The dyno test has long been a trusted Sport Compact Car yardstick for measuring performance, but keeping that yardstick consistent has taken some effort. Think how difficult it would be to build anything if every yardstick, ruler and measuring tape were calibrated a little differently. What if, say, an inch on a Stanley measuring tape were slightly shorter than any other inch, or worse, if the inches changed length ever so slightly every time you used it.
This used to be the problem we faced with dyno testing. Different brands of dynos use different measuring methods and have different calibrations, resulting in the same car, with the same output, registering different readings on every dyno. On some dynos, it was even difficult to get the same reading twice in a row on the same car.
It was into this arena that the Dynojet arrived. Dynojets have proven to be remarkably consistent from one dyno to the next, and very easy to operate. After verifying the consistency for ourselves, the Dynojet has become our standard dyno, ensuring that every horsepower we measure is the same size.
And everyone lived happily ever after...until now. Now, of course, we have the new all-wheel-drive WRX, and we desperately want to dyno test it. Since you can't fit all four wheels on one Dynojet roller, our standard yardstick is useless. We solved that problem two different ways this month. First, in Project Impreza 2.5 RS, we simply converted the car to front-wheel drive.
Simply may be the wrong word here, since the center differential had to be removed and replaced with a modified unit that didn't connect to the rear wheels, but at least the job can be done with the transmission in the car. The other solution, the one we used with our WRX comparison test, was to use the newest dyno on the block, the Dynapack. The Dynapack can handle all-wheel-drive cars just fine, but because it measures horsepower differently than a Dynojet and is calibrated differently, we have no idea how Dynapack horsepower will compare with Dynojet horsepower. Time for more testing.
Before we start looking at data, however, let's take a look at how each dyno works. Chassis dynos (that is, those that measure power with the engine in the chassis, as opposed to engine dynos that require the powerplant to be yanked from the car) come in two basic forms; brake dynos and inertia dynos. The Dynojet is an inertia dyno, so it measures power passively by watching how quickly the engine can spin a massive steel drum. Dynojets are physically very simple. The drive wheels sit on a giant steel roller that happens to have a lot of rotational inertia. Sensors on the roller monitor how quickly the drum accelerates, and because the drum's inertia is a known quantity, horsepower is just a little math away. The mechanical simplicity of the Dynojet is one of the keys to its repeatability. Unless some of that mass evaporates, there is little that could affect the calibration of the dyno.
Dynojets do have their limitations, however. While they are seemingly ideal for simply and consistently verifying power output, they are less useful for tuning than a brake dyno. Brake dynos measure power by using some sort of brake to prevent the engine from accelerating and then measuring how much torque they had to exert to resist the engine's advances. While an inertia dyno can only measure full throttle sweeps, a brake dyno can be set up, for example, to measure power output at 30 percent throttle and 3250 rpm, or any other combination of load and rpm. This ability can prove very useful for tuning engine management systems, where all the cells in a fuel or timing table need to be just right.
Brake dynos tend to be far more complicated than inertia dynos, however, since they not only have to have a brake, but also an accurate way to measure how much torque that brake is exerting. The Dynapack uses a hydraulic pump attached to each wheel. Restricting the outlet on the pump turns it into a brake, and measuring the pressure the pump builds up turns out to be an accurate measurement of how much torque the engine is exerting to turn the pump.
All-wheel-drive dynos tend to be very big and very complex, since two sets of rollers have to be used, and the distance between the rollers has to be adjustable to match the wheelbase of whatever car is being tested. The Dynapack deals with this by eliminating the complexity of rollers and, in fact, eliminating the drive wheels altogether. Each hydraulic brake is contained in a small box that can be positioned at each corner of the car. Each drive wheel is removed, an adapter is bolted in place and connected directly to the box. This makes the Dynapack amazingly flexible.
For our WRX test in this issue, we wanted to test seven WRXs and two two-wheel-drive calibration cars while simultaneously doing acceleration, braking and slalom tests. Scott Lampkin and John Card from Dynapack showed up with a truck, unloaded their four Dynapack brakes, a computer and a small generator, set them up under the wing of an airplane and dyno-tested all nine cars on the tarmac of an abandoned runway. The flexibility is incredible.
The lack of rollers offers a unique and unexpected benefit as well-no tire noise. During a dyno pull, the only noise is the engine itself, making it far easier to listen for detonation, misfire, or any other telltale aberration that might affect your tuning.
To see how Dynapack readings compare with the Dynojet readings we are used to, we brought out a pair of two-wheel-drive cars that had already been tested on a Dynojet. The Dynapack, it turns out, is capable of testing in a variety of different ways, but we chose a sweep, where it lets the engine accelerate between two set rpm points over a set period of time. We chose this method both because it most closely replicates the Dynojet's technique, and because it most closely replicates how you use a car on the street; you hit the gas, you accelerate. The procedure gets complicated, however, because the machine is so flexible. You can choose whatever starting and ending points you want (we chose 2000 rpm and redline, as we do with all our Dynojet pulls), you can choose how long the sweep takes and you can choose the settling time, or how long the car is held at the starting rpm before the run starts. The duration of the pull, known as "ramp time" is actually very important. As we have seen when testing cars in different gears on Dynojets, an engine sweeping quickly through its rev band will make less power than one that is loaded more heavily and therefore sweeps more slowly. On a turbocharged car, the engine can actually out-accelerate the turbo, slightly reducing boost. Even on naturally aspirated cars, however, the longer pulls tend to make more power, since less power is being wasted at any given moment accelerating the crank, flywheel and drivetrain.
What we found, from our two-wheel-drive tests, is that the Dynapack and Dynojet give surprisingly similar readings. If you compare a 7-second pull on the Dynapack with a third-gear pull on the Dynojet (which happens to take 9 seconds), the charts agree to within about 2 hp. The same is true when comparing a 15-second pull on the Dynapack with a fourth-gear pull on the Dynojet (fourth gear, in this case, takes about 17 seconds). The agreement between dynos was similar on both cars.
I actually performed this test in an effort to determine some sort of reliable conversion from one measurement to the other, but it appears that no conversion is necessary. As with any dyno, however, before comparing two different dyno charts, you should know the conditions under which the test was performed and be absolutely sure, in the case of the Dynapack, that the ramp time is similar.
Comparing all-wheel-drive Dynapack readings to two-wheel-drive Dynojet readings is still a little questionable, since more driveline losses will be seen on the AWD car. On the other hand, those losses are really there, and the Dynapack reading is what is really reaching the ground. If you are trying to compare engines, this isn't the best method, but if you want to compare actual performance results, we've found our new all-wheel-drive dyno standard.