Abstract: A method for measuring with a maximum error E the phase of a substantially sinusoidal signal, of angular frequency ?=2?/T, sampled with a sampling period T/r, in which the phase is calculated as the time at which a straight line crossing two consecutive samples located on either side of a median value of the signal reaches said value, including the step of selecting number r from a range included between a value r0 and a value equal to from two to three times value r0, such that: E ? max t ? [ - T r0 , 0 ] ? [ t - T r0 · round ? ( G · sin ? ? ? ? ? t ) round ? ( sin ? [ G · ? · t + 2 ? ? r0 ] ) - round ? ( G · sin ? ? ? ? ? t ) ] round(x) being the integer closest to a real number x and G being equal to 2iG1, where i is the number of bits on which are coded the samples and where G1 is a term of correction of the amplitude of the sampled signal.

Abstract: A method for measuring with a maximum error E the phase of a substantially sinusoidal signal, of angular frequency &ohgr;=2&pgr;/T, sampled with a sampling period T/r, in which the phase is calculated as the time at which a straight line crossing two consecutive samples located on either side of a median value of the signal reaches said value, including the step of selecting number r from a range included between a value r0 and a value equal to from two to three times value r0, such that: 1 E ≥ max t ∈ [ - T r0 , 0 ] ⁢ [ t - T r0 · round ⁡ ( G · sin ⁢   ⁢ ω ⁢   ⁢ t ) round ⁡ ( sin ⁡ [ G · ω · t + 2 ⁢ π r0 ] ) - round ⁡ ( G · sin ⁢   ⁢ ω ⁢   ⁢ t ) ]