It's important to place the piston as close to top dead center (TDC) as possible. Then rock the engine across TDC to properly load the rings to get an optimal seal. The best time to test is with the engine at operating temperature.
Results
Everyone wants to have a hero engine, but don't believe all the hype you hear about 2 percent leakage engines. No cylinder seals perfectly, especially if the engine is nitroused, supercharged, or turbocharged where the top and second rings are set with wide ring endgaps. Even for normally aspirated engines, respectable leakage numbers would be anywhere from 8 to 12 percent with a variation between cylinders of 4 to 5 percent, but it's possible the variation could be as high as 10 percent.
So if you test an engine and see leakage numbers even in the 15 to 20 percent range, this is not cause for alarm. Much of your concern around an engine that needs freshening should revolve around exactly where the leakage originates. That's where we'll go next.
Prop the throttle blades open to listen for air escaping past intake valve back into the intake manifold. Your ears don't have to be as big as Rat Fink's, but it helps.
Always make sure to test cylinder leakage with the piston at the top of the cylinder. This is where the greatest cylinder wear occurs. It does little good to measure leakage with the piston at bottom dead center since this is not where the engine makes power.
Leak Paths
There are three major leak paths where cylinder pressure can escape: past the rings, the intake valve, or the exhaust valve. We'll ignore the head gasket for now. With each of these three leak sources, it's easy to tell where the leakage is originating with some sleuthing. If you can hear a hissing sound coming from the valve cover breather hole or from the dipstick tube, then the air is escaping past the rings. If you prop open the throttle blades and hear that same hissing sound, then the pressure is leaking past the intake valve. If the air is escaping past the exhaust valve, you will probably be able to hear the air even as far back as the tailpipe.
Poor ring seal on a leakdown test will make itself known with a hissing sound clearly audible out of the PCV valve hole or from the dipstick tube. Be careful: Too much blow-by will make your eyes bloodshot.
Let's evaluate each of these leak paths separately. Say your engine measures 30 percent or more for each cylinder, and it's all running past the rings. This is a good indication that your engine is tired and could use rebuilding. But if the leakage number is under 20 percent and most of it's past the rings, then be advised that a rebuild is not going to be worth 50 or perhaps even 20 hp. As an example, on the nitrous small-block Chevy story in this issue, we replaced pistons and rings in the engine (granted the cylinder wall finish was not corrected) and picked up less than 5 hp, the equivalent of just over 11⁄2 percent gain. The point here is that even cylinders with 20 percent leakdown do not represent a major horsepower loss.
Tapping directly on the valve tip (or rocker arm tip) with a plastic mallet can help diagnose a leaking intake or exhaust valve. Often, a simple tap like this will reduce the leakage to acceptable levels.
The consensus among performance engine tuners for leakage past the exhaust valve is that it would have to be massive--over 30 percent or more--before it should be of concern. Flowmaster Dyno guru Kevin McClelland suggests tapping the exhaust valve with a plastic mallet to help seat the valve if you find high exhaust valve leakage. Often, this might help, but Kevin also says that cylinder pressure will help seal the valve during the compression and combustion cycles.
The area where leakage presents the most concern is the intake valve. If you detect a significant amount of air escaping past the intake valve during the leakdown test, try tapping on the intake valve with a plastic mallet to see if this will reduce the leakage. If the leak persists and is well into the 20 percent zone, it might be worth the time to yank the head and repair the problem. However, if you're looking at just one cylinder with difficulties, you can guess where we're going. As an example, we floated the valves on a 430hp small-block Chevy and tagged a valve lightly against the piston enough that the seal went away. We replaced the valve and went back on the dyno only to discover the engine generated no increased power anywhere in the power curve.
Conclusions
The overall point is that a leakdown tester is a great diagnostic tool to identify real (or imagined) cylinder pressure difficulties. But don't be alarmed if you get 15 to 18 percent leakdown on all cylinders. If the cylinders perform within a couple of percentage points of each other, it's time to look elsewhere for major power gains. CC
Don't be shocked if you see leakage from an otherwise good-running street engine around 20 percent. You could do an excellent rebuild and still not pick up appreciable power even if the leakage was substantially reduced. | 
Compression testing is another good tool to use when diagnosing an engine. Combining the results of a leakdown test with a cranking compression test can indicate an engine's potential. For a street engine, cranking pressure between 175 and 200 psi is ideal. This is a Matco compression tester. |
Another excellent tuning tool is a vacuum gauge. This is a Matco combination vacuum and pressure gauge. Besides the usual intake manifold vacuum testing with a range up to 28 inches of vacuum, you can also use this gauge to test the exhaust system up to 10 psi. | |