View Single Post
Old 11-22-2009, 07:31 AM   #1
cdnsigop
Scooby Specialist
 
Member#: 226535
Join Date: Oct 2009
Chapter/Region: E. Canada
Location: Kingston, Ontario, Canada
Vehicle:
1998 Impreza L (RIP)

Default Project Green Collar: 1998 Subaru Impreza L Wagon

I bought an impreza about a month ago, with the express intention of turbocharging it. I wanted a torquey street car, good for all weather. I have been turbocharging Honda's since I was sixteen. I do all my own work, including tuning. The only thing I dont do is welding, but Im learning...

So I started the turbo install last Thursday, and got it started Tuesday. 164,000 km on the chassis and engine. It has the stock 2.2. The car still has the original clutch, it does not slip. I have had this car for the last 1000 km. I did all the maintenance first; new front main seal, timing belt, spark plugs, wires, fuel filter, and synthetic 5w30 oil. Diff and tranny oil was good to go.

I am using the stock ecu, with a greddy emanage blue spliced into the harness; anti stall, ignition, and additional injection maps. For fuelling, I am using red top injectors from a 1998 Legacy 2.5, and a walbro 255 LPH high pressure pump. Since the turbo has been on, (About a week) I have been feeding her 94 octane Sunoco only.

Tuning: Emanage datalogging, plus an AEM wideband controller/gauge. The sensor is mounted in place of the factory front narrow band o2. I have wired the wideband o2 into the ecu harness, at the front o2 input. I have wired the stock front o2 heater circuit positive wire to the rear o2 heater circuit, so I dont get any codes, and dont have a 3rd o2 sensor dangling under the car. Using the 6th position on the rear of the aem uego, It simulates the stock narrowband o2
output, so the ecu can run normally in closed loop. And of course, a boost gauge. Oh, I am using a hydraulic speed control valve from Princess Auto for my boost controller. Cost: $8.50

Exhaust is a 2004 WRX from tailpipe to the downpipe discharge flange. New gaskets throughout, of course. Complete to the chirping at the tips of the tailpipe. I had an o2 bung welded after the cat, to mount the factory rear o2 sensor. It seems that the wrx of that year had the front o2 sensor mounted somewhere in the up pipe.... Where the factory rear o2 was mounted on the WRX, is where I have the aem wideband.

I am using the turbo, uppipe, and partial downpipe of an AVO 2.5 RS kit.
The turbo is a ball bearing garrett unit with an AVO modified compressor housing. Pretty sure its a GT25 and supposed to support 320 hp. Intercooler is a stock 2004 WRX piece, with a crushed BOV from the same car, venting to atmosphere (Didnt want to flip the TB). I made the IC fit my removing the wipers and cowling, and cutting a 1/2" strip of metal where the cowling clips into. A little bit of boot pressure, and the IC was in without cutting the TB or IC inlet/outlet. I am using Forge Motorsports silicone Y-pipe cut down slightly to meet up with the AVO turbo, and the same brand silicone TB - IC coupler. The intake consists of a 3" dry cone filter, maf adaptor, stock maf sensor, and a chopped down legacy 2.5 intake pipe. I made a catch can out of pcv pipe and fittings. All the breathers are hooked up to the can, then the can to the intake pipe in front of the turbo. Works like a champ. The downpipe was cut, then had a
partial 90 degree welded to it, with the 2004 WRX downpipe flange. Now I can use any aftermarket WRX catback..

I installed the emanage, gauges, fuel pump and injectors, and learned the support program before I ever installed the turbo.
Almost forgot: The boost controller dosnt affect the final boost much, so I see about 5-6 psi after a couple seconds in the boost, at WOT. What it does affect is initial boost, allowing an 11-12 PSI spike, tapering to the 5 6 psi I mentioned before. Because first gear is finished so quick, 11 psi is maintained throughout. 2nd, spike to 11-12, then falls to 6 at around 4800-redline. 3rd, 11 psi to 4000, then falls to 6 - redline... Same story in 4th.. The spike only lasts as long as it takes for the pressure to build at the wastegate. The valve Im using is simply an adjustable restrictor with a check valve built in. So, it restricts the flow of air to the wastegate, but allows the pressure built at the wastegate to be relieved back through the valve. Max voltage at the MAF sensor is 4.6v so far. It peaks at that, then tapers down to about 4.2v at redline. A/F's are around 11.5-12:1 at peak boost, then richens up to 10.5:1 towards redline. I am pulling timing starting at -1 degree at 4v MAF voltage, and tapering down to -8 degrees at 4.8v...

I baby the tranny on shifts, and dont floor the throttle till my foot is right off the clutch.. So far, no clutch slip at all, car just flat hauls ass. Huge torque spike when boost spikes, so the midrange power is stompin'.

Overall, I am really happy with this setup so far. That AVO turbo spools up so fast, and the 2.2 was torquey to begin with.. Now, especially with the boost spike, there is no need to downshift, reminds me of my SV1000. Just open and go. And the traction! Fist gear right handers are fun. I put on a perrin 22mm rear sway, and on the softest setting, it has almost totally removed the car's tendancy to understeer. With the new found power, it is possible to maintain a midrange power induced light drift on dry pavement in 2nd gear. Never was able to do that in the honda's!

Only two problems:
a) a glutenous thirst for premium fuel, especially the way I drive.
b) a tendancy to stall at idle a few seconds into a cold start, and the odd time after a WOT run, when idleing at a stoplight. I have adjusted the antistall maps extensively, even increased the idle speed. No go. In my obsessive quest for info on this forum and others, as well as google, I have seen one solution calling for a longer pipe between maf and turbo. I cant get my head around how this could work though. The MAF voltage is clamped at idle! Maybe someone could explain it..











* Registered users of the site do not see these ads.

Last edited by cdnsigop; 01-28-2011 at 10:34 AM.
cdnsigop is offline   Reply With Quote