欢迎访问ic37.com |
会员登录 免费注册
发布采购

MT9074AL 参数 Datasheet PDF下载

MT9074AL图片预览
型号: MT9074AL
PDF下载: 下载PDF文件 查看货源
内容描述: T1 / E1 / J1单芯片收发器 [T1/E1/J1 Single Chip Transceiver]
分类和应用: 电信集成电路
文件页数/大小: 122 页 / 371 K
品牌: MITEL [ MITEL NETWORKS CORPORATION ]
 浏览型号MT9074AL的Datasheet PDF文件第24页浏览型号MT9074AL的Datasheet PDF文件第25页浏览型号MT9074AL的Datasheet PDF文件第26页浏览型号MT9074AL的Datasheet PDF文件第27页浏览型号MT9074AL的Datasheet PDF文件第29页浏览型号MT9074AL的Datasheet PDF文件第30页浏览型号MT9074AL的Datasheet PDF文件第31页浏览型号MT9074AL的Datasheet PDF文件第32页  
MT9074  
Advance Information  
(TSLPD). The relative phase delay between the  
system frame boundary and the transmit elastic  
frame read boundary is measured every frame and  
reported in the Transmit Slip Buffer Delay register-  
(page 3H, address 17H). In addition the relative  
offset between these frame boundaries may be  
programmed by writing to this register. Every write to  
Transmit Elastic Buffer Set Delay Word resets the  
transmit elastic frame count bit TxSBMSB (address  
17H, page 3H). After a write the delay through the  
slip buffer is less than 1 frame in duration. Each write  
operation will result in a disturbance of the transmit  
DS1 frame boundary, causing the far end to go out of  
sync. Writing BC (hex) into the TxSBDLY register  
maximizes the wander tolerance before a controlled  
slip occurs. Under normal operation no slips should  
occur in the transmit path. Slips will only occur if the  
input C4b clock has excess wander, or the Transmit  
Elastic Buffer Set Delay Word register is initialized  
too close to the slip pointers after system  
initialization.  
tolerance). The MT9074 will allow 92 usec (140 UI,  
DS1 unit intervals) of wander and low frequency jitter  
before a frame slip will occur.  
When the C4b and the E1.5o clocks are not phase-  
locked, the rate at which data is being written into the  
slip buffer from the DS1 side may differ from the rate  
at which it is being read out onto the ST-BUS. If this  
situation persists, the delay limits stated in the  
previous paragraph will be violated and the slip  
buffer will perform a controlled frame slip. That is, the  
buffer pointers will be automatically adjusted so that  
a full DS1 frame is either repeated or lost. All frame  
slips occur on frame boundaries.  
The minimum delay through the receive slip buffer is  
approximately 1 usec and the maximum delay is  
approximately 249 uS. Figure 13 illustrates the  
relationship between the read and write pointers of  
the receive slip buffer (contiguous time slot  
mapping). Measuring clockwise from the write  
pointer, if the read page pointer comes within 8 usec  
of the write page pointer a frame slip will occur,  
which will put the read page pointer 157 usec from  
the write page pointer. Conversely, if the read page  
pointer moves more than 249 usec from the write  
page pointer, a slip will occur, which will put the read  
page pointer 124 usec from the write page pointer.  
This provides a worst case hysteresis of 92 usec  
peak = 142 U.I.  
The two frame receive elastic buffer is attached  
between the 1.544 Mbit/s DS1 receive side and the  
2.048 Mbit/s ST-BUS side of the MT9074. Besides  
performing rate conversion, this elastic buffer is  
configured as a slip buffer which absorbs wander  
and low frequency jitter in multi-trunk applications.  
The received DS1 data is clocked into the slip buffer  
with the E1.5o clock and is clocked out of the slip  
buffer with the system C4b clock. The E1.5o  
extracted clock is generated from, and is therefore  
phase-locked with, the receive DS1 data. In the case  
of Internal mode (pin BS/LS set low) operation, the  
E1.5o clock may be phase-locked to the C4b clock by  
an internal phase locked loop (PLL). Therefore, in a  
single trunk system the receive data is in phase with  
the E1.5o clock, the C4b clock is phase locked to the  
E1.5o clock, and the read and write positions of the  
slip buffer track each other.  
The RSLIP and RSLPD status bits (page 3H,  
address 13H, bits 7 and 6 respectively) give  
indication of a receive slip occurrence and direction.  
A maskable interrupt RxSLPI (page 1H, address  
1BH, bit 0 - set high to mask) is also provided. RSLIP  
changes state in the event of a slip. If RSLPD=0, the  
slip buffer has overflowed and a frame was lost; if  
RSLPD=1, an underflow condition occurred and a  
frame was repeated  
In a multi-trunk slave or loop-timed system (i.e.,  
PABX application) a single trunk will be chosen as a  
network synchronizer, which will function as  
described in the previous paragraph. The remaining  
trunks will use the system timing derived from the  
synchronizer to clock data out of their slip buffers.  
Even though the DS1 signals from the network are  
synchronous to each other, due to multiplexing,  
transmission impairments and route diversity, these  
signals may jitter or wander with respect to the  
synchronizing trunk signal. Therefore, the C1.50  
clocks of non-synchronized trunks may wander with  
respect to the C1.50 clock of the synchronizer and  
the system bus. Network standards state that, within  
limits, trunk interfaces must be able to receive error-  
free data in the presence of jitter and wander (refer  
to network requirements for jitter and wander  
Slip Buffer in E1 mode  
In E1 mode in addition to the elastic buffer in the jitter  
attenuator(JA), another elastic buffer (two frames  
deep) is present, attached between the receive side  
and the ST-BUS (or GCI Bus) side of the MT9074 in  
E1 mode. This elastic buffer is configured as a slip  
buffer which absorbs wander and low frequency jitter  
in multi-trunk applications. The received PCM 30  
data is clocked into the slip buffer with the E1.5o  
clock and is clocked out of the slip buffer with the  
C4b clock. The E1.5o extracted clock is generated  
from, and is therefore phase-locked with, the receive  
PCM 30 data. In normal operation, the C4b clock will  
be phase-locked to the E1.5o clock by a phase  
locked loop (PLL). Therefore, in a single trunk  
28  
 复制成功!