Tom Wilson
March 12, 2007
Westech electronic fuel-injection specialist Tom Habrzyk ran our Comp XFI cams examination for us. That means everything from the bolt-unbolt drudgery to tapping the engine-management fine points into the laptop.

Regular readers know we often run several Xtreme Energy hydraulic-roller camshafts from Comp Cams on our 347 dyno mule. Besides running well, those cams are good choices for a test engine because many of you are running them in hot street and hobby cars. At the most recent SEMA show, Comp announced a new line of XFI hydraulic-roller cams for popular V-8s, including two for the small-block Fords. After a quick call to Comp, a pair of the cams were headed our way, along with a bit of background on what these cams are all about.

It turns out that XFI is Comp's designation for its latest cam science. According to Comp Cams engineer Billy Godbolder, the basic idea is to reduce some of the overlap while otherwise bumping up the overall aggression. That means the XFI cams exceed the Extreme Energy line-which are bold units in their own right-and head right out to the physical limit allowed by hydraulic lifters. The result is rather nasty, big-lift, big-duration grinds aimed at seriously modified, larger-displacement small-blocks. These XFIs are high-rpm lobe profiles designed to run at the strip or track, with only minor concern for streetability. They're a real test of thevalvesprings and, in short, overlap flat-tappet cam territory.

When you can see a big, long dome in a cam profile, you know things are getting serious. This is the 236 XFI cam, and its mountainous lobes give a strong visual clue that the hydraulic lifter has precious little time to do anything hydraulic. Careful attention to valvespring pressure, installation, and maintenance are required, as is a constant supply of oil.

Remembering that the XFI cams are designed to work with electronic fuel injection, let's take a look at the two new camshafts' specifications, along with the next largest Comp cam optimized for electronic fuel injection, the 282 Xtreme Energy grind (see chart below).

It's clear the XFI cams are the next step up in duration, but rather than Comp's usual 8-degree jump, the 248 XFI gains 10 degrees of exhaust duration compared with the 236 XFI. Also, the gain in lift is considerably more aggressive, which differentiates the XFI line from the XE line.

Were engineer Billy Godbolder here, he'd no doubt point to how the XFI cams employ 114 degrees of lobe separation angle instead of the XE's 112 degrees. The wider separation angle is required by the long duration, which needs "narrowing" by widening the lobe separation angle, especially if it is to work with fuel injection. This helps keep the cam from being too radical at low engine rpm and, believe us, these XFI cams are already plenty snotty.

Clearly, with six tenths of an inch of valve lift, the XFI cam family gets with the power program. Our guess is they could manage to lurch around on the street given a light car, a clutch, and some help from the rear-axle gearing, but these are obviously not daily driver grinds. With a little tuning, we saw just 8 inches of vacuum at a 1,000 rpm "idle," so there are precious few good manners here.

  Duration @ 0.050 in Valve Lift Lobe Rocker
  Int. Exh. Int. Exh. Sep. Arm
282 XE 232 {{{240}}} 0.565 0.574 112 1.6
236 XFI 236 248 0.579 0.579 114 1.6
248 XFI 248 258 0.608 0.608 114 1.6
Comp's XFI cam line for small-block Fords is two strong-the 236 and 248 XFI cams. Roughly speaking, they pick up where the Xtreme Energy line leaves off, and for all practical reasons they reach the limits of hydraulic-roller cams.

With the two XFI cams in hand, we knew we could run them against each other, as well as compare them to one of our earlier 282 Xtreme Energy tests, to show where the XFI cams fit in the overall scheme of lifts and lobes. But as these cams promised to make more power than ever from our 347, it was also a chance to see what a few supporting pieces could do should we have enough testing time.

With that in mind, we chose to start the test with the 347 dressed exactly as it was during out last round of intake tests. That meant a Coast High Performance Street Fighter 347 short-block, AFR 185 cylinder heads with heavy valvesprings, and our usual electric water pump, long-tube headers, and FAST fuel injection. Because it squeaked out the highest peak power number in the previous intake test, we opted to fit Edelbrock's Performer RPM II intake manifold, along with a 70mm Accufab throttle body and spacer. This configuration returned exactly 450 hp during the intake test while using the 282 Xtreme Energy cam.

As is our normal practice, we used an electric water pump and no other front engine dress on the dyno. This greatly reduces the wrenching workload, especially during cam changes. But it also means the power normally absorbed by the water pump, alternator, and other accessories is not accounted for; you'll have to do that when figuring how much this combination would work in a car.

We should note that our engine was wearing Comp valvesprings specifically selected to handle big lifts and durations. Part of a complete Comp Cams valvetrain-lifters, pushrods, and stainless steel 1.6 rocker arms-the valvesprings were selected by Steve Brule at Westech during an earlier test to control valve float with the XE cams. This was lucky, as the big XFI cams were only going to require more spring pressure. This can be even more important when you learn just how finicky a huge, hydraulic-roller cam can be when it comes to oil pressure and valvespring pressure. These are major variables in valve-event control with big-juice cams, which are always at the edge of valve control.

Fortunately, as we had already been running these springs, we did not have to change them for this test and thus did not introduce a variable into the mix. For the record, the springs are Comp 977s; they yield a stout 160 pounds of pressure on the seat and 430 pounds open, and they proved just enough for the XFI cams.

As we got the ol' 347 on the Superflow, Westech electronics guru Tom Habrzyk draped a FAST fuel-injection system over it and the fun began immediately as nothing worked right. This was the latest version of FAST, full of so many improvements Tom eventually pulled it off and refit the original FAST gear, at which point everything came right and we could start testing.

Westech works closely with Comp, which owns FAST, and part of that partnership is getting beta versions of way-cool electronics to see just how good you are. Tom's great and so is FAST, but this time it was a real reminder of how much effort goes into getting all this science to run correctly.

Our normal exhaust system was pressed into the fray once again. The Hooker headers use 151/48-inch primaries, the collectors are 3x18 inches, and no mufflers were used. We didn't test mufflers this time around, but at this power level even the massive 3-inch Flowmasters we normally use are beginning to kill some power.

Next Tom dressed up the fuel and spark maps with a series of abbreviated power pulls. Thanks to a wideband oxygen sensor on the dyno, dialing in the air/fuel mixture to the mid-12s was simple enough. We snuck up on the ignition timing in a series of pulls, although it always ends up in the same place with modern fast-burn combustion chambers: 32 degrees.

As Tom tuned, we couldn't help but notice just how nasty the 236 XFI cam idled. Much of this was tuned out later, when we specifically tried for idle quality, which is often ignored when full-throttle testing power-producing parts. It was a powerful demonstration of how good electronic fuel injection is at providing the optimal fuel and spark at every rpm. In the old carburetor and tunnel-ram days, a cam this big would have been a pain even at the track. Now we can tune them to near civility. We shouldn't short Comp Cams either-its new cams are marvels at valvetrain longevity and idle quality considering the full-race quality of their profiles.

With the fuel and spark set, we made several power pulls to check for repeatability. The power peak was 448 hp at 6,000 rpm. In our haste to run the test after spending the morning fiddling with the fuel injection, we didn't even notice this was less than what the 282 XE had produced in the same engine.

As the clock wound its hands at cartoon fast-forward pace, we dove into changing to the larger 248 XFI cam. With that chore completed it didn't take much tuning to arrive at a 450hp peak at 6,100 rpm. By then it was starting to sink in that something else was restricting the engine, but with closing time already long gone, we opted to turn our one-day test into a two-day affair so we could chase down the airflow.

But before turning off the lights, we fiddled with the tuning to arrive at a representative idle. Eventually, it stuttered along at 1,000 rpm, rattling in a definite attention-getting way. This brought up the vacuum to 8 or possibly 9 inches. You'd get really tired of this idle after the first week in a street car, but it would be music in a special-occasion fun car.

Our dyno tests have repeatedly shown Edelbrock's Performer RPM II is a versatile intake on everything from the mildest bolt-on engine to cammed-up hot rods. But its long, torque-friendly runners are mismatched when reaching for 500 naturally aspirated horsepower at 7,000 rpm.

Returning fresh the next morning, our first thought was to try a different intake manifold. The RPM II has proven a wonderful intake in the 6,000-rpm range, but the XFI cams promised to shine at least an extra 1,000-rpm higher, and so we really wanted a short-runner intake.

It's worth noting there are plenty of large runner-that is to say, large volume-5.0 intakes currently available, but short-runner intakes are a comparatively rare breed. Fortunately, we had one of the best on hand, a Trick Flow Box R, which features abbreviated runners coupled with a huge plenum. It's just the ticket for higher-rpm naturally aspirated engines, or big-boost applications.

So for once we believed we had enough engine to support a short-runner EFI intake, a class of manifolds that usually struggles in the typical bolt-on engine dyno test. Using the Box R manifold, peak power climbed 12 hp to 462 hp at 6,500 rpm, which was also the run's peak rpm.

When comparing the Edelbrock (front) and Trick Flow lower intakes like this, the differences don't really jump out. Both are approximately the same height and the runners don't seem much different in the cross-sectional area.

Knowing we were on the right track, we took one look at the 70mm throttle body and decided it was the last choke in the system. The Box R intake is machined for a 90mm throttle body, so that's what we moved to. We also turned up the rpm limit, as the cam was clearly wanting to run 7,000 rpm, the cylinder heads seemed good for it, and we finally had an intake and throttle body ready for such action. At the other end, the long-tube headers and open collectors were ready to party as well, so we were anticipating a good run.

And we were right. The power peak jumped to 6,900 rpm, where we saw 481 hp. Obviously the Performer RPM II, and especially the 70mm throttle body, were holding back these XFI cams-no mystery there as these parts are designed and marketed for lower rpm use-and the power gains once the rest of the breathing tract was put right were substantial. The 248 XFI cam was finally able to show what it could do, which was 31 hp more than the 282 XE's jumping-off point.

What about the 236 XFI? It never had a chance to strut its stuff with the short-runner intake and 90mm throttle body. We'll go out on a short, stout limb and estimate the 236 XFI would make 259-263 hp with the higher rpm intake and throttle body.

Lay the Edelbrock (lower) and Trick Flow gaskets next to each other and the runner-size differences are much more apparent. The staggered Trick Flow runners are obviously much larger in area. Even more importantly, they're also much shorter.

We'll finish by noting we found the XFI cams sensitive to oil pressure and volume, as they should be. These monster hydraulic-rollers don't give the lifters much time to cycle, and if you allow the oil level to drop or the pressure to fall for whatever reason, expect to begin losing control over the valve events. Our test mule didn't have a dipstick in it for this test-you do unusual things with dyno engines-and we did get the oil low. The engine talked to us until we added oil. Even then we suspect the valvesprings were just passing muster, but we were far over our time limit and had to call an end to the test. Who knows, there might be another few horsepower lurking in these cams yet.

So, the XFI cams work. If you want to stick with a big hydraulic unit, they'll do the job. If you were willing to move over to a mechanical roller that might prove a little more forgiving, those are considerations that depend on the vehicle and its use. But with 481 hp on our dyno sheet, we think we could have some fun with these bumpsticks.

Horse Sense: With as much assembly/disassembly as our 347 mule motor gets, you can't blame it for letting loose with the front oil seal. But what a mess! With a few runs left to go and time running short, it was easier to simply finish the cam test than stop and change the seal. But it took a 5-pound box of Tide to clean up the dyno cell.

Once again, casually laying the Edelbrock and Trick Flow upper intakes next to each other does not directly show the immense difference in runner length between the two. At least it's clear the Performer RPM II is a long-runner design with a triangular plenum.

Open the two-piece lid of the Trick Flow upper intake and this manifold's short-runner, high-rpm personality is finally obvious. The only runners are these small stubs protruding from the casting's floor, as well as those in the lower casting. Important details are the somewhat rounded edges of the runner entries and the CNC-machining to finalize the entries' shape.

Mounted on the engine, the Trick Flow intake does what it can to give our visually challenged dyno mule a handsome profile. The Trick Flow upper is a simple two-piece clamshell; it quickly bolts together and features reusable O-ring sealing. It's a class act for high-achieving engines.

While the 90mm Accufab looks like a hero in this test because it went on last, it would have appeared lackluster had we fitted the 90mm throttle body to the Performer RPM II at first. Only when all the pieces in the airflow train work together will the power burst onto the scene. As always, it's the combination of parts that does the trick-and this time, the 90 was the final piece of the puzzle.

For years we've used this 70mm Accufab throttle body to test all sorts of parts on 350hp to 450hp engines, where it works great. But it wouldn't let our 347 get much past 6,500 rpm with the big XFI cams. Up to 4,100 rpm it made more torque, but with the power peak another 2,800 rpm higher, it was definitely time to go with a larger throttle body.

Truly large hydraulic-roller cams require that everything is optimal in the oiling system because they work the lifters so hard. Our engine lost its front seal during this test and got low on oil-it was as if the engine was hitting some sort of soft rev limiter. The misfiring was audible, the power flattened, and the engine wouldn't rev. Adding another quart of Lucas synthetic got it working again.

 236 Int.Exh.248 Int.Exh.
Rocker Arm1.
Lobe Separation114114114114
Timing at 0.006-inch tappet lift:
Intake opens32 BTDC38 BTDC
Intake closes74 ABDC86 ABDC
Exhaust opens86 BBDC91 BBDC
Exhaust closes34 ATDC43 ATDC

The Test Engine
Approaching 500 hp from a naturally aspirated 347 is hardly heroic, but it still takes good parts. Here's what was in our test mule.

Short-block{{{Probe}}} Street Fighter short-block 4.00x3.40-in bore x
  stroke, 342 ci 50-oz balancer
Cylinder HeadsAFR 185
ValvespringsComp Cams 977, 160/430 lbs seat/open
Rocker ArmsComp Cams stainless steel roller rockers, 1.6:1
PushrodsComp Cams
Timing ChainFRPP
Intake ManifoldsEdelbrock Performer RPM II TFS Box R
Throttle BodiesAccufab 70mm and 90mm
Air FilterNone
Fuel RailStock
ElectronicsFAST speed-density engine management 36-lb/hr injectors,
 FAST/GM, high impedance, 12 ohms or more injectors
Fuel RegulationWestech fuel-regulator system, dyno-mounted; fuel
 pressure 38 lbs
Fuel91-octane Rockett Brand gasoline
ExhaustHooker Super Competition headers, 15/8-in, slip joint on
 back two primaries, 3-in collector, 3x18-in tailpipes,
 no mufflers
A/F DataHoriba wide-band oxygen sensor
Water PumpCSI electric

RPM282 XE Camshaft*236 XFI Camshaft*248 XFI Camshaft*

* Edelbrock Performer RPM II

RPM248 XFI Camshaft**248 XFI Camshaft***

** Trick Flow Box R, 70mm Throttle Body
*** Trick Flow Box R, 90mm Throttle Body


Memphis, TN 38118
Trick Flow
Tallmadge, OH 44278