| Electronic Components Datasheet Search |
|
MT8931CP Datasheet(PDF) 13 Page - Mitel Networks Corporation |
|
|
|||||||||||||||||||||||||||||
MT8931CP Datasheet(HTML) 13 Page - Mitel Networks Corporation |
|
13 / 40 page ![]() MT8931C 9-83 The parallel port on the SNIC allows complete control of the HDLC transceiver and access to all data, control and status registers. Reading these registers allows the microprocessor to monitor incoming data on the S or ST-BUS without interrupting the normal data flow. Some registers are classified as asynchronous and others as synchronous. Synchronous registers are single-buffered and require synchronous access. Not all the synchronous registers have the same access times, but all can be accessed synchronously in the time during which the NDA signal is low (refer to Fig. 5). Therefore, it is recommended that the user make use of the NDA signal to access these registers. Since the synchronous registers use common circuitry, it is essential that the register be read before being written. This sequence is important as a write cycle will overwrite the last data received. These parallel accesses must be refreshed every frame. Asynchronous registers, on the other hand, can be accessed at any time. The data in TE or NT Mode Status Register, depending upon the mode selected, is always sent out on the C-channel of DSTo. However, in microprocessor control mode the user can overwrite this data by writing to the DSTo C-channel Register. This access can be done anytime outside the frame pulse interval of the ST-BUS frame. Data written in the current ST-BUS frame will only appear in the C- channel of the following frame. The least significant bit (B0) of the C-channel Register, selects between the control register or the diagnostic register. Setting the B0 of the C-channel Register to ’0’ allow access to the control register. Setting the LSB of the C-channel Register to ’1’ allow access to the diagnostic register. The interpretation of each register is defined in Tables 13 and 14 for NT mode or Tables 16 and 17 for the TE mode. It is important to note that in TE mode, the C-channel Diagnostic Register should be cleared while the device is not in the active state (IS0, IS1 ≠ 1,1). This is accomplished by setting the ClrDia bit of the C- channel Control Register to 1 until the device is activated. In serial control mode, the C-channel on the ST-BUS is loaded into the C-channel Control Register in every ST-BUS frame; the user should make sure that a 1 is written to the ClrDia bit in every frame. However, in parallel control mode the user needs to set the ClrDia bit only once to keep the Diagnostic Register cleared. Once full activation is achieved the Diagnostic Register can be written to in order to enable the various test functions. HDLC Transceiver The HDLC Transceiver handles the bit oriented protocol structure and formats the D-channel as per level 2 of the X.25 packet switching protocol defined by CCITT. It transmits and receives the packetized data (information or control) serially in a format shown in Figure 15, while providing data transparency by zero insertion and deletion. It generates and detects the flags, various link channel states and the abort sequence. Further, it provides a cyclic redundancy check on the data packets using the CCITT defined polynomial. In addition, it can recognize a single byte, dual byte or an all call address in the received frame. There is also a provision to disable the protocol functions and provide transparent access to either serial port through the microprocessor port. Other features provided by the HDLC include, independent port selection for transmit and received data (e.g. transmit on S-Bus and receive from ST-BUS), selectable 16 or 64 kbit/s D-channel as well as an HDLC loopback from the transmit to the receive port. These features are enabled through the HDLC control registers (see Tables 6 and 7). HDLC Frame Format All frames start with an opening flag and end with a closing flag as shown in Figure 15. Between these two flags, a frame contains the data and the frame check sequence (FCS). Figure 15 - Frame Format i) Flag The flag is a unique pattern of 8 bits (01111110) defining the frame boundary. The transmit section generates the flags and appends them automatically to the frame to be transmitted. The receive section searches the incoming packets for flags on a bit-by-bit basis and establishes frame synchronization. The flags are used only to identify and synchronize the received frame and are not transferred to the FIFO. FLAG DATA FIELD FCS FLAG One Byte n Bytes (n ≥ 2) Two Bytes One Byte |
|
Link URL |
| Does ALLDATASHEET help your business so far? [ DONATE ] |
About Alldatasheet | Advertisement | Contact us | Privacy Policy | Link to Datasheet | Link Exchange | Manufacturer List All Rights Reserved©Alldatasheet.com |
| Russian : Alldatasheetru.com | Korean : Alldatasheet.co.kr | Spanish : Alldatasheet.es | French : Alldatasheet.fr | Italian : Alldatasheetit.com Portuguese : Alldatasheetpt.com | Polish : Alldatasheet.pl | Vietnamese : Alldatasheet.vn Indian : Alldatasheet.in | Mexican : Alldatasheet.com.mx | British : Alldatasheet.co.uk | New Zealand : Alldatasheet.co.nz |
|
Family Site : ic2ic.com |
icmetro.com |