Project Focus SVT Part III

T minus 0.7 seconds

If editorial stupidity tends to stick in your mind, you probably remember the very specific goal we set for our SVT Focus. Drunk with one too many project car successes, we decided we should try to make our Focus outperform the SVT Cobra Mustang in all our standard performance tests, while keeping the total project bill less than the cost of a Cobra.

Yeah, yeah, we're idiots.
With the stock SVT outbraking and outrunning the Cobra through the slalom, we mostly needed a whole lot more power and a bit more skidpad grip. Last month, we started down the road to power by trying to adapt Performance Turbo and Engine's base ZX3 turbo kit to our SVT. We ran into several snags, all of which we've now solved.

First, we should point out that our kit isn't quite the standard base kit. When we mentioned to PTE that we wanted to beat a Cobra, they slipped a bigger turbo, intercooler and injectors into our kit. Last time we said we were using a Garrett GT28R turbo. Turns out we aren't. The GT28R is good for about 250 hp at the wheels. The GT2871R we actually have is good for close to 400 hp. This is actually the same as the GT28RS Disco Potato turbo in our Silvia (good for about 350 hp) with a bigger, 71mm compressor wheel. The Potato's compressor is 60mm. This means more lag, along with more power potential. Our kit also came with 42 lb/hr (440 cc/min) injectors and an intercooler that's so big, it hangs out under the bumper.

There were three fabrication hurdles and some tuning left to be done after our last installment. Here's where we stand.

The downpipe
Despite the 400-hp turbo, the kit still came with a laughably small 2.25-inch downpipe and with a flat exhaust flange restricting wastegate flow. When the wastegate is open (which is any time you're making big power) exhaust from the wastegate would slam into that flange, turn 90 degrees, run into the main exhaust flow from the turbine, and turn 90 degrees again to squeeze down the pipe. All this slamming and turning causes backpressure, and backpressure after a turbo is really bad stuff.

Besides, the downpipe designed to fit the ZX3 just didn't fit the SVT.We towed the car over to GT Fabrications in Buena Park, Calif., and asked the guys there to fabricate a downpipe that separates the turbine and wastegate flows until they can be merged smoothly. We also wanted the downpipe to connect to the stock SVT catalytic converter, since it's designed for exceptionally high flow and could probably handle quite a bit of power before it became a serious bottleneck.

Unfortunately, the stock cat was too close to the turbine, so there was no way to connect them. On the bright side, GT Fabrications whipped up a stunning, high-flow downpipe, even including a divider wall that sticks into the turbine housing to keep the two exhaust flows from interfering with each other from the beginning. We had them run the 2.5-inch downpipe to the stock 2.5-inch flex joint for factory reliability and just hooked up the stock exhaust for now.

You'll notice this setup doesn't include a cat at all. The plan is to add a high-flow cat just after the flex joint, but for now we just feel like really bad people.

How we're doing so far
TestProj SVT BaselineProj SVT nowSVT Cobra (tested 4/03)The Gap
e.t15.3 sec.14.1 sec.13.4 sec..7 sec
Skidpad.89g.90g.92g.02g
Slalom71.5 mph72.1 mph67.2 mph(tested 4/03)hah!
60-0 Braking120 ft115 ft126 ft(tested 4/03)please!

The oil pump
The PTE kit is very unusual in that the turbo is mounted so low the oil can't drain out of it and return to the pan without going uphill. PTE solved this problem by adding an electric scavenge pump. They mounted the pump very low, however, exposing it to road hazards. Last time we reported, our theory was it was mounted so low because oil needed to flow downhill from the turbo to the pump.

Turns out we were wrong about that, too. The pump can actually pull suction on the intake side, so it will pull the oil uphill. Upon discovering this, GT Fabrications moved it to the driver's frame rail where it's much less likely to get hit by a dead raccoon.

Intercooler plumbing
The ZX3 Focus has a vertical throttle body in the middle of the intake manifold. To accommodate this clumsy arrangement, PTE has a cast-aluminum pipe that incorporates the sharp, 90-degree bend you need to reach the throttle body without hitting the hood. Though this bend is no longer needed to reach the SVT's more conventional side-mounted throttle body, the casting incorporates the Mass Airflow Sensor (MAF), so we have to use it. GT Fabrications made a short adaptor pipe to connect the casting, in its new location, to the existing intercooler plumbing. The result looks a bit crude, but it turns out it's a temporary solution anyway, so we can live with it.

The MAF works by diverting a small portion of the engine's airflow through a small tube. Inside the tube, the air flows in a nice, uniform, laminar way, so it's easy to measure. This sample tube hangs in a cylindrical plastic housing and whoever calibrated this whole thing knows exactly what percentage of the air flowing through the housing is actually going through the sampling tube. If, say, 10 percent of the air gets sampled, you just multiply the mass of air sampled by 10 and you get total airflow. Problem is, we're eventually going to more than double the airflow into this engine, and the sensor isn't calibrated to read that high. We can squeeze all that air through the MAF, that's not a problem, but eventually the sample tube will reach its maximum, 5-volt reading and the ECU won't know what's happening anymore.

The PTE kit addresses this by increasing the size of the MAF. The SVT MAF is 2.775 inches in diameter, while the cast-aluminum tube the MAF now sits in is 3 inches in diameter. This increases the cross-sectional area of the sensor by about 17 percent, which increases the maximum flow the same amount. If it really did read 10 percent before, the sensor would now read about 8.5 percent of the total flow.

Tuning
We'd been faithfully following the turbo kit's instructions when we got to the last step: take your car to a dyno and get it tuned.

Huh? How exactly are we supposed to do that?

Luckily, the guys at FocusSport, who had been holding our hands through the very confusing process of working on a non-Japanese car, had a solution. Superchips Custom Tuning has cracked the SVT ECU. In fact, it has cracked every Ford ECU. This means we can tune the car simply by using its software to reflash the stock ECU. Better yet, we used Superchips' president to reflash the stock ECU. We had had Jerry Wroblewski, president of SCT, tune our car, but you can also have one of its distributors, like FocusSport, do it, or you can buy the software (about $800) and do it yourself. Not a single wire has to be cut. SCT simply plugs into the diagnostic port and lets you datalog and tune to your heart's content.

Wroblewski first loaded a base SVT program that cures many of the SVT Focus quirks. The rough idle after cold start, for example, or the erratic power delivery at the extreme low and high ends of the tach. Below 1800 rpm or so, the airflow meter signal is spastic. The high-overlap cams in the SVT cause pulsation in the intake tract that makes air flow back and forth across the MAF sensor element. The sensor doesn't know the difference between forward and backward, so when the same air flows back and forth across the sensor, it sees more air than there really is. Wroblewski tells the ECU to ignore the MAF under certain conditions, fixing the glitch.

At high rpm, the knock sensor gets overactive and timing gets erratic. Wroblewski turns down the knock sensor sensitivity a bit and smooths out some strange lumps in the timing curve.

He also makes some changes to the variable cam timing maps that he says make significant power improvements on naturally aspirated cars. How well they work on a turbo depends on how much backpressure the turbo causes.

Next, he told the ECU how big our injectors were and loaded fuel and timing maps he's used on turbocharged base Zetec engines. With these basics set, tuning could begin.

The SCT software data logs everything going in or out of the ECU and this, combined with the wideband O2 sensor hooked to the Dynojet he was tuning on, gives him enough information to make the right tuning decisions.

Fuel delivery had to be adjusted slightly because our modified MAF reads differently from the base Focus the fuel map was designed on. Timing also had to be retarded, even off boost, to cope with the SVT's higher compression (10.3:1 vs. 9.6 on the base Zetec). A few dyno pulls into the tuning session, we were putting down 265 hp, but that was dangerously lean and tended to knock at high rpm. Aiming for a conservative air/fuel ratio of 11:1 and pulling timing back until all the scary noises stopped dropped power to 245 at the wheels is still far more than we expected to see on the 8.5 psi the PTE wastegate actuator was giving us.

Making it roadworthy
It's been so long since our Focus ran that it took a little extra work to make it roadworthy. When we tore it down for the turbo install, it had recently been victimized by a curb-crawling executive. Two tires were bald, one had a finger-sized hole in the sidewall, one wheel had a mangled rim and the other three looked like we dragged them behind a truck.

Naturally, we only had four hours to fix the problem.

First, the tire search. SVT is pretty good at filling its wheel wells with all the tire that will fit, but the stock 215/45ZR-17 size is a bit limiting. There are several good tires available in that size, but step up to 225/45ZR-17 and you can choose from nearly every tire made. Plus, we like big tires.

Word on the street, however, is that 225s won't fit on a Focus. We downloaded the specs on a few 225/45ZR-17s and found most of them to be only half an inch wider and about a quarter of an inch taller than stock. Most aftermarket Focus wheels are around 42mm offset, though, while the SVT wheels are 49mm. That makes the aftermarket wheels 7mm closer to rubbing on the fenders, which explains the word on the street. The stock 17x7-inch wheel size is as narrow as you can go on a 225, and using it will make the sidewall more rounded and reduce the size of the contact patch slightly, but it's amazing the details you'll overlook when you already have the wheels.

We decided to use a file to clean up the curbed parts of our wheels and shave off the big, jagged scar on the one with a popped tire, then headed to the hardware store for some red primer (red doesn't look silver or black, which makes it easy to tell when you've primered everything) and flat black spray paint. Oh, yeah, flat black.

Then it was time to choose tires. The recently released Michelin Pilot Sport PS2 has been high on our list of tires to try, and 225/45ZR-17 is the smallest size it comes in. The PS2 is supposed to combine stellar dry grip, reassuring wet performance and good ride quality. Normally we don't care much about ride quality, but with this kind of power on tap, we do care about torque steer. The suppleness that makes a tire ride well also tends to minimize torque steer. But how to find a set of brand-new PS2s in the right size in three hours and 30 minutes? One lucky phone call to Shoreline Motoring, half an hour away in Huntington Beach, solved that dilemma.

Cocky with our success, we decided to stop by Progress on the way to see if the boys could slap in one of its stiffer rear anti-roll bars. Being only .03g away from the Cobra's skidpad grip in stock form, we figured stickier tires and a big rear bar might be enough to catch up. Turns out the stock SVT rear bar, at 21mm, is nearly as big as Progress' 22mm Focus bar. Undeterred, Progress' Ed Florez grabbed a 25mm diameter bar of spring steel and just made a bigger one. Forty minutes later, with the paint on the bar still tacky, we tore off toward Shoreline, shifting with one hand and shaking spray paint cans with the other.As you can tell from the photos, we made it.

Driving it
With just 8 psi of boost, the Focus is amazing. We've always complained that the SVT was geared like a turbo car, now we're glad it is. The gearing is so tall, you can use all 245 hp (from a roll) in first gear without wheelspin. As we had hoped, replacing our bald front tires and mismatched rears (spare tire, remember) with new, supple Michelin PS2s all but eliminated torque steer. There's a big yank to the left when you grab third gear hard, but otherwise, it's surprisingly tame. The car squats noticeably on its stock suspension in second and even third gear, and with the redline bumped up to 7800 rpm, the engine has incredibly long legs.

On its first night on the road (before the new tires), we got into a second-gear roll-on with a stock SRT-4 and simply drove past it. This in a completely stock-looking Focus with a stock exhaust and a space-saver spare on the rear. Oh, for a glimpse into that poor driver's head. The next day we barely lost to a mildly tuned WRX. Never have we driven a car that attracts such frequent street challenges. Of course, it's rare that we drive a car that looks like such an easy conquest.

The GT2871R is designed to work best at more like 20 psi of boost, so at 8.5 psi it's a little off its game. You notice this as a low-rpm lagginess that only accentuates the SVT's already soft bottom end. Roll into the throttle at anything over 3000, though, and it's instant smooth, seamless acceleration. One advantage of our oversized turbo is that it never runs out of breath. At 7800 rpm, it still laughs at 8.5 pounds of boost. With the turbo barely working and the efficient, low-restriction NS111 turbine, you hear a slightly louder, more urgent version of the normal SVT exhaust note--a blatty, edgy wail that makes you think European Mk II Escort, even if you've never heard one. At cruise, the exhaust sounds stock. We're so impressed with the sound and low restriction of the SVT exhaust, in fact, that other than adding a cat, we intend to keep it on the car as is.

Getting it out of the hole at the dragstrip isn't as hard as we expected. Launch with a bit of clutch slip at around 3500 and the turbo comes up and turns the front tires to smog. Power delivery is easy to modulate at this low boost, though, so you can feather the throttle all the way through first and lay serious stripe in second. We dropped front tire pressures only slightly for our testing (30 psi hot), wanting to see what the car could do in the real world rather than some freak dragstrip setup. Turns out the real world, on a mild, 80-degree day, is a 14.1-second quarter mile at 102.6 mph. Sixty miles per hour takes an even 6 seconds.

Our new Michelin Pilot Sport PS2s and the big rear bar didn't quite do what we expected on the skidpad. While we hoped to leapfrog from the SVT's stock .89g all the way to the Cobra's .92g, we managed only .90 with these two tweaks. The massive 25 mm rear bar seemed to have little effect on the handling balance, which suggests to us that the poly-urethane end links are deflecting enough to render thebar ineffective.

The SVT Focus already laid waste to the Corba in the slalom, and the new tires just rubbed salt in the wound, pushing slalom speed to 72.1 mph, almost 5 mph faster than the Cobra. Yikes!

What's next
There are still some fundamental problems with our turbo setup. The MAF readings are highly erratic at idle and low rpm, causing it to buck and stumble at unexpected times. Even under relatively high-load part throttle acceleration, the sensor will occasionally spike, causing an annoying hiccup. Every time the ECU sees a spike from the MAF, it assumes there's really a bunch more air coming, so it throws in fuel and retards timing. The result is a car that is great at hauling ass, but sucks to drive in traffic.

The MAF also tops out at about 6500 rpm at 8.5 psi of boost. Wroblewski managed to supply the appropriate fuel at wide-open throttle between the end of the MAF's range and the 7800-rpm redline, (you can clearly see a dip in power when this happens, though) but at part throttle, the transition is usually accompanied by a huge stumble and fire shooting out the tailpipe. Not cool.The solution is to switch to a larger MAF (possibly from our nemesis, the Cobra) and to move it farther from the pulsating intake manifold. We're going to put it before the turbo, where airflow should be much smoother.

We also desperately need a blow-off valve. It's surprising the PTE kit didn't come with one in the first place, especially when it came with a 400-hp turbo. Because the airflow meter will read all the air the blow-off valve dumps, though, we'll have to plumb it to blow that air back between the MAF and the turbo. No big deal. We can drive the car to GT Fabrications now, and they can do all the work.

We'll also have to upgrade the fuel
pump as fuel pressure is rolling off from the target 40 psi to about 35 psi at high rpm. The Ford ECU is smart enough to adjust pulse width to compensate for the falling boost, but as this problem continues and we try to make more power, we'll run out of injectors. With a bigger fuel pump, pressure will stay up and our current injectors will be big enough. With the extra pump capacity, we can also increase the target fuel pressure with a few simple keystrokes, stretching our injector capacity even further.

We had a problem with oil getting into the intake after a few laps of the skidpad. The result was a storm of detonation and a frightening blue cloud. We suspect the scavenge pump is to blame. Oil is pumped into the oil separator used by the crankcase ventilation system, and it seems likely that this was filling up a bit and the high cornering loads caused it to barf into the intake. An oil catch can on the vent line from the valve cover to the intake might help, but we may go all the way and just reroute the pump's flow to the oil pan. It would mean welding, but we know people who do that.

With these problems fixed, we should be able to start cranking the boost. With the stock 10.2:1 compression and the 91-octane rancid pond water we use for gas, we'll crank it up slowly. With luck we can shave .7 seconds from our e.t. before a hole appears where none was before.

Of course, we'll also need a clutch. The stock clutch handles this power fine, even with a few quarter-mile passes, but try to do burnouts for the camera and it's toast. Needless to say, ours is toast. While we're in there, a limited slip is a must. On our measured tests it probably won't do that much, but in the real world, where we like to accelerate and turn at the same time, the open diff really sucks.

Our cost so far
2003 SVT Focus:$17,480
PTE turbo kit$3,995
GT Fabrication downpipe:$450
GT Fabrication misc. fabrication:$200
FocusSport partial installation:$375
Odyssey PC680 battery:$148
FocusSport torque mount:$99
SCT reflash:$350
Michelin Pilot Sport PS2 tires:$692
Spray paint:$15
Mounting/balancing:$50
Progress rear bar:$199
Total:$24,053
Minus 2003 SVT Cobra:$34,750
Total left to spend:$10,697
SOURCEBOX
Precision Turbo and Engine Odyssey Batteries
FocusSport The Progress Group
GT Fabrications Shoreline Motoring
  • «
  • |
  • 1
  • |
  • 2
  • |
  • 3
  • |
  • 4
  • |
  • View Full Article
Enjoyed this Post? Subscribe to our RSS Feed, or use your favorite social media to recommend us to friends and colleagues!
1 comments
Jay Mitchell
Jay Mitchell

what ever happened to project focus after sport compact car went under ... and after the piston melted


Modified