Just in case you aren’t up on current events in the LS community, there seems to be a divide wedged firmly between cathedral port and rectangular port enthusiasts. For those unfamiliar with the terms, they refer to the shape of the ports in the most popular LS head configurations. Cathedral port heads came on the original (Gen III) LS1, but were also used on the performance-oriented LS2 and LS6 versions. Cathedral port heads were bolted on literally millions of truck variants, which included (among others) the 4.8L LR4, the 5.3L LM7, and 6.0L LQ4. By contrast, rectangular port heads were offered on the Gen IV version of the LS, the most famous being the LS3, but they were also installed on the L92 and L76 truck engines.
When GM introduced the LS3-based rectangular port heads they were the best flowing small-block heads ever offered on a production vehicle (although they were eventually surpassed by the 7.0L LS7 heads). The best factory cathedral port heads (243, 799, and 317 castings) flowed near 250 cfm, which represented a huge step up from the conventional small-block heads that preceded them. However, the LS3 heads flowed a whopping 317 cfm. This begs the question, how much power are the rectangular port heads really worth, and do the power gains remain consistent once boost is added to the equation?
To answer the proposed question about the rectangular port and cathedral port heads we needed the following: a test engine, two sets of heads, and a turbo system. The test engine actually solved another of our needs, as the crate LS3 engine featured a pair of Chevrolet Performance “as-cast” LS3 heads. In preparation for the test we augmented the LS3 short-block with a cam upgrade that featured a 0.617/0.624-inch lift split, a 231/243-degree duration split, and 113-degree LSA. In preparation for boost, the short-block also received a set of Fel-Pro MLS head gaskets and ARP head studs.
Though some might argue with our choice of cathedral port cylinder heads, we selected the 317 truck heads primarily because they offered the best combination of flow and chamber volume. The 71.5cc chambers nearly matched the 70cc chambers of our LS3 heads and the 317 heads flowed just under 250 cfm (just a few ticks below the 799/243 heads we’ve tested). The 64cc chambers of the 799/243 heads would increase the static compression and certainly help the power production of the cathedral port heads, but we liked matching the compression ratio as best we could for this test. Both the 317 and LS3 heads received valvespring upgrades to work with the revised cam profile and boost.
While the cathedral port camp will lament us not using the 243 heads, the LS3 camp will take exception to our use of a FAST intake with the 317s. The reason for using the FAST intake was that the LS3 heads were run with what is arguably the best factory intake ever produced. The LS3 intake is better on that combination than any of the respective cathedral port (or LS7) intakes (yes, including the TrailBlazer SS). Rather than restrict the 317 heads with a substandard intake, we installed what is basically the LS3 equivalent for the cathedral port heads. There, both camps have something to complain about.
Additional components used in testing included a set of 120-pound Holley injectors, 1 7/8-inch Hooker headers, and a Holley Dominator ECU. The modified LS3 engine was run with both sets of heads in naturally aspirated and turbocharged trim. Run naturally aspirated with the 317 heads, the 6.2L produced 551 hp at 6,600 rpm and 507 lb-ft of torque at 4,900 rpm. Run with the LS3 heads, the peak numbers jumped to 584 hp at 6,700 rpm and 526 lb-ft of torque at 5,300 rpm. The high-flow LS3 heads sure made themselves known on the naturally aspirated 6.2L, but how would they do once boost was added?
The application of boost to our LS3 came from a single Precision 7675 turbo. Capable of supporting over 1,100 hp, the turbo is more than capable of supporting the power levels our 6.2L would achieve. All we wanted to do was to find out if the gains offered by the LS3 heads in naturally aspirated trim remained under boost. In this case, we set the boost pressure at a conservative 7 psi using a pair of Hyper-Gate45 wastegates from Turbosmart. Turbosmart also supplied a single Race Port blow-off valve for our turbo LS. The homemade turbo system featured a pair of stainless turbo headers from DNA coupled to a custom 3-inch Y-pipe crossover. The crossover was home to the pair of wastegates as well as the T4 turbo flange.
The exhaust system for the turbo consisted of a single 4-inch section of tubing, while the cold side featured 3.5-inch aluminum feeding an air-to-water intercooler from ProCharger. We measured both boost and backpressure during testing, as we made no effort to adjust the boost. We relied on the 7-psi wastegate spring to control the boost, and any variation would be a result of changes in backpressure. Backpressure in the system works with the boost pressure to open the wastegate valve. An increase in backpressure will effectively open the wastegate earlier for any given spring setting.
After installation of the turbo system, we ran the LS3 once again with the two cylinder head combinations. Equipped with the 317 heads, the turbo 6.2L produced 807 hp at 6,800 rpm and 747 lb-ft of torque at 4,900 rpm. The boost pressure peaked at 7.6 psi, before dropping to 6.9 psi at 7,000 rpm. The backpressure started out at just 6.6 psi (lower than the boost pressure), but soon peaked at 13.8 psi (giving a 2:1 ratio of backpressure to boost pressure). After installation of the LS3 heads, the peak numbers jumped to 826 hp at 6,700 rpm and 771 lb-ft of torque at 4,600 rpm. Just as they had on the naturally aspirated engine, the LS3 heads offered more power under boost, but the extra power also increased the backpressure. The boost pressure peaked at 7.6 psi, but dropped to 6.4 psi at 7,000 rpm. The backpressure started out at 6.7 psi, but eventually climbed to a peak of 14.2 psi (a ratio of just over 2.2:1). The increase in backpressure caused an early opening of the wastegate (by 0.5 psi) which would only add to the gains offered by the LS3 heads, but would further increase backpressure.
This test shows the LS3 heads are certainly better than the 317 heads on both a naturally aspirated and a turbo LS. But what about them against a set of 243 heads? Then, the question becomes, what about a set of ported LS3 heads against ported cathedral port heads? The questions and dyno testing never ends.
1. The test mule used for the test was a factory LS3 crate engine. We added a set of Fel-Pro MLS head gaskets and ARP head studs (not shown). Although the LS3 cam was a decent performer, we installed a performance-oriented hydraulic roller cam. The new profile featured a 0.617/0.624-inch lift split, a 231/243-degree duration split, and 113-degree LSA.
2. Both the stock LS3 (shown) and 317 heads received a valvespring upgrade to work with the cam upgrade.
3. Both heads were run on the flow bench prior to installation on the test engine. The 317 heads flowed 245 cfm and the LS3 heads flowed a whopping 317 cfm.
4. The stock LS3 heads were run with the impressive factory LS3 intake (shown). The 317 heads received a FAST truck-style intake manifold.
5. To ensure adequate fuel delivery, especially under boost, we installed a set of Holley 120-pound injectors. For the dyno tests, engine management was under the control of a Holley Dominator ECU.
6. First up on the dyno were the 317 heads. Run in naturally aspirated trim, the 317-headed LS3 produced 551 hp and 507 lb-ft of torque.
7. After the test, off came the intake to allow access to the 317 heads. In truth, no one would replace the stock LS3 heads with 317s, but they would definitely go the other way on a 6.0L truck engine.
8. Run on the dyno with the LS3 heads, the aluminum 6.2L produced 584 hp and 526 lb-ft of torque. After comparing the heads in naturally aspirated trim, it was time for some boost.
9. For our healthy LS3, we selected a single Precision 7675 turbo capable of supporting over 1,100 hp.
10. The turbo headers fed a custom Y-pipe crossover equipped with the T4 turbo flange and a pair of Hyper-Gate45 wastegate mounts.
11. Boost from the Precision turbo was fed through an air-to-water intercooler from ProCharger.
12. We drilled and tapped the Y-pipe to monitor backpressure during the turbo testing.
13. Run with the 317 heads, the turbo 6.2L produced 807 hp at 6,800 rpm and 747 lb-ft of torque at 4,900 rpm. The peak backpressure checked in at 13.8 psi.
14. After installation of the LS3 heads, the peak numbers jumped to 826 hp at 6,700 rpm and 771 lb-ft of torque at 4,600 rpm. The backpressure rose slightly with the more powerful LS3 heads to a peak of 14.2 psi.
15. Given the difference in flow rates offered by the two cylinder heads, it is not surprising that the LS3 head easily outperformed the 317 heads. After all, the cathedral port 317 heads were designed for a pedestrian 6.0L truck application, while the LS3 heads were designed for the performance-oriented Corvette. Given the minor increase in chamber volume of the 317 heads (71.5 vs 70), the gains attributed to the LS3 heads are primarily down to flow, though the chamber design of the LS3 is also better than the 317. Thanks to the cam upgrade, the LS3 produced 551 hp and 507 lb-ft of torque with the 317 heads. After installation of the LS3 heads, the power numbers jumped to 584 hp and 526 lb-ft of torque. Not many people would consider putting 317 heads on their LS3, but the LS3 heads (along with the intake and rockers) are a solid upgrade for any 6.0L LQ4.
16. This is really the dyno test that fans of two respective head camps were waiting for. Run with a single Precision 7675 turbo, ProCharger air-to-water intercooler, and 317 heads (at 7 psi), the 6.2L produced 807 hp at 6,800 rpm and 747 lb-ft of torque at 4,900 rpm. After running the same setup with the LS3 heads, the 6.2L produced 826 hp at 6,700 rpm and 771 lb-ft of torque at 4,600 rpm. The gains offered by the high-flow LS3 heads continued under boost, but there is evidence that the gains diminished slightly. A peek at the boost and backpressure curves might offer some insight.
17. The two sets of curves illustrated here represent the boost pressure and backpressure offered by the two turbo combinations. Both turbo combinations were run with 7-psi wastegate springs and no controller. Equipped with the 317 heads, the combination produced a peak of 7.6 psi, which dropped to 6.9 psi at the top of the rev range. Run with the LS3 heads, the boost curve peaked at 7.4 psi and fell to 6.4 psi at the top of the rev range. The reason for this can likely be traced to the backpressure curve, as the extra power offered by the LS3 heads (in naturally aspirated trim), increased the exhaust flow in the system and elevated the backpressure. The peak backpressure registered during the runs was 13.8 psi for the 317s and 14.2 psi for the LS3s. The higher backpressure worked with the boost pressure to open the wastegate at a slightly lower boost level with the LS3 heads than with the 317s.