A flat lifter is, in reality, not flat on its working face, and would quickly fail if it was. The lifter's face is actually crowned with a 60- to 100-inch radius. The cam lobe itself is tapered. This, in conjunction with an offset of the lifter over the cam lobe, causes the lifter to rotate. This means it is not a simple rubbing action between lifter and lobe, but a semirolling motion. Without the correct rotation, the lifter will have a short life. The contact patch between the cam form and lifter is one of the highest-stressed areas in an engine. If heavy springs are going to be installed, it pays to use armor-faced lifters such as those sold by COMP Cams. Flat lifters are inexpensive and they can get the valve off the seat really fast. What they don't do quite as well as a roller is build lifter velocity or deal with super heavy springs. The amount of velocity that can be imparted to a flat lifter is dependant on the lifter diameter. The wider it is, the faster it travels. Because a flat lifter can initially accelerate faster than a roller, we find that with cams under about 270 to 278 degrees of off-the-seat duration, a flat tappet can produce as much or more area under the curve. It also used to be claimed that a flat lifter design was substantially inferior to a roller lifter in the friction department. Though this was, to an extent, correct 20 years ago, we find that new super oils have eaten substantially into whatever friction reduction advantage the roller may have had over a flat lifter.
If maximizing output is the goal and the budget extends to the higher price of a roller lifter cam, then, for anything over about 275 degrees, the roller is potentially a superior power producer. Where it scores is the velocity it can impart to the lifter is higher than a flat lifter. If it has enough duration to get up to speed, then the lifter can be pushed higher. As we have already seen, valve lift with a two-valve engine is an important factor, and a roller can deliver in this area.
Once a flat or roller lifter decision has been made, the next question is, solid or hydraulic? Of the two, hydraulic is the most popular for street use as it is quiet and virtually service-free. The hydraulic innards adjust the working length of the lifter such that it takes out all the lash and no more.
You may have heard the term "anti-pump-up lifters." These are intended to fix a problem that can occur toward the top 25 percent of the engine's rpm range. What happens is that the spring starts to lose full control of the valvetrain and separation between various components takes place. This, as far as the lifter is concerned, looks like lash, so the lifter does it's job and takes it up. When the valve now tries to close, the lifter, which is now a little too long, holds the valve off its seat and heavy-duty power loss takes place. For many years, the accepted fix for this was an anti-pump-up lifter, which was a much leakier, faster-collapsing lifter that allowed the valve to physically close unimpeded. But it also collapses easier and consequently cuts valve lift. The real fix is a spring with better control. That's a serious topic we will talk about later.If valvetrain noise is of little or no concern, then a solid cam is the way to go, as there are no worries about hydraulic lifter collapse.
Before moving on to pushrods, some vital info about flat-lifter cams could save you a lot of hassle. Because the surface loading is so high, flat lifters require a break-in procedure that needs to be carried out conscientiously. Never re-use lifters; always use new ones. When installing, generously lube the cam profiles and the lifter faces with the special cam lube supplied with the cam. Make sure the lifters rotate freely.
On start-up, idle the engine at 2,000 rpm for 20 minutes. Do not allow the engine to idle slowly until initial break-in is completed. When very aggressive race profiles are used, the break-in should be done on lighter springs or with low-lift break-in rockers.
Pushrods
A pushrod looks like a simple item. Just make it stiff and light and you are home dry, or at least it would seem so. Sure, low weight and stiffness are key ingredients, but there is another important factor related to pushrod manufacture. Its ability to suppress vibrations. If it can do well here, it acts as a simple damper between the lifter and rocker, rather than a spring. The modern aftermarket pushrod has countless hours of spin-testing to optimize the balance between low mass, stiffness and damping. Doing so ensures a more optimal valve motion. This ultimately translates into additional output. Because a pushrod can influence the valve's motion from as low as the middle of the speed range, additional output can be seen from there on up if it is canceling spurious motion at the valve. Since it is difficult to know for sure that your valve springs are dynamically perfectly behaved, it pays to use a good aftermarket pushrod as a little insurance. But there is more to buying pushrods than just shelling out for a good brand.
When changes to the valvetrain are made and maybe the block and heads have been milled, we can find that the pushrod/rocker/valve stem geometry with a stock length pushrod is anything but correct. The technique here is to use an adjustable pushrod and adjust its length until the roller or rocker tip sweeps out a patch that is centralized on the valve stem. Also, with high lobe lift cams, you should check that the ball end of the pushrod is not moving through an arc larger than the ball will accommodate. If it does, than check your cam supplier's catalog as there are pushrods with tips to allow greater angularity.
Rockers
Rockers not only operate the valves, but they are also extremely useful as a valvetrain tuning aid. Most of the V-8s we deal with have stamped-steel rockers. These are cheap, but if used for lifting the valves much more than about 0.500 (500 thousandths) will cause accelerated guide wear. If the cam is a short, flat-lifter design for a small-block Chevy, you may, if the budget is tight, want to consider Crane's 1.6:1 ratio items. The stock rocker is supposed to be 1.5:1, but rarely comes higher than about 1.45:1. If the budget is a little above basic replacement level, then both Crane and COMP have some roller-tipped rockers that still utilize the stud-mounted ball pivot.
For a little more money, you can get a fully rollerized aluminum race-style rocker. Other than being able to get them for most applications in a variety of ratios, these rockers have the advantage of lower friction and the ability to accommodate ultra-high valve lift.
For the most part, aluminum makes a good rocker material because it has an inherent internal damping capability. On the debit side is that aluminum fatigues, so when heavy race-type springs are used, the rockers will need to be replaced at regular intervals. Stainless steel rockers such as those produced by Comp Cams and Crower have an almost infinite life.
Another option if you're planning on an up-scale engine is to convert from stud-mounted rockers to shaft-mounted ones. These are not cheap, but they do provide the best in valvetrain control.
If you have a better understanding of rocker geometry, you will almost certainly have a power advantage over a less-informed racer. The principle point of a rocker is that it steps up the motion delivered to the valve by the lifter. This is known as the "rocker ratio" and is usually the ratio of the radius of the valve side of the rocker divided by the radius of the pushrod side. If the centerline of the points A, B and C as per the drawing on p. 76 all fall on a straight line, the rocker ratio will remain constant throughout the rise and fall of the lifter. This is OK, but not necessarily the best situation for maximum output. In any undervalved engine, and that's pretty much all two-valve engines, greater output can be had by lifting the intake valves as fast as possible off the seat. The faster the intake is lifted, the less cam duration is required to make peak power. A shorter cam with a faster-opening rocker, as we discussed earlier, will make more torque while still making the peak power of a longer-duration, lower-lifting valvetrain.
A question often asked is how high a rocker ratio is best. If you consider that the purpose of a valvetrain is to move valves, not pushrods and lifters, then it becomes apparent that the higher the ratio, the better within the mechanical constraints imposed by the materials involved. Current maxed-out valvetrains are lifting the valves to about an inch and running to about 10,000 rpm. All this is being done with rockers in the 1.9 to 2.1:1 range.
Stepping up the rocker ratio is often a good way to increase output with no more than a simple bolt-on mod. Higher-ratio rockers can spread the engine's required LCA. This means that if the existing cam has too wide an LCA, as is so often the case, bolting on a set of high-ratio rockers can pay a handsome dividend. On the other hand, if the LCA was such that the overlap triangle was optimum, installing a higher-ratio set of rockers can drop output rather than increase it. My own tests have indicated, within the ratio range of 1.5 to about 1.9:1, that for every 0.1 ratio increase on the intake, the LCA needs to be spread by 0.75 to 1 degree.
As for the exhaust, we find that it is relatively insensitive to valve acceleration but is sensitive to duration. For this reason the rocker ratio used on the exhaust is best kept about 0.1 to 0.2 of a ratio lower than the intake ratio.
As for power increases as a bolt-on deal, for something like a small-block Chevy or Ford, experience shows you can reasonably expect to see results similar to or better than shown in the graph below.
Rocker Geometry
A rocker's ratio is usually defined as R1 divided by R2 but, because of the geometric relationship between points A, B and C, the ratio changes as the rocker progresses through its lift range. If pickup point "C" is moved in the direction of the ghosted arrow, the ratio of the rocker, as it lifts the valve off its seat, will be higher. So long as no dynamic problems are encountered, this is good for added output, especially with short-cammed engines. | 
Overlap Triangle--1.5 Rocker Ratio Versus 1.8
The graph shows how the overlap triangle increases with rocker ratio increase. If the overlap was optimal on the lower ratio rockers, the cam LCA must be spread to restore the status when higher ratio rockers are used. | 
COMP Cams Magnum 1.6 Rockers Versus Stock 1.5
If the cam's LCA is a shade too wide for the application (as it is in most cases), a set of higher ratio rockers can really deliver the goods, especially from the mid-range up. The cam in this instance was a 272/278 flat hydraulic on a 110 LCA. The Magnums delivered an instant 14 hp with no significant low-end loss. |