If you've been in racing any amount of time at all, you know an engine dynomometer is vastly more useful as a tuning tool than for producing a single max-horsepower number for bragging rights at the track. Whether you're doing the work yourself or paying an engine builder to do it for you, a dyno can help tune an engine to your needs.

You can map out the powerband to determine the best gearing; finesse the torque curve to make the power produced more tractable; and even experiment with air/fuel ratios, timing, and fluid temperatures to determine what a particular engine likes best. (We all know no matter how you try to build two engines exactly alike, they always have their own personalities.)

These are fine-tuning steps--meaning you're often dealing with fractions of horsepower. Plus, given the restrictive rules in almost every class of stock car racing today, most engine builders admit all the big gains have been found. The winner in today's horsepower race is the one who can uncover the most "little" gains. Again, we are talking about fractional bumps on the horsepower scale.

It all leads to the question: How reliable are your dyno numbers?

It's relatively easy to get shifting numbers from different dyno runs at different times not because the equipment is poor or difficult to use, but because there are so many variables that can drastically affect an engine's output. Everything from the obvious (air/fuel ratio, air temperature, oil temperature) to things that may not be quite so obvious (barometric pressure, contaminants in the air, how the dyno controls are operated, different batches of fuel) has to be considered. The key is to control as many variables as possible; by keeping them constant from one test to the next, they cannot contaminate the data.

Of course, that's easier said than done. Before you begin controlling variables, you have to identify them and determine a way to measure and track them. "Measuring the variables is part of the issue of getting to control the variables," explains Harold Bettes of Superflow. "In the good old days, it was easier because we simply didn't worry about a lot of those things. There were fewer gauges then, so therefore, less problems. But come to find out, those things we weren't measuring were equally important to those things we were-sometimes more."

Rob Benson heads up the dyno department for Hendrick Motorsports' engine division and is the proprietor of one of the most state-of-the-art systems anywhere on the planet. With seven Winston Cup teams (four in-house) and at least one Busch Series team to support, approximately 800 complete engines will be built at Hendrick Motorsports this year, and every single one will see dyno time. Because of the enormous stakes involved with a poorly performing or broken engine at this level, you can bet Benson is fanatical about the information he gathers from every dyno pull.

"The major component is to control the temperatures," he says. "We try to limit the number of pulls we put on a race engine and make sure when we do our pulls all the temperatures for the air and fluids are within a set range. We try to hold the water temp to within two degrees Fahrenheit and the oil to approximately the same range. We also measure the oil temp in and the oil temp out, which is usually 220 and 240. We even go so far as to control the temperature of the water used in the water brake, which is held within a range of 10 degrees.

"Of course, we've invested in a lot of equipment that allows us to maintain these parameters that an engine builder who isn't involved in Winston Cup probably doesn't have access to. It isn't critical that every engine builder stay within these limits to be successful, but it does give you an idea of how important it can be."

To give you an idea of what Benson is talking about, we've included several photos of one of his dynos and the environmental controls that go with it. Unless you are competing weekly with Hendrick Motorsports for NASCAR's Winston Cup, it's unlikely you can afford to duplicate this level of hardware. (One conditioning system designed solely to provide combustion air that is both temperature and pressure-controlled costs approximately $1,000 a month to operate.) The idea, however, is to emphasize the importance of chasing every variable possible.