| 摘要 |
An edge multiplier circuit comprises a chain of N phase-looped delay cells (130, 131, 132, 133, 134). An order of cells to be delayed is determined by action loops. A first action loop (116, . . . , 128) is utilized for values of j varying from 1 to N, each corresponding to a total delay equal to j times an elementary delay (Te) of a cell. The delay is applied to the chain of N delay cells. An action of the first loop comprises a second action loop (118, . . . , 127) for values of i varying from 1 to N, each corresponding to a rank of a cell in said chain. An action of the second loop calculates a delay error (a (j, i)) output from the cell of rank i relative to an ideal delay that distributes the total delay of the chain equally to each cell. A first value of the error is calculated prior to activating an additional elementary delay (Te) of the cell of rank i and a second value of the error is calculated if the first value is higher than a predetermined threshold, after activating an additional elementary delay of said cell of rank i. The rank i in this case completes the order to be determined.
|
