the rear where the traction control computer is. The fronts only have to go from the front to the back. I have a few things I need to try, but after I get this sorted, I am ready to start mounting the body! Very excited about that. It's been a long time coming.

[Scargo]

Good luck! Sounds like growing pains...
I've read about issues with Frozen Boost on here before. Thanks for the heads-up about the clips. I saved that MM-C part for future use. Now, I'll discard.

[Frank818]

I knew about the HE! I was sure there was something fishy on that o-ring, when I tighten it up it torn apart. I told Craig to be careful about this one. Bought a set of o-rings a long time ago and one fit but guess what, I bought my set in China! So even though they are supposed to be coolant-proof, I'll soak it in coolant to test. I will use Permatex sealant on the overflow barb as I don't weld alu.

How long did it take for your o-ring to fail in your face?

[Hindsight]

Hi Frank, the o-ring failed under coolant exposure in about two or three weeks.

[Hindsight]

I've spent the week fussing with my race logic traction control system. The issue I'm having is that both front wheel speed sensors show a really crazy signal into the race logic traction control unit (RLTCU). They are ok from about 10mph to maybe 15 and above that they just go nuts. Rear sensors work just fine. I even made a little spacer sleeve that allowed me to mount a rear sensor in the front and this worked fine for traction control but probably won't work for ABS. I even bought a brand new OEM front wheel speed sensor and it didn't resolve the issue.

I tested about 100 different things and will save the blow-by-blow, but in the end, I have narrowed it down to the fact that the 2007 Subaru WRX front wheel speed sensors simply won't work with the RLTCU, and oddly, the front sensors are quite different from the rear sensors. They are a different size and shape, but also, they produce different signals. The rear produce about 1 volt at 20mph while the front product about 2.5. Also interesting is that the rear sensors are the same part number for WRX from 2002 to 2007, while the FRONT sensors are the same from 2002 to 2006 only - there is a different part number for 2007 (and some 2008). So this leads me to believe that Subaru changed something in 2007 with the sensors and whatever that change is (perhaps they turned them into active instead of passive sensors - active won't work with the RLTCU), it won't work.

Since I want working ABS, I'm going to try a couple routes:
1. Test a used older version of the wheel speed sensors (got one coming from eBay off a 2003 I think). If that works with the Race Logic unit, I'll see if it works with my ABS module. I doubt it will, but I do see some part numbers for the modules showing up as working for both 2006 and 2007. At any rate, if it works with my ABS module, great, done, but if not, I will then buy a used 2006 ABS module and swap it out for mine. I've checked the wiring diagrams and everything is basically the same so it should work.
2. If #1 fails, I'm going to have to see if I can swap the RLTCU out for one that works with aftermarket wheel speed sensors, then add additional tone rings and sensors to the outer CV joints front and rear and fabricate mounts for the aftermarket sensors. That will be a lot of work and RaceLogic may not have any stock to send me.

Since I'm stuck on that for now, on to the body.......

[idf]

If it turns out that what you need is an '06 ABS module (or any other related part) I have one for you.

[Hindsight]

Nice thank you! I should know by late next week or next weekend. Will let you know.

[Loring]

IIRC, the inductive 3 wire sensors send a digital square wave, vs an AC signal for the 2 wire magnetic pickup type. If that's the case, I don't think you'd get a signal, even at 10-15 MPH.

Can you measure the resistance of the sensors?

[Hindsight]

Yeah resistance checks out. Like 1100 ohms I believe. All the wiring has been ruled out, same for the connectors. The most telling test was that a brand new OEM front sensor using the same wiring harness won't work with the RLTCU, but my existing REAR sensors work just fine both in the rear, and when made to fit in the front via spacer bracket (again, using the same wiring harness, connectors, everything). Since the rear sensor works in the front, it can't be and issue with the tone ring, air gap, or wiring. And it can't be an issue with the front sensor since I tried a brand new one.

I might try getting two used rear sensors so I can run with four rear sensors on the car and see if the ABS computer has a problem with it. I doubt it will work if I'm getting vastly different signals between the front and rear sensors now, but it's worth a shot, especially if I can get a pair of rear sensors cheaply.

[Loring]

Yeah, that's weird. 1100 would indicate it's a pickup sensor, and if they generate two different signals (front vs. rear), it would suggest that Subaru set it up like that with the required software. The RLTCU isn't setup to run like that. I agree, run rear sensors in the front and call it a day.

And that also makes sense that the RLTCU would clip out after a certain MPH, when the voltage profile exceeds what it's designed to read.

[Hindsight]

Not much progress. The used front ABS sensor I received was busted (open circuit) so I ordered another and will have to test it next week when it arrives.

The wife got me a nice big 60 gallon air compressor for my 40th birthday (I guess it's more of a consolation prize) so I spent a couple days installing that (running 50 amp wire, adding a regulator and plumbing, and anchoring it to the cement floor).

I brought all the body panels out of storage today and gave them a good cleaning. I'm read to start mounting the body tomorrow. Very excited about that.

Yesterday I was bad and took the car out for a spin on the main road outside of my neighborhood. Tried to do it at a time when there would be no traffic and the street isn't a very busy one. I was finally able to wind it out a little in second gear and part of third. With the Cobb S2 tune in it, the thing is scary fast. I'm sure I'll get used to it and crave more power soon but for right now, especially with how low it is and how big the other cars look, it's definitely scary. I data logged the whole thing and think I may change my mind about tuning the car myself. I saw some significant knocking on two occasions, and my wideband is showing leaner than the command fuel final is calling for, often times under boost which isn't right. The injectors are stock (and were pro tested), the entire fuel system is new and upgraded, and the fuel pressure is right so I can't imagine what would cause this. It also seems like it doesn't flip over into open loop when it should; there are times at WOT in 10-15 pounds of boost where it is still running in closed loop though I don't know for sure if that's an issue but it doesn't seem right. Oddly, the values for AF Correction #1 are generally very reasonable (under 10 and usually under 5) and AF Learning #1 is 0, so I don't think the MAF needs to be recalibrated. The only thing I can think of is maybe it needs a new MAF but I get sick of throwing money at parts that don't fix any issues. That's a long winded way of saying I think I may actually save myself some time and money by paying for a pro-tune.

[Hindsight]

Started fitting the body today. Already stuck.....

I have the side sails clamped in place in the door area. They are 5/8" back from the firewall, about. I put the engine cover on and shifted it around a little bit until the front of it is flush with the door jambs of the side sails. I have experimented with supporting the rear center of the engine cover to get rid of some of it's midline sag.

At any rate, the issue is that I have 3/8" gaps between the side sails and the engine cover on each side! It should be half that or less no? I can get the gaps to about 3/16 at the very front of the engine cover, but then it widens as it goes back until reaching 3/8" at the rear. The trunk cover has the same problem. My issue is that I can't figure out how to correct it. I thought I could just pull up on the top of the wheel well arch and push in at the top of the skirt to slide it inward and then maybe clamp it that way to the shock tower mounting surface, but that won't work because it's too close to the rear bumper. The rear bumper sets the width of the side sails in the back. There really isn't much you can do to adjust the wide of the side sails..... at least, not that I can figure out.

Am I missing something? I see others have closer gaps.

Thanks in advance!

[nkw8181]

There isn't much you can do other then just play with it and get it the best you can. That's what I resigned myself to

[AZPete]

You can change the width of the rear bumper a little by pulling the center forward to make a nice gap with the rear engine cover. I attached a wire from the center of the bumper to the 1 x 1 frame with a turn buckle. After a few months the bumper remembered its shape and the wire wasn't needed so it's now gone. Also, consider the 5/8" placement of the side sails as a suggestion, not gospel, because you may have to adjust if your humps hit the roll bar.

[Hindsight]

Thanks guys.

Pete, if you pull the center of the bumper forward, would that not reduce the curvature of the rear bumper and actually lengthen the width of it? But you gave me an idea.... I could reduce width by increasing the bow of the rear bumper by attaching a wire across a bolt on each side and using a turnbuckle as you did, to tighten. That would bring the sides of the bumper inward. I would have never thought of that thank you!

Also, how does the front-back orientation of the side sails affect the width of them? I didn't think it would change that. I seem to be good at 5/8".... the engine cover comes forward enough to line up at the doors.

[Mechie3]

The rear fender brace I made helps somewhat with this. It pulls the middle of the side sails in to help with the outward bowing. You have access to machine tools right? It's just two tabs on the end with a 1/4-28 rod end attached to that to support a shaft in the middle with LH and RH threads.

[Hindsight]

Ok much better. I pulled harder than I thought I should on the fender to squeeze them further in on their perch and it worked. I won't need to mess with the rear bumper now I don't think.

I do have that ugly gap in the trunk cover, and Pete I think that's the gap you were referring to in your post above right? IE not the side-side gap but the gap between the bumper and the back of the engine cover? I feel like I have seen this was an issue for a lot of people in reading other threads. I think I read Aloha built his trunk cover up to deal with it. I could shift my side sails and engine cover forward but if I did that, I'd have to move it a full 1/2" in order to close the gap between the bumper and trunk and I would worry that I'd then be too far out of alignment for the doors or front fenders or something.

[Scargo]

No silver bullets! Make some shims for gap spacing then use duct tape to bring panels together. I had to trim down some vertical edges to bring things flush and get panels closer.

[Hindsight]

Thanks Craig, I missed your post earlier. I think I was able to get the fenders inward enough by forcing them and bolting them to their perch. Basically same thing as your brace I guess... hopefully they will "learn" their new fitment on the car.

Glyn, thanks. And thanks for your recommendation of the recip air saw. I bought one and it has already been indispensable!

I may wait to do much more to the engine cover until after the doors are on. I really want the doors to line up with the front of the engine cover and even with the door skin just sitting in place, it's very tough to tell. A 1/8" change will make a big difference. I also noticed that the door skins are quite different left to right. I'm going to end up with a big gap on the right side of the car not matter how I adjust the engine cover unfortunately. You can see it in the pic below. Sanding down the returns on the engine cover is going to be tricky as well. The body has a curvature where the side sails come up and meet the engine cover.... it continues that curve.

[Frank818]

Nice to see you can make things somewhat ok by just playing around with figerglass flexibility and not machine too much.

Keep updating with details man, I'll get there soon enough, I wanna to keep issues fresh in my mind.

[Hindsight]

Quick question: My clutch pedal doesn't always return to the full upright position automatically. It has the return spring and the return spring is working and functional, but for some reason, sometimes it gets stuck on the last half inch or so have travel ("last" really meaning "first" half-inch of travel when you are starting to press it down). So after pressing the clutch pedal in then releasing, it stops about 1/2" behind the brake pedal instead of even with it. Clutch master cylinder is new. Clutch slave is old. When the pedal is all the way up and I push on it with my hand, there is a tiny bit of play (1/16" or less) before the clutch master cylinder rod starts moving forward which, I believe, is correct... basically means the clutch master is JUST barely set to not be permanently pressing when the pedal is all the way up.

Thanks!

[AZPete]

Are you sure you have bled the clutch system completely? My clutch pedal would not stay up until I had all the air out.

[Hindsight]

It WAS bled completely yes, but my master cylinder ended up being bad (found that out when I tried to bleed the system and couldn't), and I'm going to assume that since my master was bad, there are good odds that the slave is also bad. So it could be that I bled it, and now after some time, air has leaked in through the slave. I only noticed the clutch pedal thing recently... it hasn't done that in the past few months (I bled it several months ago).

[Bob_n_Cincy]

Hindsight,
The pressure plate (clutch cover) is what pushes the pedal back about 70-80% of the way. There is a spring mechanism that brings the pedal the rest of the way to the top. Disconnect the clutch master cylinder clevis pin and make sure the spring mechanism is working.
Bob

[Hindsight]

Thanks Bob, the spring is there and working. For some reason, it just doesn't always get me that final half inch or so. Maybe it just needs to be greased or something.

[Scargo]

Do you have the spring (shown) in place on the clutch lever? Did you lube the shaft for the clutch lever when you put it in? The only other thing I can think of is the pressure plate itself. It having some friction at the end of its travel.
Subaru-Impreza-Clutch-System-and-Clutch-Disc-Components-Parts-Assembly.png

[Loring]

You should be able to clamp the slave in the rest position and check for sponginess on the pedal. This doesn't take a lot of force, btw. Don't blow your slave or master seals out.

Should feel like there's a brick under the pedal.

[Hindsight]

Thanks guys. Glyn, I don't have the spring in there but I think it's because I have the later model style clutch where the slave is on the engine pushing toward the back of the car. Per the Subaru parts diagram for the 2007, there is no spring there. I did lube the transmission input shaft yes.

Thanks Loring, that's a good idea... that should also determine whether or not the slave cylinder is dragging.

[Frank818]

Hey man, after what, 3 weeks now? I have checked my 3 o-rings that I soaked into coolant and they are in perfect condition. I'll cross fingers heat won't disintegrate them...

2016-04-02 08.14.58.jpg

[Scargo]

Uh-oh! One looks like it is turning brown!Is it Viton?
On a serious note, some materials that are being used just don't live up to the specs of Buna-N. It should be good to 250°F. A Buna-N (NBR) Nitrile is good to 275. A good reference.
Don't feel alone. I have square ring seals and Buna-n looking o-rings that are not holding up on my oil filter adapter. I fixed one issue by ordering Viton seals from McMaster Carr.

[Hindsight]

Nice Frank, I hope they work for you.... and look at that pink VW coolant! I miss that stuff. Works great and seals small leaks wonderfully.

[Frank818]

Brown? It's probably the silver paint that was on the o-ring mixing with the coolant. But I'll take a closer look tomorrow after getting them out.

Yeah pink stuff! I think it's used in Porsches too, that's what my mech uses, or at least a few years ago it was.

I think I'll buy a set of viton (aka Fluoroelastometer) rings good for 400F, just to make sure. My local store has some.

[Hindsight]

Spent a little time on the body today. Seems there isn't much adjustment intended for the fender to side sail, and bumper to fender connection points, which is good because it makes it easier. Wheels are centered with side sail 5/8" back as shown in manual.

Looks like the fender has a mount near the windshield but that isn't referenced in the manual.

Did any of you add some washers between the top of the side sails and the rear perch? I need about 2 or three in there it or it looks like the body is pulled down a bit.

[Mechie3]

That little front fender mount wasn't on the first 30 or so frames. FFR actually has three different variations of the frame at that location.