Pontiac Camshaft Engineering - Latin Square - Derived Camshafts Revealed

In the late '60s, Pontiac Special Projects certainly had its work cut out for it. The development program for the SCCA Trans Am-based 303 Pontiac was one that was fraught with setbacks, low project budgets, and some engineering deadends.

Perhaps the biggest hurdle was the initial use of the small-chamber version of the tunnel-port Ram Air V cylinder head on the short-deck block. If the ports didn't produce sufficient velocity in a drag-oriented 400ci engine, how were they supposed to work on a road-racing 5.0L V-8, which needed a much broader powerband?

The bottom line was, they didn't. Several fixes were tried, including a smaller exhaust valve to increase scavenging and other velocity- enhancing measures. In the end, they realized this path was never going to be competitive. A new strategy was required.

Reaching back into the parts bins, the engineers found a much more suitable alternative-the smaller, but still large, Pontiac Ram Air IV cylinder heads. Here was a readily available head that could be ported to produce the desired power levels and, more importantly, in a wide enough rpm band to be suitable for road race use.

Actually, it was more than just the heads that were used. Indeed, the whole top end from the Ram Air IV was swapped on the short-deck 303. The heads, of course, bolted right on the short-deck block, and the aluminum intake had its runners shortened to fit. Later versions of the 303 used the standard-deck block, and no other alterations were needed.

In the September, October, and November '05 issues of HPP, we presented our interview with former Pontiac Special Projects Engineer Tom Nell. In the course of the taping, he went into great detail on the process they used in camshaft selection for the Ram Air IV-headed 303 engines. His description was so fascinating and the process so brilliant that we decided to run that section of the interview as a separate article, which we proudly present at this time.

HPP: Please explain how the Latin Square camshafts came into being. It has been touched upon in some books in the past but never explained in detail.

Nell: Honestly, Jeff Young gets all the credit because he was the guy who did this program. We were trying to find out what the engine wanted. Jeff worked with a guy named Don Tewless, who was just starting a cam company called General Kinetics, in Detroit. Tewless would make us masters, and Jeff had him make up I don't know how many cams to run a basic test to determine what was the most important factor in valve timing that influenced the engine's power. Was it the lift or the duration, where you opened the valve, where you closed it, whatever?

Anyway, the end result was the engine said that of 96 percent of what you did, where you close the valve is what I like. So we said, "OK, where you close the valve is the most important thing, so that's what we're going to base this on." I don't know where this "Latin Square" thing came up, but here's how it works:

Take a piece of paper and a pencil and draw a square, maybe 2x2 inches (see figure 1). Next, within that square, draw a circle that touches all four sides. Then, draw two diagonals through the center of the circle from point to point on the square. At the center of the X, make a dot and write down the number 1. Keep in mind, that the numbers used here are representative only and are chosen for simplicity's sake.

Next, go up to 12 o'clock where the circle touches the topside of the box and make a little dot, and make that number 2. As you come around the circle in a clockwise direction, where the diagonal crosses the circle, some 45 degrees later, that is number 3. Number 4 is at 3 o'clock, 5 is where the diagonal crosses at about 4:30. Keep doing that right around the circle at the points and you end up with a total of nine points.

Now, what we were looking for, the test criteria, was to find a cam that would give us the maximum area under the torque curve from a certain rpm to a certain rpm. We wanted the maximum area under the torque curve between 5,000 and 8,000 rpm. We didn't care what the peak was; we just wanted the most area under the curve. The idea was to alter intake and exhaust closing points 5 and 7 degrees in both directions from a baseline grind.

The baseline was Cam #1 (see figure 2). It had 310/330 degrees advertised duration on a 105-degree centerline. All the cams had 0.500-inch lift because that was the limit we could run with the domed pistons we needed to get the compression up. That gave an intake closing point of 80 degrees and an exhaust closing point of 60 degrees.