The cylinder test for deriving detonation energies is fully described and analyzed, including the relation between streak camera and Fabry-Perot interferometer data. The 6, 12.5 and 19 mm scaled displacements are modified to represent average relative volumes of 2.2, 4.4 and 7.2. CHEETAH V3.0 is introduced with its new all-Hugoniot calibration, which produces the most accurate detonation energies to date. Using CHEETAH V3.0 energies at the three cylinder volumes, we find that 107 cylinder shots show full burn. Of these, 52 are taken as standards with an error of +/-2.6%. Another 26 show higher energies, which cannot be explained. Finally, 111 shots show lower energies and these are taken to represent slow kinetics. The composite explosive PBXN-111 is taken as a special problem, because unconfined and confined Size Effect data plus two different-size cylinder tests have been done. The form of JWL++ with one fully-reacted JWL, as copied from Ignition & Growth, failed to fit the data. The two rate constants interact too strongly, so that the form works for 90 % cylinder energies but not for the 25 % found in PBXN-111. We, therefore, constructed a 2-JWL form with the first, fast rate describing the detonation velocity and the second, slow rate the cylinder energies. This form of JWL++ isolates the rate constants and works for slow-reacting explosives. The simple form of JWL++ allows the fast and slow rates to be estimated directly from the data. The Gurney velocity divided by the detonation velocity for the standard explosives is found to be 0.30, 0.33, 0.34 and 0.35 for the relative volumes v = 2.2. 4.4, 7.2 and the maximum measured.