Tom Wilson
December 1, 2005

Horse Sense:
We received a phone call from Carroll Shelby once. When the secretary said there was, "a Carroll Shelby" on line five we didn't believe it. Would you? Turned out he was calling to chew us out for calling it an, "AC Cobra."

You have to admire Carroll Shelby for still playing the game. After all, the man flew during World War II, had a great career as a driver (he won Le Mans way back in 1959 in an Aston Martin), built the most revered sports cars in the U.S. and possibly world-wide history (the GT-350 Shelby, 289, and 427 Cobras), won the sports car constructor's championship in 1965 (the one and only time by an American), helped make the original GT40 a winner, brews great chili, runs a successful philanthropy for children, and is back with his name front and center on the new Shelby Cobra GT500 from Ford.

That might be enough for a lifetime, but Carroll can still be found running from one event to another, building cars, working his children's fund, and generally staying in the spotlight. And to that end, here we are, tagging along on a dyno test of one of Carroll's personal engines. At press time, it wasn't known if it was going in a '66 GT350 convertible or a Cobra (the two-seater with no roof, not the Mustang namesake more common in these pages), but it was going to be one or the other.

No matter, really, as with 525-plus lb-ft of torque and always more than 500 hp, it will do the job nicely in either chassis.

Built at Trans Am Racing by Mark Jeffery, the engine is a 427ci stroker Windsor. Befitting the early cars it is destined for, the engine is carbureted, and as such, could go in any sort of modern toy-car Mustang. As you'll see, the combination of generous displacement and conservative compression and camming makes this an easy-to-drive, yet satisfyingly powerful, powerplant. Lightly constrained with big dyno mufflers, the 427 was between 505 and 526 hp depending on the induction, and growled out 518 lb-ft of torque at its worst and 531 lb-ft at its best. Considering the small camshaft, these are good, car-moving numbers that ought to replicate easily in the real world of 211/42-inch exhaust, tight-fitting hoods, and the like. And unmuffled, we saw 529 hp and 541 lb-ft, which ought to be fun on track day.

Our testing included an exceptionally well ported Edelbrock Torker II single-plane intake, a not so heavily modified Edelbrock RPM dual-plane intake, and a pair of carburetors-a Pro Form and a 650 Mighty Demon. While the trick, ported intakes are a bit of a curve ball, the single- versus dual-plane data is always good to review (modern dual-planes are tough to beat), and we had enough time to fiddle with the mufflers, too.

The Tinkertoy
Engine builders prefer seasoned blocks, and starting life in 1972, we figure this regular two-bolt production block is as seasoned as a Thanksgiving turkey. The expected 0.030-inch overbore takes the cylinders to 4.030-inch diameter; while the SCAT cast steel crankshaft swings a generous 4.170-inch stroke to give the 427-inch working volume.

Mark didn't go to the trouble, or considerable expense, of internally balancing the early style 28-ounce crank, but did have a close tolerance balance worked into the reciprocating assembly. The main journals are stock Ford dimensions, so no rpm heroics are planned. The rod journals are the expected 2.100-inch small Chevy size, with matching I-beam SCAT connecting rods. A Romac damper was fitted to the crank snout, while a steel main stud girdle from Probe ties the main caps together.

The pistons are from Probe, "the bathtub design," Mark says. For this engine, Probe provided the finished piston, but Mark says he typically makes his own dished pistons by buying flat-tops then milling out the dome as required. For this engine, the off-the-shelf Probes worked out fine as-is, and provide a 10.7:1 compression ratio.

Cometic built the three-piece metal head gaskets, with Mark opting for the 0.036-inch thick offerings. His primary goal with these is to keep the quench area thinner than 0.040 inch to maximize in-cylinder charge motion. That reduces detonation; a concern when the compression is creeping as high as this engine's with pump gas in the tank. Conventional Sealed Power moly rings and Clevite rod and main bearings were used.

For cylinder heads, Mark chose Edelbrock Victor Juniors. Now, like the two intake manifolds we tested, the heads have been ported and the valve package enlarged and custom valve angles cut. The porting science comes from Roger Helgesen, a freelance airflow specialist Mark collaborates with, while the vast majority of making the aluminum chips fly was Mark's doing. Thus, the Victor Juniors have been fully ported and 2.08x1.625-inch valves fitted in favor of the stock 2.05x1.600-inch units. Mark reports that he tried 2.100-inch intake valves, but there was nothing gained by it.

A careful five-angle valve job finished the heads, which delivered 312 cfm on the intake and 242 cfm on the exhaust. That's up from 275/204 cfm out of the box. This is a good flowing set of heads, obviously.

Valvetrain duties are handled entirely by Comp Cams hardware. With the small hydraulic-roller cam you'd think the valvetrain was pretty standard stuff, but there's a fair amount of thought-and money-put into it. Beehive valvesprings were chosen, which almost doubles the spring budget in return for a progressive increase in spring rate near coil bind and combined with oval wire, better control of harmonics. These springs also have a relatively high seat pressure, but at 300 in-lb over the nose, are not high-effort springs when the valve is open. Titanium retainers keep the package light and Comp 1.6 Magnum rollers provide the rocking motion.

As for the cam, it's a custom grind from Comp, but just barely custom. A conservative 248 Extreme Energy cam is the starting point, and then Roger has the lobe centers put at 108 instead of the standard 110 degrees. This is the only change, but does make the cam a custom piece. Roger also pointed out that he avoids dual-pattern grinds because with good porting the exhaust-to-intake flow ratio is up to 80 percent, so the dual pattern isn't needed. The cam was installed 4 degrees advanced.

Of the two intakes, we started with the Edelbrock Torker II. This single-plane casting is not the newest in the Edelbrock catalog, and comes out of the box with miniscule runners (they seem matched to the almost 45-year-old 289 head's intake port). Many of you are asking why anyone would run such an old-school intake on a stroker Windsor, and the answer is two-fold. For starters, the Torker II is a shallow-profile unit, so it clears Cobra hoods. Secondly, with all that meat, the Torker II leaves all sorts of porting room.

So, out of the box, the Torker II is hopeless on modern, large-port cylinder heads feeding equally large short-blocks. But after being introduced to the Trans Am Racing carbide bit, things looked up. A simple port match is worth 50 cfm of flow, Mark says, and fully ported, like this one, there is a generous 363 cfm of flow on tap.

Of course, great flow or not, the Torker II is a single-plane intake, and low-end power is bound to suffer accordingly. Enter Edelbrock's much newer Performer RPM Air Gap intake manifold. A dual-plane design, the Performer RPM was designed to equal the top-end power of older manifolds, such as the Torker II, while providing excellent low-end power as well. One price it pays for that is height-it towers over the Torker II.

The Performer RPM Air Gap Mark brought along had been ported as well, but not as substantially as the Torker II because it wasn't necessary. He characterized it as more of a port matching than anything else.

Topping everything Mark had in his bag of tricks was a pair of carburetors to try. The first was a Holley-like unit from Pro Form. It uses Holley metering plates and bowls with the Pro Form throttle body. Mark says the Pro Form flows 90 cfm more than a standard 750 Holley, and is, "a nice piece for the price." The second carburetor was a stock 750 Mighty Demon.

This engine benefited from experience and talent when it came to matching and porting the critical head and intake manifold castings. On the other hand, it was cammed fairly short to make it easily streetable, so anyone looking to up his game for strip or track duty ought to find some easy horsepower in a more aggressive bumpstick. In the meantime, we're sure Carroll Shelby will have a good time with it just as it is.

Dyno Results

Torker II, no spacer, Pro Form, mufflersWilson Manifolds 1-inch Spacer750 Mighty DemonPerformer RPM Air Gap No Spacer
RPMHPTQHPTQHPTQHPTQ
3,500314472311466313470343515
3,600322469318464321469357521
3,700332471333473338479371526
3,800351485349483346479386533
3,900363489362487363489397535
4,000377495375492378496402528
4,100389499384492390500413528
4,200403504400500405507422528
4,300416508420514423517430526
4,400431515435519437521441527
4,500444519448523450526447521
4,600454518461527461526456520
4,700463518469525470525464519
4,800470515476521478523470515
4,900475509484519485520476510
5,000479504489513492516483507
5,100485499494509497511490505
5,200489494499504502507497502
5,300492487501496502497498494
5,400494480504490504490501488
5,500496474506483506484505482
5,600501470510479508476511479
5,700508468514474514473512472
5,800508460516468515466513464
5,900508453517460515458514457
6,000508445513449516451513449
6,100505435513441516444512441

Barry Grant spacerNo MufflersPipe ExtensionsDifference
Baseline vs. Final Pull
RPMHPTQHPTQHPYQHPTQ
3,500 336 504 335 502 344 516 30 44
3,600 345 504 350 510 357 521 35 52
3,700 362 514 371 527 372 527 40 56
3,800 375 518 383 529 384 531 33 46
3,900 386 521 393 529 394 531 31 42
4,000 397 521 404 531 403 530 26 35
4,100 406 520 416 533 412 528 23 29
4,200 418 523 429 537 422 527 19 23
4,300 432 528 443 541 437 534 21 26
4,400 445 531 452 539 451 538 20 23
4,500 454 530 462 539 459 535 15 16
4,600 462 528 470 536 467 533 13 15
4,700 473 529 478 534 477 533 14 15
4,800 480 526 486 532 484 529 14 14
4,900 488 523 492 527 492 527 17 18
5,000 494 519 498 523 498 523 19 19
5,100 500 515 504 519 504 519 19 20
5,200 504 509 509 514 507 512 18 18
5,300 508 503 512 507 511 506 19 19
5,400 510 496 518 504 515 501 21 21
5,500 515 492 523 500 519 496 23 22
5,600 520 488 528 495 523 491 22 21
5,700 522 481 529 488 526 484 18 16
5,800 525 475 531 481 526 476 18 16
5,900 525 468 528 470 523 466 15 13
6,000 526 460 527 461 525 459 17 14
6,100 524 451 527 454 523 450 18 15

Dyno Results Cont.
Mark's plan for his day on Westech's 901 Superflow dyno was simple. He simply wanted to verify the power output of the 427, then try the two intakes and carburetors. He started with the Torker II intake and Pro Form carb. After the usual timing and jetting sessions, the power came in at 508 hp and 519 lb-ft of torque. As the charts show, the power was relatively soft down low with the single-plane intake, but in a light car, this would not be much of an issue as a 427 inherently makes some torque when cammed this short.

Looking for more top end via a carburetor spacer, Mark tried a 1-inch combo spacer from Wilson Manifolds. Dyno operator and engine builder Steve Brule at Westech noted he's had good results from this tapered, 4-into-1 spacer. Sure enough, starting at 4,650 rpm, the 427 liked the extra plenum volume, resulting in peaks of 517 hp and 527 lb-ft of torque. These are good returns for such a simple change, but we should note the spacer didn't do anything for the low-end.

Next, the Demon was substituted for the Pro Form, still with the Wilson spacer in place. Out of the box, the Demon proved just a little rich on the air/fuel mixture compared to the Pro Form (but nothing that wouldn't work just fine in the real world). It was a couple of horsepower ahead of the Pro Form across the tach, too. Rejetting it three steps leaner resulted in just a little more power, and the official results after multiple pulls put the Demon approximately 6 hp ahead of the Pro Form. That is not a huge improvement, but the Pro Form is a good semi-custom piece, so we figured the off-the-shelf Demon was doing OK.

With the clock ticking away, it was on to the Performer RPM Air Gap manifold. The Demon carburetor and no spacer was the first configuration tested, and what a difference in the bottom-end power. At 3,500 rpm the RPM Air Gap was ahead by no less than 30 hp and 45 lb-ft of torque. There simply was no comparison at the low-end. Advantages this large are nearly impossible to beat with top-end power, and with the Torker II ahead by just 3 hp and 3 lb-ft of torque at 6,000 rpm, the winner of this contest was clear to see.

In fact, the crossover point-where the Torker II out-performed the RPM Air Gap-was somewhere north of 5,000 rpm. That shows just how good the modern dual-plane intakes are. Unless you have the engine cammed and the chassis geared for stratospheric rpm, the single plane intakes give up too much torque in the low and midrange for almost all applications. That is, unless you have too much torque down low and can't hang onto the tires. But there aren't many of us with that problem.

We should also mention this combination idled happily enough at 950 rpm with 8 In Hg showing on the vacuum gauge. It would idle down at 850 if you asked it to; in any case, easily streetable.

If you had pickup-truck-hood clearance, fitting a simple open 1-inch Barry Grant spacer (not the Wilson tapered 4-into-1 unit) will nicely tune the RPM Air Gap for useful top-end power. It's a classic case of rocking the curve, at 4,400 rpm in this case. Below that, the spacer gave up some power-it was down 7 hp and 11 lb-ft of torque at 3,500 rpm, for example-but above 4,400 rpm it was definitely making more umph-plus 13 hp and 12 lb-ft of torque at 6,000 rpm.

The final tests of the day were to remove the 3-inch dyno mufflers, letting the headers blast into just 18-inches of 3-inch tubing, and ultimately, adding another 18-inches or so of pipe onto those. Removing the mufflers gained 12 hp in the mid-range, 5 hp at the top end, and brought the power peak down several hundred rpm, always a good sign when the power goes up as well. However, the mufflers seemed to take the torque with them, as their absence positively murdered the bottom end; at 2,500 rpm, power was down 40 hp.

That's why the pipe extensions were tried. They helped a bunch simply by adding pipe length, but curiously did not find all the low-end power the mufflers could make. Compared to the short open collectors, however, the longer pipes returned 19 lost horsepower at 3,000 rpm and 9 hp at 3,500 rpm. Results like this make us think we should all be paying more attention to exhaust tuning than we normally do with the typical open-exhaust hobby car. There could be real power there.