欢迎访问ic37.com |
会员登录 免费注册
发布采购

PDSP16488A0 参数 Datasheet PDF下载

PDSP16488A0图片预览
型号: PDSP16488A0
PDF下载: 下载PDF文件 查看货源
内容描述: 单芯片的二维卷积器与积分行延迟 [Single Chip 2D Convolver with Integral Line Delays]
分类和应用:
文件页数/大小: 33 页 / 414 K
品牌: MITEL [ MITEL NETWORKS CORPORATION ]
 浏览型号PDSP16488A0的Datasheet PDF文件第3页浏览型号PDSP16488A0的Datasheet PDF文件第4页浏览型号PDSP16488A0的Datasheet PDF文件第5页浏览型号PDSP16488A0的Datasheet PDF文件第6页浏览型号PDSP16488A0的Datasheet PDF文件第8页浏览型号PDSP16488A0的Datasheet PDF文件第9页浏览型号PDSP16488A0的Datasheet PDF文件第10页浏览型号PDSP16488A0的Datasheet PDF文件第11页  
IP7:0  
IP7:0  
512  
512  
512  
512  
BYPASS  
BYPASS  
512  
512  
512  
512  
IP7:0  
512  
BYPASS  
L7:0  
838  
ARRAY  
838  
ARRAY  
BYPASS  
16  
16  
512  
512  
BYPASS  
434  
OR  
834  
ARRAY  
512  
512  
512  
512  
512  
512  
512  
512  
512  
16  
16  
L7:0  
512  
BYPASS  
L7:0  
IP7:0  
512  
512  
IP7:0  
1024  
1024  
BYPASS  
1024  
BYPASS  
512  
434  
OR  
434  
OR  
1024  
1024  
L7:0  
834  
834  
1024  
1024  
ARRAY  
ARRAY  
1024  
L7:0  
Fig. 4 Line delay configurations  
To support real time applications an option is provided in  
which the length of the line delay is defined by the number of  
clocks obtained while the HRES input is low. HRES would  
normally be composite sync when the convolver is directly  
attached to an NTSC or PAL video input.  
The line delay is achieved by reading the previous contents  
of a RAM-based line store, and then writing new information to  
the same address. When HRES is high, write operations are  
inhibited, and the address counter is reset. During an active line  
the counter is incremented by the pixel clock. If the maximum  
count is reached before the end of a line, then write operations are  
terminated and wraparound effects avoided.  
The rising edge of HRES, marking the end of a line, is  
normally asynchronous to the pixel clock, and it is possible for an  
additional pixel to be stored on some lines. This has no effect on  
the convolver operation, and will not cause a cumulative shift in  
the pixel position from line to line.  
An alternative means of defining the line length is, however,  
provided when an exact number of pixels is needed. HRES going  
low then starts the delay operation for every line, but it ceases  
when the 10-bit value contained in two registers is reached. This  
method can avoid the need to store blank pixels at the end of a  
line before HRES goes high. With this method the line must  
contain an even number of pixels but the value loaded into the  
control registers, defining the line length, must be one less than  
the even number required.  
HRES input goes active. This is present on the DELOP output  
pin. This will ensure that no pixels on the right hand edge are lost  
due to the internal pipeline delay. If the pixel clock is a continuous  
signal, the user must ensure that the HRES high to low transition  
meets the timing requirements defined in Fig. 10. The HRES  
rising edge at the end of a line need not be synchronized.  
When pixels are read or written to a frame store, an alternative  
line delay configuration is needed. Within the frame store lines  
would be stored in contiguous locations, with no gaps caused by  
the flyback period between the lines. This method of use makes  
the HRES defined line delay operation difficult to use, and an  
alternative mode of operation is provided. The HRES input is then  
driven by a system-provided signal, which defines a complete  
frame store update period. It is not a line defining signal. The high  
to low transition of this signal will initiate the line store update  
sequence and allow the internal address pointers to increment.  
These pointers will be synchronously reset at the end of a line,  
when they reach the pre-programmed value. They will then  
immediately start a new operation using address zero. The actual  
line delay must be pre-loaded into two control registers as  
described previously.  
Write operations back to the frame store must allow for the  
total pipeline delay. This can be achieved by inhibiting write  
operations until DELOP goes low. Write operations then continue  
until it goes back high. The PDSP16488A assumes that data is  
valid when a clock signal is applied, and that it also meets the set  
up and hold requirements given in Fig. 10. If data is not valid due,  
for example, to a frame store DRAM refresh cycle, then the user  
must externally inhibit the clock. The clock supplied to the  
convolver will in this mode be a signal which defines a frame store  
cycle time.  
In an image processing system, the pixel clock is often re-  
synchronized, or even inhibited, during blanking or sync. The  
next line is then started with a precise time interval from the end  
of sync (falling edge of HRES) to the first pixel clock edge. This  
avoids any visible pixel jitter at the beginning of the line, which  
would otherwise be present since pixel clock is asynchronous  
with respect to video sync pulses.  
The use of the convolver in a line scan system is similar to its  
use with a frame store. These systems have no flyback period,  
and the address counter must be synchronously reset at the end  
of the line and then allowed to continue.  
When using the PDSP16488A the pixel clock should not be  
inhibited, or re-synchronized, until the delayed version of the  
7
 复制成功!