by Hib Halverson
Copyright 1997 Shark Communications
No Use Without Permission
The return of performance to production Chevrolet in the late-80s and early-90s was driven by two key events: continued development of computer controls and roller hydraulic lifter camshafts.
Roller lifters or "Rollerized valve tappets" first saw widespread use in World War II military aircraft engines. In the late-1950s, hot rodders discovered the idea and, within ten years, roller lifter camshafts or just "rollers" were an accepted ticket to maximum horsepower from engines with pushrod-operated valves.
In the early-'80s, enamored with the roller lifter's reduced friction and ability to work with higher valve lifts for a given duration, General Motors developed them for production applications. Eleven years ago, Chevrolet introduced hydraulic roller lifters in the small-block V8s used in Corvettes and Camaros.
Aftermarket camshaft manufacturers soon hopped on the bandwagon with roller hydraulic lifters for engines not originally equipped with them. Since then, a popular upgrade is switching from a flat-tappet hydraulic cam to a roller hydraulic. We decided that an interesting exercise would be to do this swap, but make the cam change the only modification affecting performance.
Our test case was a '65 Chevy Malibu powered by a 400 cubic-inch, small-block V8. It has a two-bolt-main block bored .020-in. over and a production crankshaft that was Magnafluxed and deburred. Stock rods were Mag'ed, deburred, had their beams polished and were fitted with B&B Performance rod bolts (p/n 38630). Pistons are Bill Miller Engineering forged, 9.5:1 flat-tops with floating pins. Reciprocating parts were balanced by Evans Speed Equipment in South El Monte, California.
The cylinder heads are cast iron, L82 Corvette units modified by Mark DeGroff's Cylinder Head Service. The original camshaft was a flat-tappet hydraulic working rapid bleed-down lifters and production Chevrolet valve gear. Exhaust is a set of Hooker Super Competition adjustable race headers fitted with Flowmaster collectors and blowing through three-inch pipes and Flowmaster mufflers. The induction is a rare, Rochester, mechanical port fuel injection system from a 1964 Corvette.
We did this job like most readers would do it: in a small garage with the engine in the car. The first step was to tear the motor down enough that we could get the camshaft out. In many "flat-to-roll" conversions, this means pulling the heads because the valve spring seats will probably have to be enlarged.
Space is limited, here, so we're skipping disassembly basics. If you need help with that, consult a service manual. Year One, Inc. sells Chevrolet Chassis Service Manuals for '65-'72 Chevelles and other, older platforms. Helm, Inc. has factory manuals for newer models. In some cases, less expensive, non-factory manuals for '68-up cars are available from Haynes Publications. If you want a book that covers just engine assembly and disassembly, Motorbooks International has several choices. Lastly, consider David Vizard's book How to Build Chevrolet Small-Block Camshafts and Valve Trains, published by Motorbooks. It's the best work on the subject in the last 15 years and a must-read resource for anyone wanting to understand the high-performance, SB Chevy valve train.
How about some Chevelle-specific tricks not in books? First, you can't remove the oil pan with the engine in the car because the front crossmember is in the way. However, except for that rare engine equipped with a two-piece timing chain cover, you must drop the pan to remove the cover and gain access to the cam. The solution?
Once the engine is stripped down, raise it as high as it will go then support it with wood blocks (4"x2.5"x1.5") placed between the motor mounts and the motor mount brackets on the frame. Be careful to keep you fingers out of the space between the mounts and frame while placing the blocks. If the jack fails, your fingers will be crushed.
Depending on what transmission the car has, sometimes you can get the oil pan off this way. At minimum, you can drop it enough to get the timing cover off, clean the pan rails and replace the gaskets.
Getting the cam the first few inches out of the block is difficult because of inability to exert enough leverage to guide its lobes over the cam bearings. To make that process easier, a pair of 5/16, coarse-thread bolts, 4-6 in. long, can be screwed into the front end of the cam to serve has "handles".
You may get the cam almost all the way out, then find it interferes with the hood release mechanism. Position the engine such that the point of interference is the very bottom of the release, then simply pull the lever back and the cam will come out.
The step was to clean parts and repaint. We used Eastwood Chassis Black (p/n 1244Z) for accessory brackets and Eastwood Chevrolet Orange (1232Z) for the harmonic damper, valve covers, water pump and timing chain cover.
Crane Cams is an innovator in roller lifter camshaft technology. It was first in the aftermarket with roller hydraulic cams. Though Crane did not originate the idea, in conjunction with its subsidiary and OE GM supplier, Camshaft Machine; it developed high-volume production of steel billet, roller cams with interference-fit, cast iron distributor gears. This advancement allowed roller hydraulic cams to meet OE durability goals paving the way for their use in millions of production Chevrolet engines.
The roller hydraulic lifter camshaft we selected was a Crane HR276 (p/n 119621) that has the cast iron gear. It's a dual-pattern cam and, if used with 1.5:1-ratio rocker arms, has intake and exhaust lifts of .488-in. and .509-in., respectively. Its durations are 276°/284° (advertised) and 214°/222° (at .050 lift). It has no overlap at .050-in. lobe lift and is ground with its intake centerline at 107° and a 112° lobe centerline angle which means it has 5° advance built into the profile.
We wiped Red Line Synthetic Oil's Engine Assembly lube on the cam's bearing journals and distributor gear, then inserted it part way in the block, installed our "handles" and, carefully, finished the job.
We added a Crane, a double-row, true-rolling timing chain assembly (p/n 11993). Its crankshaft sprocket has three keyways allowing installation "straight up" (which we used, since the cam is ground with 5° advance already) or with either: four degrees advance or four degrees retard. We installed the camshaft sprocket and lined its index mark with crankshaft sprocket mark. We left the cam sprocket bolts only "snug." The final tightening would be done later.
The next step was to "degree" the cam. This validates both the cam manufacturer's work and your installation. Briefly, you establish top dead center using the "positive stop" method then compare the cam's actual timing with the manufacturer's timing data. A "degree wheel" is used to measure crankshaft rotation and a dial indicator is used to measure lifter or valve movement. We recommend degreeing a new cam in any street high-performance or racing application. A full discussion of the process can be found in How to Build Chevrolet Small-Block Camshafts and Valve Trains.
Crane sells a handy package of measuring equipment, the "Tune-a-Cam Kit." (p/n 99030), that has everything you need to degree a camshaft with the heads on the engine. However, we degreed with the heads off, so we needed B&B Performance's TDC Indicator Stop (p/n 40200). We also used B&B's 14-inch Degree Wheel and Chevrolet mounting kit (p/ns 45050, 45010). TDC checking requires changes in degree wheel indexing and we prefer the B&B set-up because it makes indexing changes easier. Lastly, to mount the dial indicator to the head decks, we used a magnetic base one can find at any tool supply.
During the degreeing process, we checked lobes for #1 and #6 cylinders and found cam timing accurate within 6/10ths of one percent, well under accepted standards. Clearly, Crane makes a quality product.
We removed the camshaft bolts to install a Crane cam button spacer (p/n 99164) and locking plate assembly (p/n 99168). Both items must be used with a roller to control endplay. The camshaft bolts were treated with Valco Cincinnati Thread Locking Compound then torqued to 20 ft/lbs.
We installed a Fel-Pro timing cover gasket and test fitted the cover for a check of camshaft endplay. Unlike a flat tappet cam, a roller "floats" in the block so endplay is an issue. It's controlled by the cam plug at the back of the block and by the cam button spacer and the timing cover at the front.
A reasonably accurate endplay measurement is possible using a dial indicator that has a mounting ear on its back. Aim the indicator stem through the engine block's right-front oil return hole such that it touches the back of the camshaft sprocket. Position the indicator body such that the mounting ear drops into the #2 exhaust lifter bore. Gently pry the camshaft back and forth and watch the indicator dial. The measurement will be a teensy bit off due to the indicator axis diverging slightly from the camshaft's, but your reading will be of acceptable accuracy for endplay purposes.
In some cases with sheet metal timing covers, there will be no endplay, so it will have to be modified for clearance using two sockets and a large bench-vise or a hydraulic press. Put machinists blue on the cam button, then push the cover in place to mark the area to be formed. Use a 1-1/4-in. socket on the outside and a 9/16-in. socket on the inside. Position the cover between the two sockets and sandwich all that in the vise. Compress the vise and the 9/16ths socket forms a depression on the inside of the cover. Increase the size of the depression a little at a time until end-play falls within the desired range. We modified our cover such that endplay was the .005-.008-in. suggested by Crane Cams.
Lastly, we cleaned the oil pan gasket surfaces and end seal grooves. To keep debris from falling into the pan, we followed the gasket scraper with a shop vacuum. We put a set of Fel-Pro Performance oil pan gaskets (p/n 1821) in place, reinstalled the pan and lowered the engine back into its mounts. We reinstalled the fuel pump using a Crane, bronze-tipped, fuel pump pushrod (p/n 11985) required for use with roller cams.
If you've seen people install harmonic dampers with a 10-lb. sludge; don't go there. You'll damage the crankshaft thrust bearing and probably the damper itself. Use a B&B Performance Damper Installation Tool (p/n 40660). It has a threaded pilot, a big nut and a ball-bearing to smoothly press the damper onto the crank snout.
Using B&B's degree wheel and TDC stop, we set the engine at TDC #1, then compared the timing pointer to the timing mark on the damper. The match-up was perfect. We torqued a B&B damper bolt (p/n 26300) to 60ft/lbs. and added an MSD timing tape (p/n 8985), then turned our attention to the cylinder heads and valve gear.
We returned the heads to Mark DeGroff's Cylinder Head Service for a freshen-up and to have the valve springs changed. We originally chose L82 Corvette heads for the large combustion chambers needed to get 9.5:1 compression in a 400 with flat-tops. Additionally, of the production, big-chamber, iron heads; they have the best intake ports. The first time around, DeGroff ported and polished the castings, added larger, 2.02 in. intake valves and did a multi-angle valve job. For the freshen-up, Mark disassembled the heads, inspected all parts and grit-blasted the chambers. He then "touched-up" the valve faces and seats and verified that valve stem-to-guide clearance was within tolerance.
Compared to flat tappet cams, roller profiles have higher acceleration rates and higher lifts so more aggressive valve springs must be installed. Invariably this means a dual spring of larger diameter. Our stock spring seats had to be enlarged to accept them, so Mark DeGroff machined the heads accordingly.
Valve stem lubrication is a tricky issue with a street high-performance engine. Liberal valve stem oiling is ok for a race motor, but on a streeter, it can cause excessive oil consumption and even visible smoke. Too little oil will fry the stems and guides. Originally, Mark DeGroff set stem-to-guide clearance at .0015-in. for intakes and .0025-in. for exhausts. For sealing, he used OE valve stem o-rings and a set of late-model, OE rubber seals installed on the ends of the valve guides. After 30,000 miles, stems and guides showed minimal wear and there was no evidence of excessive oil use so, clearly, valve stem lubrication was ideal. Since the Crane retainers allow production o-ring seals to be used; Mark duplicated our original seal set-up. We feel that DeGroff small-block heads, set up like ours, will have 50,000 mile durability…outstanding life in a street high-performance application.
The Chevrolet valve springs we used with the flat tappet cam had greater installed height than the Cranes. Also, the OE exhaust valve rotators, typical of small-block heads of the mid-to-late-70s, had been removed, resulting in additional, surplus installed height for the exhaust valves.
To address the intake valve installed height issue, we installed the Crane spring and retainer (p/ns 99838, 99936) recommended for use with the HR276 but used a set of special, Crane split-locks (p/n 99096) that reduced installed height by .050-in. For the exhaust valves, we used a taller Crane spring (p/n 99893) but with the standard retainers and locks (p/ns 99936, 99097) that go with the cam.
If you are converting an iron head originally equipped with rotators to Crane dual springs and no rotators, consult Crane's technical support department before you order valve springs, retainers and locks.
Mark DeGroff checked all springs. The measured: 100 lbs., closed and 300 lbs. at .500-in lift for the intakes and 120 lbs. closed and 300 lbs at .500-in. lift for the exhausts, figures appropriate for the camshaft profile we are using and the rpm range of our 400. That all were consistent in both seat and open pressures is demonstrative of Crane's good quality control. That and choices in springs, retainers and locks allowed DeGroff to assemble the heads without valve spring shims.
Back in our garage, we shot a fresh coat of Eastwood Chevrolet Orange paint on the heads then reinstalled them, using Fel-Pro Performance head gaskets (p/n 1014) which combine Fel-Pro's, excellent stainless steel core/laminate design with a steel ring around the cylinder for improved sealing in high-compression engines. Though its intended use is marine and racing applications, this gasket has the steam holes necessary for the 400 small-block in a street application.
We used B&B Performance head bolts (p/n 3810). SB Chevy head bolts screw through the decks and into the water jackets and, thus, their threads should be coated with a sealer. We use Loctite PST which offers good sealing and adequate seize protection. The B&B bolts use head bolt washers and they should be lub'ed before installation. We used Red Line Assembly Lube for that. We tightened the head bolts in the pattern listed in the service manual to 70 ft/lbs. for the long bolts and 65 ft/lbs. for the short ones.
We dunked each pair of Crane lifters in Red Line Synthetic 10W-30 then dropped them in place. Crane pushrods (p/n 11628) specific to this application went in next.
We added Fel-Pro intake manifold gaskets (p/n 1205) intended for stock or slightly modified intake ports. They have Fel-Pro's clever, Printoseal feature that allows them to be reused during induction system changes. We laid a 1/4-in. bead of Valco RTV silicone sealer on the end-seal surfaces, waited 20-min. for it to "skin," then set the manifold in place. We opened a set of B&B intake manifold bolts (p/n 26050), dabbed Red Line Assembly Lube on the washers and PST on the threads then torqued them to 30 ft/lbs.
While researching this article, we found comparative data on rocker arm ratios in How to Build Chevrolet Small-Block Camshafts and Valve Trains indicating that, in many street high-performance applications, additional power with minimal loss of low-end torque comes with 1.6:1 rocker arms on the intakes only.
We installed Crane. 1.6:1, extruded aluminum rockers (p/n 10759) on the intakes which gave us .520-in. valve lift. We used Crane one-fives (p/n 10750) on the exhausts. Both these rockers were designed for '86 and up heads but can be used on older heads. Their advantage is they fit underneath the old-style, production, sheet metal or cast aluminum valve covers that must be used on engines equipped with Rochester fuel injection.
The initial valve adjustment was 1-turn of preload. This was accomplished by adjusting each intake when its companion exhaust started to open and adjusting each exhaust when its intake started to close. Each rocker nut was hand tightened until zero lash was reached. A good way to feel for this is to pull up on the pushrod end of the rocker arm while twirling the pushrod. When you feel a tiny resistance to your twirl, that is the zero lash point. Each nut was given a turn more, then the locking screw was tightened.
The valve covers were reinstalled using Fel-Pro gaskets (p/n 1604) and B&B hold downs (p/n 68500). Information on the remaining reassembly can be found in a service manual.
Since we had the exhaust off the car, we improved the system's reliability and appearance. First, we had our Hooker Super Competition headers (p/n 2375) coated with Hooker's optional Metallic-Ceramic treatment. This coating tolerates thermal stress, resists rust, is somewhat impact resistant and has an attractive, polished silver finish. Hooker Industries offers this coating on new headers or as a stand-alone service. The headers on this car had been treated twice before by other suppliers with metallic-ceramic coatings that failed in short order. Hooker's process is the only one that has lasted.
For a durable seal, we use Fel-Pro header gaskets (p/n 1407), specific to flange-plate-type Hooker headers. Lastly, we use the Stage 8, locking header bolt system which never loosen up.
Upon reinstallation of the distributor, we set the initial timing at 15 degrees. This ensures and easy first start-up, even if the engine is cold. We installed a MSD distributor cap (p/n 8437), one of MSD's new, racing rotors (p/n 8467) and a set of MSD Super Conductor spark plug wires (p/n 31659). Then, we screwed in a fresh set of AC RapidFire #1 spark plugs, filled the crankcase with Red Line 10W-30 Synthetic oil, installed a K&N Oil Filter (p/n 07-0025), and filled the cooling system with 75% distilled water, 20% antifreeze and a bottle of Red Line Water Wetter. The last ingredient contains a "wetting agent" that enhances transfer of heat from the engine block and head castings to the coolant. Because of this, once the thermostat fully opens, coolant temperature stabilizes at a lower level.
The engine fired immediately and we adjusted the fast idle to 2000 rpm. Then, we backed off the timing to about 20.° Once the engine reached normal coolant temperature, we ran it up to 3500 rpm and set the total ignition advance to 34°. We continued to run the engine at a fast idle until the oil temperature reached 175°. Then we shut it down and checked for leaks.
Now it was time for some development driving. The significant change in airflow characteristics brought by the new camshaft forced recalibration of the engine's fuel delivery schedule. We equipped the Malibu with an Air/Fuel Ratio indicator (p/n 85-2437) from K&N Engineering. This devices uses data from an oxygen sensor installed in the exhaust to derive air/fuel ratio and displays it on a small, LED panel we mounted in the car's dashboard. We used it to set the fuel injection's part-throttle-cruise calibration and the point at which its power enrichment occurs. With a carburetor engine, these operations correspond to changing the main jets and/or power valve.
We also used the K&N A/F Indicator to make a preliminary wide-open-throttle fuel mixture adjustment, however, we do not recommend O2S-driven devices as the final measure of WOT fuel calibration. It is better to use vehicle performance testing or a precise exhaust gas analyzer for this purpose.
With the engine's fuel curve in the ballpark, we put some more miles on the car getting subjective impressions. The first thing we noticed was a smoother idle, slightly lower idle speed and about 2-inches more vacuum. Drive ability at low speeds and part throttle also improved. At highway speeds, we saw more vacuum and about 1.5 more miles-per-gallon. All of these are signs of an engine that is running more efficiently.
After about a thousand miles, we got the car back in the shop checked for leaks again, pulled the valve covers and readjusted the lifters, then, replaced the oil filter with a fresh K&N and added enough Red Line 10W-30 top bring the level back up to "full".
Obviously, the last step was to validate our modifications and the results of our testing are covered in detail elsewhere (see attachment). Suffice to say that, clearly, when done to a street high-performance Chevrolet V8 having appropriate induction and exhaust system modifications, replacing a flat tappet, hydraulic cam with a a roller hydraulic of similar or slightly less duration pays off in better performance.
|AC Rapid Fire
see your ACDelco dealer
B&B Performance Sales
PO Box 1329
Riverside, CA 92502
Mark DeGroff's Cylinder Head Service
NGK Spark Plugs, USA
Red Line Synthetic Oil Corp.
Stage 8 Fasteners
Year One, Inc.
The performance increase that came from switching from a flat tappet hydraulic to a roller hydraulic camshaft was significant.
Before we tore the engine apart, we took the car to K&N Engineering in Riverside, California. At K&N's research and development facility is a Dynojet chassis dynamometer that is available to the media on a limited basis for testing purposes.
Prior to a baseline test of our Malibu, we installed a set of NGK R5674-8 V-Power racing spark plugs (about two heat ranges colder than the Rapid Fire 1s we run on the street) and added a K&N Air Filter. We made three runs on the Dynojet that averaged 295hp@4960 rpm and 347lbs/ft. torque at 2660 rpm. If we apply the conventionally-accepted rule-of-thumb that says rear wheel horsepower and torque of a V8, rear-drive car are approximately 80% of the gross flywheel numbers, the Malibu's 400 small-block was good for about 370hp and 435 lbs/ft. torque.
Once the flat-to-roll conversion was made and the car was driven about 1000 miles, we took it back to K&N for another session on the Dynojet. The NGK V-Power racing plugs went back in.
The first thing we did was a final check of wide-open-throttle fuel mixture. Actron Manufacturing, which makes Sunpro automotive diagnostic equipment for consumer use, also markets the KAL Equip line to the service trade. Actron's Bruce Heath was kind enough to loan us a KAL Equip model 5000, exhaust gas analyzer for our tests. This piece of equipment's accuracy is more than acceptable for setting WOT fuel calibration.
Our first runs on the Dynojet with the new cam attempted to dial-in the fuel mix at WOT. On the first pass, the KAL 5000 showed A/F to be an ideal, 12.5:1 at high rpm. On the next two passes, we leaned the mixture somewhat, then set it a bit rich. While the KAL 5000 clearly indicated the changes, the power and torque numbers the Dynojet gave us were almost identical proving that our initial adjustment was the best choice so we reset WOT fuel mixture to that.
Then, this time for the record, we made three more passes.
The results were impressive: an average of 339hp@4840rpm and 396lbs/ft. torque at 3800 rpm, a whooping 44hp and 51lbs/ft. more at the rear wheels. Using the "80% rule", estimated gross power and torque at the flywheel have risen to approximately 425hp and 500lbs/ft…not bad for a low compression, iron head 400.
It is important to note that this kind of performance increase from a flat-to-roll conversion is not specific to just engines like ours. Most any small-block Chevrolet of 350-420 cubic inches is going to see a similar improvement provided the induction, heads and exhaust are properly configured.
The best passes from both the baseline and the final tests are listed below.
It was interesting to note that peak torque both increased and moved up in the rpm band. Of course, we expected the increase but the change in where that peak occurred had us wondering at first.
We feel the reason the torque peak changed position is that the engine is running much more efficiently in the 3000-4000 rpm range. When we looked more closely at the data we got from our sessions on Dynojet, we noticed that in both tests, the 400's torque curve actually had two peaks. In the baseline test, torque peaked at 2600 rpm then again at 4000 rpm. The numbers were only 2.6 lbs/ft. apart, with the first peak being larger. In the second test, torque peaked at once at 2700 rpm and again at 3800 rpm and those peaks were 15.9 lbs/ft. apart, but this time, the second was higher.
Our conclusion was that the Crane roller hydraulic cam had the most effect between 3500-5000 rpm. Because of that, in our second round of tests; it was the second peak at 3800 rpm that became the larger of the two.
We also noticed that, 1) the power peak moved down just slightly and 2) while horsepower between 5000 and 5500 rpm increased, the rate at which it fell off above 5000 was more rapid. These attributes are no fault of the camshaft, but indicative that the induction system, specifically the Rochester injection unit, is now airflow-restricted above 5000 rpm. However, this problem is of little consequence since 1) power between 5000 and 5500 increased, 2) the engine is shifted at 5500 and 3) because it uses a stock crank and rods, its MSD ignition is set with a rev limit of 5800 rpm.
The payoff with a roller cam replacing a flat tappet grind of the same or slightly greater duration is that you can sometimes increase mid-range torque, improve drivability and have more horsepower. While this may seem like having the proverbial cake and eating it, too, it can happen because, for a given duration, a roller profile can have less duration, more lift and get the valve open more quickly than can a flat tappet cam.
The valve timing chart shows the intake valve events and overlap of both the flat tappet cam we took out and the roller we installed.
|Flat Tappet||Roller w 1.5||% Change||Roller w 1.6||% Change|
|intake lift||.468-in.||.488-in.||+ 4.3||.520-in.||+ 11|
at .050-in. lobe lift
|lobe centerline angle||110°||112°||+ 2||112°||+ 2|
|int. centerline, as installed||106°||107°||+ 1||107°||+ 1|
|advertised duration||275°(.006-in. lift)||276°(.004-in lift)||n/a||276°||n/a|
|duration at .050 lobe lift||219°||214°||- 2.3||214°||- 2.3|
|overlap at .050 lobe lift||4°||none||- 100||none||- 100|
above .050 lobe lift
|25.735-in/deg.||27.493-in/deg.||+ 6||29.326 in/deg.||+ 14|
above .050 lobe lift
|.007-in/deg.||0||- 100||0||- 100|
|lift at peak piston velocity||.259-in.||.273-in.||+ 5||.291-in.||+ 12|
|lift at TDC||.085-in.||.075-in.||- 13||.080-in.||- 6|
Note that the intake lift and the valve event area numbers take the rocker arm ratio into account and are figured at the valve while the other data is measured at the cam lobe.
The key conclusions one can draw from this chart are that, 1) though the roller cam's duration at .050-in lift is a bit less, because of higher lift and quicker valve acceleration, the overall valve event (quantified by the "intake area above .050 lift" and called "power potential" by some camshaft company engineers) is greater and allows more air flow into the engine. All that additional airflow really comes into play between 3000 and 5000 rpm.