Tom Wilson
February 21, 2004
Horse Sense: As is the science in medical journals, dyno reports are notlong on passion. But who wouldn't love standing in the dyno celllistening to 1.3 hp per cubic inch thump out its tune? It almost makesup for the 100-degree temperatures and pizza-only diet.

It's not taking us long to figure out a 347 is the way to go for easy horsepower with a 5.0 engine. Now that we've had our 347 on the dyno for a couple of tests, we're impressed with how easily it churns out the horsepressure. Compared to a stock-bore-and-stroke 5.0, it might as well be cheating! And if you're class racing, it probably is--but that's a different story.

Besides easily made horsepower, a 347's longer stroke means it makes power a little differently. At its most basic, a longer stroke is simply more displacement, and displacement means torque. Furthermore, the longer stroke means the piston is accelerated faster up and down the cylinder for a given rpm, so piston speed is higher in a 347 than in a 302. That produces a stronger vacuum signal at the intake valve, so the incoming column of air is accelerated more forcefully. This has camming and intake manifold consequences, so we thought we'd try a few cams in our 347 mule and find out. Along the way, we were able to also take a quick look at a second intake, flashback to carburetion, and update our rocker studs.

Our three test cams were Comp Cam's largest, most powerful streethydraulic roller bumpsticks. With more than half an inch of lift and 230degrees of duration at 0.050 inch of valve lift, these cams are designedfor hot street and mild strip cars. They'd also do well in open-trackapplications, and we'd be tempted to run the 266 in a slalom engine.

When adding displacement by either boring or stroking, the engine puts a larger demand on the intake and exhaust manifolding. This is no secret, as a larger engine will obviously consume more air than a smaller one, and that air has to pass through the manifolds. Thus, a larger engine can use larger-diameter runners without the airflow becoming "lazy." The same is true for camming. A 347 can tolerate more lift and duration than a 302. Practically speaking, this means a 347 more easily produces vacuum to run the brake booster and smooth the idle. It also means the 347 draws harder through the lift and duration limits set by the camshaft. So, a 347 owner can use a slightly larger camshaft than the 302 owner, yet still retain workable vacuum and idle characteristics.

To put those concepts to the test, we had Westech put our 347 on the dyno and select three of the largest hydraulic roller cams in the Comp Cams catalog, noted in Table One. These are large, power-producing camshafts aimed at serious street and moderate strip engines. They're not going to pass smog and won't win any fuel-mileage contests when putting around town (although they are quite fuel efficient when making power--it's just that they make so much power). What they will do is make great power and torque when combined with other good-breathing parts.

We ordered two of our cams from Comp as cam-only parts, and one cam aspart of a kit. The latter move netted us everything necessary to installand update the valvetrain short of the rocker arms and pushrods. Wealready had roller rocker arms (also from Comp), and the pushrods arefit-to-length units that will vary from engine to engine. Ours workedout to be 8.700 inches long.

Note that all three cams have lobe separation angles of 112 degrees. This is a key measurement when building an electronically fuel-injected engine, as the computer can't handle the erratic vacuum signal produced by a lobe- separation angle greater than 112 degrees. And, because we were running these cams with our standard EEC IV engine management, this was a limit we needed to respect.

Also, the cams increase lift and duration by fairly constant values. Each step up in grind brings about 0.010 inch more valve lift, along with 8 degrees more duration. Of the two, it's the duration that makes disproportionately more power, because a little more duration has the valve open longer and the increased lift has it opened farther.

Comp supplies two new dowel pins and a cam card with each cam. The dowelpins simply tap into place. The length depends on what sort of timingchain and fuel pump eccentric may be fitted. The cam card has all thecam numbers on it; it's necessary when dialing in the cam. We installedall our test cams 2 degrees retarded.

A limitation of our test engine was its 24-lb/hr fuel injectors. These are simply not large enough for the mid-400hp range expected from these cams, so larger injectors were required. We could have gotten by with 30-lb/hr units, but while trying to dig up anything that would work, we were kindly offered a set of 36-lb/hr injectors by our old friend Richard Holdener. As Richard had the matching mass airflow meter in 36-lb/hr trim, it didn't take any additional thought to say "yes, thank you" to his offer.

You don't often see Steve Brule with his hands in his pockets, butthat's the way we caught him as he eyeballed the just completed engineone last time before he hit the starter switch. Our 347 CHP StreetFighter short-block and Ford Racing Performance Parts Z304 heads haveproven to be reliable power builders.

Going into the test, there was some talk around Westech that a long-runner EFI intake manifold such as the Performer RPM II we would run would control the power output rather than the camshaft. Obviously, every part in the engine has to be matched to produce maximum power, and the idea of the intake manifold providing the choke sounded like a real threat. Many times we've had intake and exhaust manifolds prove the airflow restriction on more powerful engines, especially on early long-runner EFI intakes such as the GT-40 or Cobra combined with higher rpm.

In the end, we figured we were going to find out if the Performer RPM II on our 347 would provide the main power characteristics or if the cams would. To determine this, we would, of course, run the RPM II throughout the test. We also had an old Edelbrock Victor 5.0 EFI intake with much shorter (but narrower) runners on hand, as well as a Performer RPM Air Gap carbureted intake. Substituting those intakes ought to tell if the RPM II was proving restrictive.

When we moved from 24- to 36-lb/hr injectors, we also wanted to movefrom our Cobra kit computer, which FRPP optimized for 24-poundinjectors. We ended up liberating this A3M1 computer out of anotherproject car in the shop. It ran great. It's possible the processor swapaccounted for some power, as the Cobra's processor is notoriouslyconservative.


We began with the 266 cam and Performer RPM II intake. The engine lit immediately and settled into a nice, smooth idle as the cold-start program ran the engine in open loop. While the idle was a tad high, its smoothness proved the cam was easily capable of acceptable street manners. Unfortunately, by running without oxygen sensors (they're on our to-do list), the engine quickly took on a ragged attitude at idle once the cold start program melded into the regular idle code.

Of course, it didn't take but a minute for the nagging problems to show up. First the dyno's Lambda A/F meter went Tango Uniform, but simply recycling the controller got it going again. Next the engine went rough and 20 hp lazy on a couple of exploratory power pulls. Eyeballing around the spaghetti atop the engine revealed the No. 8 injector's wiring lead had come adrift, so that too was an easy cure. Then it was on to set the ignition timing. For the firefly-brief power pulls on the dyno, which is a near totally controlled environment, it's pos-sible to run incredible timing. We ended up at 20 degrees initial for best power.

Now that Edelbrock's Performer RPM II intake wears a triangle-shaped,expanding crossover "tube" as do other deep-breathing EFI intakes, ittoo offers impressive EFI performance that matches carbureted horsepowerwhile still delivering EFI torque. There's no power reason not to runEFI anymore.

Once dialed in, our 347 ran 1 hp less from the last time we ran the engine with its Coast High Performance cam, for a power peak of 426 hp at 6,000 rpm. Continuing to split hairs, we noted the Comp cam was ahead on torque by 2 lb-ft, peaking at 4,500 rpm with 440 lb-ft. There was nothing overly scientific about this, but interesting that the two different cams came out so close.

We figured this "small" Comp cam in our test did a dandy job making power. Somewhere between 400 and 450 hp is ideal for a real-world Mustang because it is all the power street tires can hold onto without spinning at the drop of the throttle. It's entertaining without being overbearing.

Wanting much more rpm, camming, and cubic inches, the Victor 5.0 intakemanifold fell 20 hp and lb-ft shy of both our injected and carburetedPerformer RPM combinations. This is definitely a manifold meant forrevving past 7,000 rpm.


Moving on to the middle--or 274--cam was done by stripping the engine on the dyno, swapping cams, and rebuilding the engine right there on the "pump," as water-brake dynos are often called. A couple hours later, we were ready to make noise again. As the engine fired, we noted 10 inches of manifold vacuum at a slow idle with this "middle" 274 cam. We'll confess to not noting the manifold vacuum registered by the 266 cam, but it had to equal or slightly exceed the 274 in vacuum production. Either way, 10 inches of vacuum is just sufficient to keep the brake booster powered up.

With the ignition timing set to 20 degrees initial, we made several power pulls to ensure repeatability while seeing what the combination would do. It turned out the larger 274 was making less power than the 266 all the way up to 5,200 rpm, where the horsepower curves crossed over so that the 274 ended up peaking several horsepower higher than the 266. The improvement change was subtle, and Westech engine expert and dyno driver Steve Brule seemed to think we were seeing the limitation of the intake manifold at this point. As he put it, "This doesn't surprise me. It's exactly what I'd expect a cam to do [lose power down low and gain it at higher rpm], but it's trying to move the torque peak up and the intake manifold isn't letting it."

To investigate, we turned the timing up another 2 degrees, for a stout 22 degrees initial. This added several more horsepower at the top of the curve for a total gain of 11 hp over the 266 cam, while torque was down 6 lb-ft from the 266 cam. Furthermore, the torque peak had moved up 300 rpm, from 4,500 to 4,800 rpm. After gazing at these figures, Steve was more comfortable with the idea that we were simply seeing a typical torque-for-horsepower trade, which is what camming-up typically does. So is the trade worth it? And is the area under the curve greater for the 274 or the 266? Does the larger cam gain too little too late, or is it just that much better? To find out, we averaged the power figures from across the powerband.

CAM/Avg. Power/Avg. Torque

This was also our first time bolting up our Edelbrock Performer RPM AirGap intake manifold. As always from Edelbrock, it looked good, fit, andmade power. It would seem the ideal street carburetor intake for all butthe most extreme applications.

Not much to say for the larger 274 cam, huh? Now let's talk application. If you have a street car, you're much more interested in the midrange. If you have a drag car, then the top end is all that really matters. In the midrange, the two cams are making identical numbers--or nearly so--thus, there's little sense in averaging them. As we go up the tach, however, the larger 274 cam begins to pull slightly away. A good way of measuring such an increase is to average the power figures over the rpm range in question. So we had the dyno average the power between 4,500 and 6,000 rpm.

CAM/Avg. Power/Avg. Torque
266 /412/ 414
274/ 415/ 417

Yes, we're sort of splitting hairs, but there's a definite bend in the fig- ures in favor of the 274 cam when we consider just the high rpm. Taking that idea a bit further along and averaging from 5,000 to 6,000 rpm, we see the big cam helps to the tune of 6 hp and 6 lb-ft.

Westech is supported by Demon carburetors, so the company has a largesupply of the gas mixers on hand. This 750 Speed Demon had already stoodthe watch on numerous dyno tests, so it was about perfectly dialed inthe moment we bolted it on. The dual float level windows aresurprisingly handy when tuning or troubleshooting.

CAM/Avg. Power/Avg. Torque
266/ 422 /403
274 /428/ 409

So, for a drag car with high-rpm gearing, the larger cam would help--not hugely help, but some. For a road racer, it would depend on the track. Big, open venues where the engine is kept buzzing all the time would favor the larger cam; tighter tracks favor the shorter bumpstick. Of course, there are gearing changes to accommodate either cam, but we'll leave all that bench racing up to you.

After years of fitting the entire front engine dress on EFIsmall-blocks, we've accepted reality and begun testing with CSI electricwater pumps. These pumps eliminate hours of bolting brackets,alternators, and idler pulleys on and off. They also eliminate theapproximately 8 hp lost to the stock, belt-driven water pump. Add thatback in along with catalytic converter, street muffler, tailpipe, andintake hose restrictions if trying to duplicate as-installed flywheelhorsepower figures. All told, we guesstimate our dyno mule registers 20-25 hp higher at the flywheel than it would running in a car.

Victor 5.0

With all our discussion about intakes and power curves, along with the closeness of the 266 and 274 cams in practical power output, we decided to try a sportier intake. Rummaging through the parts bins at Westech, we unearthed an Edelbrock Victor 5.0 and swapped this unported, box-stock intake for the Performer RPM II we'd been running. This takes a bit of doing because the throttle body is moved toward the center of the engine due to the Victor 5.0's shorter crossover tube. We had to fiddle with the throttle linkage bracket and find the bolts, so it took us awhile. If doing the same job with a new intake, you'd have all the bolts, gaskets, and so on.

Looking at the Victor 5.0 as we installed it, we began to wonder how it would work. It was designed for higher rpm than the stock 6,250-rpm limiter we were working with and thus has short runners. Furthermore, the crossover tube that takes air from the throttle body to the runners is narrow. Offsetting all this was the Victor's small cross-sectional area on the runners. Maybe the short runners would pro- mote a bit of torque due to the increased air velocity they were obviously going to lose.

With the ignition timing set to 21.5 degrees, we let her rip and she fell flat on her face. Torque was down as much as 20 lb-ft and the horsepower just lie there as well. The intake was simply designed to run at much higher rpm than what we were showing it.

As usual, our MSD billet distributor and wiring harness did the sparkingthing without fail. This is our carbureted distributor; we use an EFIbillet MSD distributor when running injected.

After laying the previous runs over these runs, it's clear the Victor has nothing to offer this combination and would lose big torque and horsepower unless paired with a considerably larger or higher rpm engine.

In fact, that's what Richard Holdener said. He noted good results with the Victor when coupled with a street mechanical-roller cam and 408 ci. And then the torque it loses way down at 2,500 rpm is good because traction is such a problem with these monster engines that a little less torque down low helps. So, with a mechanical-roller cam, cubes, and rpm, the Victor Jr. does well, but not so on a 6,500-rpm street engine


We removed the Victor and 274 cam, then put the engine back together with the 282 cam and Performer RPM II intake. This is the largest cam in our series, and it's sort of getting out there for a street hydraulic roller.

DYNO RESULTS! Part 1 - Click to Enlarge
DYNO RESULTS! Part 2 - Click to Enlarge

By now the drill of getting the engine together, setting the timing, and running the engine for repeatable numbers was routine. It didn't take long to arrive at representative 282 numbers, which were 447 hp way up at 6,000 rpm and 436 lb-ft of torque at a lofty 5,000 rpm, for a good incremental increase in horsepower. Being conservative, we'd think of it as a 445hp engine and not have to worry about making any excuses.

While not well configured to judge such things the way we were running it on the dyno, we can say the 282 cam wasn't that far behind the other two in driveability or idle characteristics we witnessed. Vacuum was still hovering around 10 inches, and we wouldn't shy away from running this beast between stoplights when necessary.

Cam Specs - Click to Enlarge

As always, the torque peak had migrated upward as the cams grew larger, and by this point the torque peak was 500 rpm higher than the first cam. As is also typical of cam changes, no more torque was produced, but rather higher rpm was supported, which makes more horsepower.

Carbureted Flashback

With our three cams running fuel injected, the last question was how our cams would fair carbureted. We've long observed that carburetors and their short-runner intake manifolds did a great job of making top-end power, but they gave up meaningful torque in the low and mid range. However, the latest thinking says the current crop of EFI intakes were approaching parity with carburetion, and we wanted to test that concept--or see if the EFI was a choke.

It was simple to remove the Performer RPM II injection manifold and its attendant wires, injectors, fuel rails, regulators, and whatnot, then bolt on a brand-new, box-stock Performer RPM Air Gap intake with one of Barry Grant's Demon 750 Speed Demon carburetors. This gave us a direct comparison between the latest Edelbrock fuel-injected and carbureted intake manifolds, and thus an accurate look at the difference between EFI and carburetion.

We used the usual Westech dyno exhaust system for this test, so there'sno change there from the last several years. The big 3-inch pipes andmufflers offer no restriction over open headers

It didn't take long to physically change the equipment, and soon we had the carburetor dialed in--a process made simple by the dyno's plentiful air and fuel data. Interestingly, with the carburetor, the big 282 cam pulled a steady 12 inches of vacuum at an admittedly frenetic 1,150-rpm idle.

On our first official power pull we saw 457 hp, with the torque down slightly from the EFI value. By the time we had the oil fully warmed and our official power pulls in the bag, the horsepower was 462 at the peak and the torque 433 lb-ft.

Comparing the EFI and carburetion, we were impressed by how close the two ran. Gone are the days of a big torque advantage and a horsepower deficit for EFI. When you get down to it, these two manifolds barely swapped torque and power, with the "advantage"--it's fairly small to call it that--going to EFI for all but the peak horsepower number. This is great news for everyone, as now the street enthusiast can take the route for him and keep the fuel injection without fear of losing power to a carbureted combination. The EFI guy also nets the driveability, legality, and fuel economy pluses of fuel injection.

The carbureted guy doesn't lose either. For an '85-or-older Mustang, or the dedicated strip car where simplicity rules, a carburetor fan can run the latest stuff and know he's not giving up torque for his easy living.

So there you have it--425-450 hp, injected or carbureted, from your choice of hot, hydraulic roller cams. Not a bad place to be. 5.0