8/1/2001
GM 302: Our High Output OBD II Compliant Engine
By Norm Brandes
In my last column (Automotive Rebuilder, May 2001, page 24), I described an OBD II emissions-compliant 302 we built for a Chevrolet project car. This engine, running at 6700 rpm, developed 370 horsepower at the rear wheels through exhaust without converters. This thoroughbred not only shared the same displacements as its predecessor, it also had the same exhaust rumble and could run a 12.0 second quarter time.
This month I’ll tell you how we did it.
First of all, we were in the unique position of being able to work directly with the GM engineers and other aftermarket companies on this project. This opportunity allowed us to try things that are not ordinarily available in building a custom vehicle.
Our goal was to take the perceived 1960s Z28 image and transplant it into a 2000 Camaro using all of the OBD technology. However, because of the rumble we wanted to achieve, we had to eliminate the catalytic converters. This was accomplished by reprogramming the computer which turned off the catalyst efficiency monitoring. However, not being a tree hugging environmentalist, my motivation to keep the OBD II system intact was for the protection of this engine.
Something today’s engine rebuilders have to keep in mind is that OBD II can detect problems that no emissions test or engine analyzer can. OBD II is 10 times more sensitive than an I/M 240 loaded mode emissions test. If a vehicle can be driven without turning on the OBD II light, you know it’s running good and will pass virtually any state-mandated emissions test and achieve the engine durability that is expected in today’s vehicles.
One of the secrets in keeping OBD happy is to achieve the right balance between compression and cam duration. High compression improves idle stability with a performance cam. The only thing you have to watch out for here is that the quality of gasoline varies widely from one area of the country to another. If you can’t get 93 octane premium fuel, you might have to run on 91 octane which increases the risk of detonation.
In coming up with the 302 displacement, we used a stock GM 4.8L truck crankshaft (1999 vintage) in a specially cast aluminum block with undersized bores that GM provided. With a bore of 3.840˝ and a stroke of 3.268˝, we were right at the 302 displacement. +300° Below of Chicago cryogenically treated the crank along with the block and other engine components to stabilize and strengthen the parts.
We used custom-made JE forged pistons with an 11.5:1 compression ratio. JE digitized the shape of the LS6 cylinder head combustion chambers, then machined the piston crown to match the new contour. The final machining of the valve pockets in the pistons was done by Westech, after checking the needed clearance for the valves.
Due to the rpm we needed to run because of the small displacement, we decided to use Dyer Lite Weight steel connecting rods. Piston protrusion is critical and we found these connecting rods to measure well below .0005˝ center-to-center variation. The throws on the crankshaft were also carefully measured and found to be below .001˝ in stroke variation.
With all of these components being held to tight tolerances, we knew that the rotating assembly would have less than .002˝ protrusion height variation. Our experience with other LS1 engines has shown us that variations in protrusion heights greater than .008˝ - .010˝ will set a random misfire (P0300) code. No matter what we did, we couldn’t extinguish the MIL light. However, when we reduced the protrusion variation to less than .002˝ the problem was cured. That’s how we arrived at this spec.
The block was machined to ensure identical deck heights on both sides, and to minimize any variations in flatness, cylinder taper, concentricity, straightness, etc. The size of the main bearing openings was also optimized for bearing clearance. A change in main bearings, Federal-Mogul’s H-14 Competition Series Alloy (#152M1), gave us our decrease in vertical oil clearance. This was needed due to the use of an aluminum block and the growth it undergoes as the temperature changes. We essentially built the bottom end of this block as if it were going into a NASCAR application.
The LS6 cylinder heads that were used on this motor were flowed and lightly ported by Advanced Air Flow. There’s a boss protrusion in the intake runners for a rocker arm stud that creates a slight restriction to airflow.
Grinding the boss flush improves flow, but requires using sealer on the rocker stud threads to prevent air leaks. We left the flow diverter in the guide area alone because this helps provide swirl and tumble into the "fast burn" combustion chamber. The combustion chambers were contoured and matched to 64 cc each. Matching combustion chambers is critical as far as OBD II is concerned, so we try to keep all the chambers within 1/2 cc of each other.
For the camshaft, we installed a Competition Cams mechanical roller with 234/240 duration, 110 lobe separation with .629˝ gross lift. This cam uses mechanical roller lifters, which we felt were needed for the high rpm potential of this engine. The fast opening ramps on this cam really boosted the mid-range power of this engine. The cam was ground with 4 degrees of advance built-in and installed straight up with zero offset.
The heads were fitted with special 1.900 aluminum rocker arms made by Jesel. The high lift rockers required adding a one-inch spacer between the valve cover and head to provide extra clearance for the rockers. Some grinding also had to be done in the head to increase clearance for the rockers and pushrods.
To route the exhaust gases out of the engine, we chose TTS Power Systems coated headers. These headers have a heavy gauge flange that resists warping and provides a leak-free long lasting seal with the exhaust ports. Because of the ground finish on the header flange, no exhaust manifold gaskets were used, only a little high temperature RTV around the ports. Even a small exhaust leak that you can’t hear can make the O2 sensor run richer.
The design of the collector at the end of an exhaust header, as well as the placement of the oxygen sensor, is very important. If the collector isn’t long enough, or the oxygen sensor is located in the wrong spot, it may not get an average reading of all four cylinders. This may cause the O2 sensor to read rich or lean depending on which cylinder(s) has the most influence on its readings.
The intake manifold we chose for this engine is from a 2001 Corvette Z06 and features better runners and a larger plenum. This manifold is good for an extra 20 to 25 horsepower over the earlier manifold. There is no EGR valve, thus we needed the 2001 program code in the PCM. GM uses valve overlap to control NOX on these engines.
When the engine was assembled, we used the new revised head bolt torquing procedure that GM came out with in 1999. The new procedure is to tighten all M11 bolts in sequence to 22 ft.lbs (30 Nm) on the first pass, give them all a 90 degree twist on the second pass, and an additional 90 degree twist on the final pass (except for the medium length bolts at the front and rear of each cylinder head).
Because of the headers, some of the engine’s wiring had to be rerouted to keep it away from the heat. It’s also important to make sure you don’t have any spark plug wires near any sensor wires because this may cause erratic sensor readings.
Once the engine was installed in the car, we had TTS Power Systems in Compton, CA, reprogram the PCM for optimum performance. This included correcting the fuel injector base pulse width for proper idle and fuel mixture, and increasing the fuel flow at wide open throttle for maximum power. Ignition timing is a maximum of 30 to 32 degrees of advance.
We were very pleased with how the 302 project turned out because it proved you can build an OBD II compliant motor without having to sacrifice power or performance.