欢迎访问ic37.com |
会员登录 免费注册
发布采购

PDSP16116AMCGGDR 参数 Datasheet PDF下载

PDSP16116AMCGGDR图片预览
型号: PDSP16116AMCGGDR
PDF下载: 下载PDF文件 查看货源
内容描述: 16 x 16位乘法器复 [16 X 16 Bit Complex Multiplier]
分类和应用: 外围集成电路输入元件时钟
文件页数/大小: 17 页 / 272 K
品牌: MITEL [ MITEL NETWORKS CORPORATION ]
 浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第7页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第8页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第9页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第10页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第12页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第13页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第14页浏览型号PDSP16116AMCGGDR的Datasheet PDF文件第15页  
PDSP16116  
CLK  
SOBFP  
EOPSS  
A, B, W,  
WTA, WTB  
n
21  
2
3
1
2
3
4
5
6
7
n
1
1
n
2
3
n25  
n
24  
n23  
n22  
n21  
A, B, BTOUT  
GWR  
START OF  
FIRST PASS  
END OF FIRST PASS/  
START OF NEXT PASS  
(MINIMUM NUMBER OF  
LAY CYCLES SHOWN).  
PERIOD BETWEEN  
OTHER INTERMEDIATE  
PASSES IS SIMILAR.  
NOTES  
1. 1 = FIRST CYCLE OF DATA IN PASS  
2. n = LAST CYCLE OF DATA IN PASS  
Fig. 7 Use of the BFP control signals  
In practice, data output may never approach the theoretical  
maximum. Hence, it may be worthwhile to try various universal  
exponents and choose the one best suited to the particular ap-  
plication.  
NB It is easier to simply add the word tag to the exponent for the  
purpose of determing the shift required, instead of modifying it  
according to Table.6. To compensate for this, the universal ex-  
ponent may be increased by one.  
Data is output from the butterfly processor with a two-part  
exponent: the 5-bit GWR applicable to all data words from a  
given FFT and a 2-bit WTOUT associated with each individual  
dataword. To find the complete exponent for a given word, the  
GWR for that FFT must be modified by its WTOUT as shown in  
Table 6. The result is the number of places the binary point has  
shifted to the right during the course of the FFT.  
WTOUT  
GWR  
16-BIT DATA  
SIGN  
BIT  
UNIVERSAL  
EXPONENT  
4-BIT ADDER  
This value must be compared with the universal exponent to  
determine the shift required. This is done by subtracting it from  
the universal exponent. The number of places to be shifted is  
equal to the difference between the two exponents. The shift  
can be implemented in a PDSP1601/A (the shift value is fed  
into the SV port).  
As FFT data consists of real and imaginary parts, either two  
PDSP1601/As must be used (controlled by the same logic) or a  
single PDSP1601/A could be used handling real and imaginary  
data on alternate cycles (using the same instructions for both  
cycles).  
An example of an output normalisation circuit is shown  
in Fig.8. Only 4-bit data paths are used in calculating the  
shift. This means that we must be able to trap very small  
values negative of GWR and force a 15-bit right shift in  
such cases.  
4-BIT SUBTRACTOR  
1111  
4-BIT MUX  
SV PORT  
B PORT  
PDSP1601  
C PORT  
ASRSV  
NORMALISED OUTPUT DATA  
Fig. 8 Output normalisation circuit  
11  
 复制成功!