Sam's Engine Swap |
![]() |
|
|
Introduction: |
|
This page is no longer being updated. After you've
looked at the engine swap check out my latest
engine swap page.
Disappointed with the engine, and shocked with the prices, I decided it was time to take that piece of garbage out of my engine bay and drop in a REAL engine. Something that would not only last a lot longer than my old engine, but also make my car really, really fast. |
|
|
The EJ20 Engine and JDM VS USDM: |
Since I wasn't about to replace a used blown engine with another,
identical used engine, and there was no way around spending a couple thousand dollars
whichever way I went about this, I decided I needed a stronger power plant.
Subaru has been developing their 2 liter turbo engine for many years now
(the 2.5 liter naturally aspirated engine has far less development behind
it). In the US, the 2L turbocharged design found in the 2002 - 2005 US WRX makes
a very respectable 227 Horsepower. However, in Japan, a very similar
iteration of this block makes 280HP out the door, totally stock. This is
accomplished through a larger turbo, better flowing heads, different cam profile,
different injectors and better internal components all around
allowing more boost, sooner and making better use of the available fuel
(Japanese engines are made to run on what is roughly the equivalent to 95 / 96octane in the US).Swapping in a US WRX engine into a non WRX car will net a car that is perfectly compliant with emissions testing, but the engine's complex wire harness will require a professional 45 hours of installation, at a cost of about $4000 from a tuner shop. Swapping in a Japanese (JDM) engine, however, will not only be much easier (as the wiring harness is smaller, simpler and separated from the rest of the car), but it can also be done cheaper and make 53extra horsepower STOCK. The downside is that there is no such thing as OBD2 in Japan, netting such a swap an instant fail at most computerized emission tests.
Turbocharged 2-Litre 4 cylinder boxer engine. VF29 Turbo running 15PSI, 8-1 compression. Factory Specifications: 280HP@6500rpm, 260ft torque@4,000rpm, redline 8250rpm Interestingly enough, virtually all Subaru engines are interchangeable: Mechanically, this engine will bolt right up to my Chassis needing only the turbo front crossmember from the donor car... The electrical side of it, however, is a different story... |
|
|
||||||||||||||||||||||||||||||||
Wiring Harness. |
||||||||||||||||||||||||||||||||
|
Then all the ECU wires are traced to their end, bundled together with tape and tagged out on their end connectors: And all unnecessary wires are cut off. There are considerably more non-ecu related wires in the car than engine harness wires. This leaves us with JUST the engine harness. This is all the ECU needs to control the engine: This required 7 hours of SOLID work. The process was as follows: Once all the wires were free, I held on to all the wires coming from the ECU plugs and taped them together. I kept taping all ECU wires together as I separated them from wires that did not go into the ECU until all wires had been run into their respective plugs. There, I tagged all the ECU wires out, then cut and removed everything that was not ECU related. The reasoning behind this is simple: If it does not go into the ECU harness, then it is going into the rest of the car. Since all I am replacing is the engine, there is no reason to replace any of the other wires. It was not difficult, but it took a very long time, and a LOT of focus and discipline. If you have A.D.D., this would be the time to take some extra Ritalin (just kidding!):)
I do not believe this swap would be possible without a wiring diagram;
the wires going into the ECU from the dash and fuse box go into plugs that
do not necessarily correspond with the RS plugs, so I can neither use the
JDM harness on the dash (nor would I want to: THAT would be a wiring
nightmare) not figure out from just the plugs which RS plugs correspond to
it. With both the RS and the JDM wiring diagrams I can simply pull the dash
wires from the RS ECU and splice them into the JDM ECU though. Much easier.
|
||||||||||||||||||||||||||||||||
In-Car Wiring. |
||||||||||||||||||||||||||||||||
|
In order to integrate both wire harnesses in the car it is necessary
to gain access to the stock wire harness; this is done by removing the ECU
(take out carpet, unbolt ECU plate, remove ECU, remove plugs).
Unfortunately it is necessary to remove the air conditioning box; this
requires bleeding the AC lines and removing the AC unit from inside the
car. Once this is done the heater core is still somewhat on the way, but I
have decided to do my wiring without removing it; this saves me the time
and
trouble of removing the entire dashboard. In retrospective, it may not
have been that much easier, if at all. But it worked. With the AC unit out the rightmost firewall feed through is pushed back into the passenger compartment and the wires going from that engine harness to the ECU are cut. Anything specific to the 2.5RS engine can be removed, anything coming from the fuse box, the dashboard, the fuel pump, relays, etc has to be spliced into the JDM harness. I made this PINOUT MATRIX to make my job easier; every sensor/connector is named, identified, and put next to its plug location on the JDM ECU. Sounds easy, right?
Below left: JDM harness next to car, wires and plugs being spliced. Below right; JDM harness inside car, most wires and plugs spliced, leftover plugs on passenger seat.
It took an entire extra 2 days to complete the in-car wiring; between wires that simply did not exist in the 2.5RS wiring diagram, wires that were not listed in the JDM diagram, plugs that matched up but had wires in different locations, plugs that were not anywhere in the car, wires that simply did not exist in the JDM ECU at all, the last two days comprised in one of the most frustrating tasks I have *EVER* undertaken. At some point in the swap I actually almost regretted starting it.
With all the wiring done, the ECU wires are bundled together, taped, and the cover is put on. Since the right hand drive ECU has mounts which mirror the US mounts, it is not possible to bolt it down. Instead, I used the cover to hold it in place. Its secure right now, but I will seek a more permanent solution next time I'm under there.
The following 2.5RS ECU wires do not appear to go anywhere in the JDM ecu:
The car starts up, idles smoothly, and runs perfectly, all functions operate normally and it does not show a Check Engine Light signal (thought the CEL IS connected). I guess the 2.5RS wires go into functions that do not exist in the JDM ECU. |
||||||||||||||||||||||||||||||||
Mechanical: |
||||||||||||||||||||||||||||||||
![]() ![]() Here you can see the completed engine installation. Notice the intercooler water spray nozzles on the intercooler plate diffuser underneath the hood scoop, and the dual-water pump JDM windshield washer bottle (pump #2 sprays the intercooler to reduce heat soak).
Virtually any Subaru engine can be put into any Subaru chassis. The engine bolts right up to the transmission, engine mounts, etc etc... There are only 3 sticky points: 1- The power steering lines need to be bent
slightly in order to fit around the larger heads (DOHC vs SOHC). The
power steering pump is not the same, and the JDM lines will not fit due to
the fact that they are for a right hand drive vehicle. Underhood wiring:This was very straight forward; wires are run through the firewall to their proper locations, taped together, covered in wire loom, and the grommet was installed again. A few notes: 1- The rubber grommet had to
be split down the middle to allow the wires to pass. I taped it back
together when done.
|
||||||||||||||||||||||||||||||||
Braking and Handling: |
||||||||||||||||||||||||||||||||
The added horsepower from this swap makes it possible to reach very high
speeds, very quickly. With the stock brakes, any form of "spirited" driving
usually resulted in brake fade as the rotors blued and the pads went up into
smoke. It very quickly became apparent that even my upgraded stock brakes
were not adequate for this power level. The brakes were upgraded by swapping
WRX brakes up front, and Legacy Turbo brakes in the back. The WRX brakes use
slotted Racing Rotors and the back are Brembo cross drilled. Both rotors are
vented and Carbotech Bobcat ceramic street+track pads are used all around.
Front cooling is further improved by air ducts.Update: I am now using Subaru STI 4-pot front calipers. For handling, I am running Whiteline heavy duty swaybar endlinks all around, a cusco 21mm front swaybar, a whiteline 18 - 24mm adjustable rear swaybar (set for 20mm for daily driving) on whiteline heavy duty mounts, and a rear strut bar. Struts are KYB Adjustable Gas Shocks (AGX) and springs are H&R sport springs. Here is a list of the upgrades I have installed. Click on any one of them to see an image of the part in the car. Drivetrain: Handling:
Other: Exhaust: |
||||||||||||||||||||||||||||||||
Initial testing: |
||||||||||||||||||||||||||||||||
|
Here are some ACTUAL tests, performed with snow tires, full interior, a full tank of gas, spare tire, car jack, tools in the trunk, and a passenger:
0 - 100km/H (0 - 62MPH); 4.07 seconds. (this is *FAST*. Faster than a Ferrari 360 Modena, Faster than a Corvette Z06, Faster than a Ferrari 550Maranello...) 1/4 mile: 12.61 seconds @113MPH (launched from 4000RPM. Its decent. I hope to hit 11s with a bigger turbo.)
|
||||||||||||||||||||||||||||||||
Completed Swap: |
||||||||||||||||||||||||||||||||
|
Two bugs currently: No Air Conditioning.
Update: 10000 miles driven and still strong as ever. To date I have only raced one car that was faster in a straight line; a 500HP mustang :) |
||||||||||||||||||||||||||||||||
Credits: |
||||||||||||||||||||||||||||||||
|
1st and foremost I would like to thank Tim Sanderson, who did
the mechanical aspect of the swap (lifting out the old engine and
crossmember, putting the new one in, adapting all the lines and fittings,
etc). Tim is an outstanding mechanic who lives near Milwaukee, WI. If you
need any work done on your car, he's my first recommendation. Thanks goes to Myles Hechtman at Fine Line Imports. FLI supplied the engine for my swap, along with my fuel pump, rear brakes, radiator and a clutch I'm not using yet. Myles was very helpful with my endless questions about the swap and what would or would not fit. If you are planning on an engine swap along the lines of what I did here, be sure to give FLI a call, ask for Myles, and tell them Sam sent you ;) Thanks to Carl Seidel for all the outstanding port and polish work on the engine manifold, exhaust headers and throttle body. I'd also like to thank Robert Shmitt and his brother for giving me a hand with the swap, Hugo for letting me stay over while I worked on the car, and Kelly Drews for giving me a ride and sticking around when it was 100F and there was nothing to do but watch me cut wires for 7 hours. And a HUGE thanks for all the guys at NASIOC, specially turbosubysteve, with his 11 second JDM-swapped monster, for all the technical info without which I couldn't have done this. |
||||||||||||||||||||||||||||||||
Pictures/Videos. |
||||||||||||||||||||||||||||||||
![]() First autocross with the car; stock suspension, upgraded swaybars, too much horsepower. I had a hard time controlling the car smoothly for a fast run; it would get going too fast on the straightaway for me to corner without sliding, and boost during cornering always meant tire spin. On this video you can see a nice wipeout during the slalom and then coming out of the final corner I get on the gas and smoke the tires on the straightaway (frame capture on the left). Nothing like a 4 wheel burnout! :) 1st / 2nd gear pull video (2.3MB). Hear the turbo whine... Autocross with new shocks, springs and boost set for 18.5PSI (23MB) NEW: The Swap Video, produced by my friend Kelly. New: Flyby |
||||||||||||||||||||||||||||||||
Update: The blown engine: |
||||||||||||||||||||||||||||||||
I drove just under 14000 miles in a bit less than 8 months (June 27 2005
to February 18th 2006) on my JDM WRX
Version 5 Type RA engine. The engine was bought used, with an
estimated 50thousand miles or so on it. Following the link above you will
see what modifications were done to it. Essentially, I was running the
original engine, with all original internals, original turbocharger and
intake. I upgraded my intercooler for the larger USDM STI top mount
intercooler and fitted the car with a 3" stainless steel turbo back
exhaust. The free flowing exhaust caused the engine to boost slightly
higher than normal; typically I would see 17PSI whereas the engine should
be doing about 15PSI. After doing a full port and polish on all intake and
exhaust components (minus turbo) the engine was boosting approximately
22PSI, without any kind of aftermarket boost controller. The stock ECU
didn't seem to care; I never experienced fuel cut, and the Air/Fuel ratio,
Exhaust Gas Temperature and Knock Sensor activity were all within spec. I
did have to run higher octane fuel on it to manage those boost levels
without seeing ignition timing retard.One day after driving 200 miles in below 0F (-20C and below) weather the oil light flashed and I realized that the engine had consumed over 4 quarts of oil. There didn't seem to be any smoke out of the exhaust or any leaks under the car. I topped the oil off, drove another 100 miles, and the dip stick had been shot out with a frothy brown mess coming out of the dipstick tube. The suspected cause of failure was a blown head gasket. A compression and leak down test yielded the following results:
Compression:
cyl1 140
cyl2 140 cyl3 133 cyl4 127 Leakdown tests done at 90PSI, percent leakage:
cyl1 3.3%
cyl2 10% some leakage detected in the pcv cyl3 8.9% some leakage detected in the pcv cyl4 17.8% some leakage detected in the exhaust
Factory spec is 3% leakage I believe. With those numbers it was time
to crack open the block and see what it looked like. Here are the
pictures on the drivers side of the block (Cylinders 2 and 4). The
opposite time looked normal.
My Comments:
The low compression test and high leakage on the exhaust side seems to indicate a problem with the engine heads. Most likely the valves weren't seating properly due to accumulated carbon. A sludge of oil and coolant can be seen on the piston tops; ordinarily they should be dry and coated by a thin layer of carbon, as was the case with the other side (pistons #1 and 3). Most importantly, the pitting on the cylinder walls indicates that they have been damaged. I do not know the cause of this. It could be detonation, though I suspect that was not the case, since it was something I monitored and watched out for. Perhaps the head gasket allowed coolant to seep into the combustion chamber and this cold coolant hit the hot cylinder walls and pitted them somehow? I am not thoroughly convinced by this explanation but can't come up with a better one at this time. Regardless, the solution would have been to re hone the cylinder walls out past the pitting, which would require a new set of pistons since the cylinders would now be of larger diameter. These new pistons would require new piston skirts, and the entire engine block would have to be disassembled, cleaned, and re-assembled professionally. I would have to pay for the honing, the pistons, the skirts, new gaskets, new head studs, and then I would still have to find a solution for dealing with the exhaust side leakage on the engine heads.
|
||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||
|
Copyright
� 2002 -2003 by Sam Barros. All rights reserved. Removing any material from this site for display without consent from its author consists in an infringement of international copyright laws and can result in fines up to $50000 per infringement, plus legal costs. So ASK ME before you remove anything from here. |