Deserter GS to Suby EJ20 Turbo

[GS guy]

Looking at the original PCV tube, I couldn't remember where the 2nd inlet was located. I dug out an old pic showing the original set-up:



And you know what? My system only has One PCV system inlet! You can see it on the right hand side with the open T fitting. I never realized before now that my system differed from the diagram in that detail. There are no other inlet paths as originally installed on my JDM Subaru (Legacy) engine.

So the diagram is a little off from my OEM layout. That was an old diagram I found from a Google search, I always though it was pretty universal for the turbo engine layout.

Here's another view:



Sorry about the confusion - and the restrictor is in the T fitting coming off the main cross-over line going into the F connector.

[Vancouver98STi]

The restrictor is positioned exactly as you have it depicted in the sketches, and flow paths as I believe they take during both operating modes. Nice depictions! The restrictor size is a #1 drill, 0.228" exactly (nice snug fit in the OEM T fitting hole).

I can't take full credit for the depictions I posted. I weeded through multiple versions found online, and I then added my guess as to the location of the restrictor. Thanks for the info on the size of the restrictor opening!

Looking at the original PCV tube, I couldn't remember where the 2nd inlet was located... And you know what? My system only has One PCV system inlet! There are no other inlet paths as originally installed on my JDM Subaru (Legacy) engine.

Heh heh... that's pretty funny that you'd forget how the original PCV system was set up. I suppose I can't say for sure how mine is set up either, but my excuse is that I've only had the car for a short time and I've not yet disassembled the PCV system (or even looked at it closely). Thanks for posting the additional pictures. It'll be interesting for me to see how different the EJ20K PCV system is on my Impreza STi compared to how it originally was on the Legacy that your engine came from.

Closely examining a VW 1.8T PCV diagram is suggested.

After pouring over so many Subaru forums looking at this stuff, I don't know if I want to look at anymore PCV system diagrams. I think I'm starting to get a handle on how to "tweak" the Subaru system to suit my needs, and I don't want to get confused again.

[GS guy]

The saga continues.....
Haven't posted on here for quite some time, the bulk of the work (challenges?) on this car has been chassis and general car oriented - not specifically converting to Subaru power. Much progress has been made, much still to do. There's also "life" outside car projects, so forward progress gets delayed. Enough whining - here's a bit of engine related stuff.
Latest tackle is re-working the electronics layout in the driver's side pod. This area is specifically for engine management and controls - so applicable here. Your basic somewhat 'vintage' Megasquirt set-up. MS - ECU, (enclosed) relay board, block box directly above RB is an adjustable pulse-width modulated low speed fan control for the A-A intercooler (high speed bypasses the module for full 12V), fuses and additional relays for boost solenoid power, IC fan delay-on & ignition power. All this crap takes up a lot of real estate! And I don't even have much wiring completed yet. Electrical gone wild.....
Earlier in my thread showed this location with the MS/RB, mostly in a 'finished' and much simpler layout - or so I though at the time. That had to be re-thought and re-worked as I started fleshing out how I wanted things to work, placement of components and general flow of the electrical system. This keeps the main fuse panel under the dash dedicated to typical chassis/body functions (plus fuel pump and cooling fan - located nearby). Amazing how quickly you start running out of space when you're working with physical hardware vs. ideas and plans on paper!
Also making some updates on the fuel system, including a slick fuel pressure regulator from Mario @ The Dub Shop. Lacking availability of some necessary fittings is blocking further progress on that front.
Here's where we're at:

DSC01050.JPG

DSC01054.JPG

Jeff

[GS guy]

More electrical prep - this time on the passenger side. The American Autowire fuse panel came with a large Mega-fuse to locate at the battery for "fuse-link" protection of the electrical system. I decided to carry this further for all the electrical sub-systems, except the Mega fuse packaging was just too big. I decided instead to use MIDI fuses with modular fuse holders to supply the main fuse panel plus auxiliary relays, the fuel pump/coolant fan, plus all the "stuff" in the drivers side pod. These are located as close as reasonably possible to the battery and provides four independent/fused direct battery feeds, plus the main starter connection.
Also on this side up under the cowl I located a small panel with an inertial safety shut-off switch for the fuel pump along with individual fused feeds for fuel pump and coolant fan. These will supply a couple of high amp & waterproof relays forward of the firewall in close proximity to the pump & fan. The location should work for easy access to the safety switch if necessary.
Finally the large buss bar is for ground connection throughout the car. No direct to chassis grounds.
Jeff

DSC01067.JPG

DSC01070.JPG

MIDI Fuse Holder Assy 2.jpg

[GS guy]

Holidays over and weekly deposits of snow looking to be the norm this winter, time to bring the electrical bits inside for some serious wiring fun.
Starting with the drivers side pod electrics which is all engine control (vs. regular chassis electrical). Decided to re-work some of my previous wiring harnesses to provide a more modular layout. Some previous ideas on paper didn't work out that great in reality. The main thing is wanting the engine to be easier to R&R without having to completely remove all the ignition components.
Anyhoo, much nicer to work in the basement workshop for making all the tedious connections and circuits. This shot shows the relay board, intercooler fan controller, with separate fused circuits and relays for the coils, EDIS module, boost solenoid and time-on delay relay to power the IC fan.
To recap, I decided to run the IC fan continuously (except at start-up) at a low speed setting ~1/2 speed via a pulse-width modulation controller. When the MS ECU detects >5lbs of boost or moderate increase in IAT it will command the fan controller to switch to full speed. All settings TBD once this thing is actually on the road. Still have a couple more circuits to complete but this part is about 85% done - although still to be bench tested.
Jeff

DSC01079.JPG

[GS guy]

Completed a bunch of labeling, connector re-labeling, wiring diagram upgrading and bench testing the relays. 2-step fan control working awesome, about a 10 second delay @ fan start-up following ignition on. ALA I can figure out the MS programming to activate the high speed part - should be good to go! Wiring for these parts looking good and mostly finished.
Also re-worked the EDIS harness, updated diagram and plug labels. The plan is for this harness to stay with the motor during R&R, just un-plugging the harness plug from the chassis mounted module.
Next step is tackling the main chassis fuse and relay panel - get as much of the tedious auxiliary relay wiring completed as possible so I can focus on general chassis wire running when the time comes. Good projects to work on with the temps 20's - 30's outside!

[GS guy]

Calling this part done. Glam shot with the EDIS cable added. Worse part is I have to break it down and store until ready for final install on the chassis!
Jeff

Done 1.JPG

[GS guy]

Sheesh - what a mess! Not sure how I'm going to bench test this nest. Good reason for having a surrogate dash mock-up - heck of a lot easier doing this stuff on the bench vs in the car.
Jeff

DSC01084.JPG

[panel]

Beautiful !

[GS guy]

Got all the fuse panel wiring done - that I can do on the bench anyway - and tested using a surrogate taillight assembly. An interesting (read: aggravating!) finding was that the VW headlight high-low beam relay has to have a load on it to function properly. With no lamp attached, it won't switch high-low! Wipers, emergency flashers, turn signals and fog/driving lamp circuits all working correctly.
I decided to take the plunge to LED lighting (at least for the tail lamps so far) using Sylvania replacement bulbs (ZEVO 2357R). Lamp looks good in the taillight, works well with significant lumen output difference between low and high output (tail and brake light). For correct turn flashing I installed an AAW LED flasher, which worked well to flash the installed LED bulb. However, the flasher is totally silent, so other than the lamp flashing there is no "clicking" noise to remind you you left the TS on going down the road. Not sure if this is a non-issue, or if I'll have to add another small relay in the circuit just to have the clicking noise! I think I'll wire up a spare relay to try out and at least have the plan ready If I decide to use it.
Jeff

DSC01086.JPG

[Ol'fogasaurus]

Ooooooo: when wiring is done right and neat is looks so purdy rather than the "rats' nest" you so often see. When it comes to problem solving time the "neat" stuff is so easy to follow and find the problems so much faster.

Good job!

Lee

[GS guy]

Houston: We have a clicker!
After prototyping the idea with a spare micro-relay and couple of N4001 diodes and seeing it work, I decided to (in my best Schwarzenegger impression) "Just Do It"! Diodes prevent backfeeding from active blinker and had a convenient location with existing screw to mount it with simple aluminum angle bracket. Seems like a waste to use a relay just for the coil action and resulting "click", but it works and requires minimal space on the panel.
On to the next challenge.....
Jeff

DSC01091.JPG