FIELD: computer engineering. SUBSTANCE: device is designed to compute functions by setting reference values of functions in interpolation blocks. Function increments are interpolated between interpolation blocks in function increment unit. Errors in reproduction of functions at device result output caused by digital data presentation in interpolation blocks and in function increment unit are relatively compensated for with finite number of bits and reduced impact of error in setting rate of function variation between interpolation blocks by first-order polynomial on resultant computing error. Four structures are proposed for implementing function increment unit depending on bit number n of code being converted and read-only memory capacity. Bit number n is between 11 and 32. Function computation error corresponds to 0.5-0.75 units of less significant bit of result. EFFECT: improved precision of elementary function computation. 6 dwg, 1 tbl
Title | Year | Author | Number |
---|---|---|---|
DEVICE FOR COMPUTING ELEMENTARY FUNCTIONS BY TABLE METHOD | 0 |
|
SU1442984A1 |
APPARATUS FOR CALCULATING FUNCTIONAL RELATIONSHIPS | 2015 |
|
RU2602989C2 |
INTERPOLATOR | 0 |
|
SU1396140A1 |
DEVICE FOR EXTRACTING SQUARE ROOT | 0 |
|
SU1099318A1 |
INTERPOLATOR | 0 |
|
SU1405074A1 |
METHOD MEASURING NONPRIMARY RADIATIONS OF RADIO TRANSMITTERS | 1995 |
|
RU2126976C1 |
DEVICE FOR COMPUTING SINE AND COSINE FUNCTIONS | 0 |
|
SU1494002A2 |
SINE-TO-COSINE CONVERTER | 0 |
|
SU1327095A1 |
DIGITAL FUNCTION GENERATOR | 0 |
|
SU1361547A1 |
DEVICE FOR COMPUTING OF LOGARITHM | 1991 |
|
RU2006916C1 |
Authors
Dates
1999-08-27—Published
1996-06-04—Filed