Valve Float
Before we go any further, it's important to define what we mean by loss of valve control. This is most often referred to as valve float because the common misconception is that the lifter launches itself off the nose of the cam. While this can and does occur, the more common result of a loss of valve control is when the valve bounces as it approaches the seat. Generally, this will happen to the intake valve first because of its greater weight. This results in a loss of engine power because cylinder pressure is pushed back into the intake manifold instead of remaining in the cylinder. The biggest problem is identifying when the power loss occurs, because it may happen a few hundred rpm earlier than what is considered normal for a given application.
Classic valve float is usually accompanied by a dramatic loss in power and an obvious misfire, but most engines are already suffering power loss from valve float before it becomes audibly noticeable. This is what happened to our big-block and why our engine picked up significant power at the top end of the curve.
Don't Freq Out
Most springs are rated for a certain load at a given installed height. That load is the pressure imparted on the valve to control it. Historically, load has been the main factor in matching a spring to a valvetrain, but given all the variables of rocker ratios, valvetrain weight, pushrod deflection, and a couple of dozen other items that affect valve operation, it's easy to see that matching a spring to a cam lobe and valvetrain requires much more data than just a load rating. Much of this has to do with what is called a spring's natural frequency.
At a given rpm, any valvespring will hit a frequency where it will naturally vibrate or resonate like a tuning fork. When this happens, the spring loses much of its ability to maintain control over the valve. This is why most single-wire valvesprings come with a flat wire damper. This damper is designed specifically to dampen the spring's natural frequency, especially when the spring resonates at an rpm where the engine spends time. Dual or triple springs use the friction between the inner and outer springs to perform this damping action.
The beauty of the beehive spring is that its conical shape and variable rate creates numerous, yet less dramatic, natural frequencies, which makes it much less susceptible to a loss of control. There's much more to this concept that would take volumes to discuss (and frankly we don't pretend to understand all of it anyway), but it's enough to say that the beehive spring takes advantage of the physics of its design to give it much more control.
Beehive Buzz
Comp Cams now offers new valvesprings that take advantage of this conical-shape technology to complement our favorite engine. One of the engines we have always had difficulty controlling is a big-block Chevy outfitted with a set of hydraulic-roller lifters. This is due in large part to the weight of a typical big-block valvetrain, which uses large, heavy, 3/8-inch-diameter valve stems. We've experienced mild valve float as low as 5,500 rpm on a hydraulic-roller-cammed 454 H.O. engine.
To prove this point, we bolted up a GM Performance Parts 454 H.O. big-block to Ken Duttweiler's dyno and outfitted it with a Comp Cams Xtreme Energy 282 hydraulic-roller cam, along with a set of World Products Merlin Jenkins oval-port iron heads equipped with a set of Comp single-wire valvesprings (PN 911), which are the standard springs recommended for this cam. We outfitted the rest of the engine with an Edelbrock Performer RPM Air Gap intake, a 750-cfm Holley mechanical-secondary carb, and an MSD distributor sparked by a 6AL box.
Dyno man Ed Taylor performed the dyno test, equipping the Rat motor with the Comp 911 springs. At around 5,500 rpm, the engine went into serious valve float and would not rev past this point. Given the cam's duration and lift, it should have made power at least up to 5,700 rpm, but clearly the engine was limited by valve-control difficulties.
Comp then sent us a set of brand-new PN 29120 beehive springs designed specifically to address this valve-control problem on the Rat motor. As you can see by the accompanying spring-pressure chart, the beehive spring offers 30 pounds more load on the seat (part of this load is due to the shorter installed height), with open pressures within roughly 5 to 10 pounds. We installed these springs directly on the big-block heads with no prep work and with no other changes so we could do a direct comparative test.
Once the engine was back up to temperature, the first test with the new beehive Comp springs was a bit of a shock. We expected to see a slight increase in power above 5,000 rpm, but what we saw instead was a power increase virtually across the entire rpm band from 2,200 rpm to 5,700 rpm. There was only one point at 4,300 rpm where the power numbers were the same. Amazingly, we picked up 26 lb-ft at 2,400 rpm and then 20 hp at the top end at 5,600 rpm. The large gains also both occurred at the opposite ends of the rpm scale, with minimal changes in the middle.
We discussed this test with Comp Cam's spring design engineer Thomas Griffin, and he attributed the increase at low speeds to the additional load pumping the lifter down slightly, which could shorten the cam's duration and boost torque. To test that theory, we installed a set of Comp Cams mechanical-roller lifters and different pushrods to eliminate the hydraulic lifter as a variable. This is not a recommended procedure, but with a very tight 0.004-inch lash for a quick test, Comp felt we could get away with it. This test revealed 11 lb-ft less torque at 2,400 rpm than with the beehive springs and hydraulic lifters. This was at the 432 lb-ft that was the same exact rpm where we had gained as much as 26 lb-ft. This tells us that Griffin's theory was worth at least 11 lb-ft of the total 26 lb-ft change. However, this still leaves at least 15 lb-ft not accounted for.
Conclusion
This one test should by no means create the illusion that this new beehive spring is the ultimate solution to everyone's valvespring problems. Load is certainly still important to control valves at virtually any rpm, but it's also clear from Comp's dynamic testing and our own quick dyno flog that adding in the concept of valvespring frequency and what the entire valvetrain needs to ultimately control the valve is also very important. We've also learned that the lifter side of the valvetrain will benefit greatly from additional stiffness, especially for pushrods (even at the expense of additional weight), while the valve side of the rocker arm will work better by reducing weight. This makes life much easier on the valvespring.
You can expect to see more of this kind of information in the future as valvetrain testing and experimentation continues, but don't be surprised if you begin to see valvespring pressures become more conservative, especially with the use of these new beehive type of spring designs. Are you prepared to become a cone head?