SN54ABT18646
SCAN TEST DEVICE WITH
18-BIT TRANSCEIVERS AND REGISTERS
SGBS306 – AUGUST 1992 – REVISED AUGUST 1994
4–7
POST OFFICE BOX 655303 • DALLAS, TEXAS 75265
state diagram description
The T AP controller is a synchronous finite state machine that provides test control signals throughout the device.
The state diagram is illustrated in Figure 2 and is in accordance with IEEE Standard 1149.1-1990. The TAP
controller proceeds through its states based on the level of TMS at the rising edge of TCK.
As illustrated, the T AP controller consists of 16 states. There are six stable states (indicated by a looping arrow
in the state diagram) and ten unstable states. A stable state is defined as a state the T AP controller can retain
for consecutive TCK cycles. Any state that does not meet this criterion is an unstable state.
There are two main paths through the state diagram: one to access and control the selected data register and
one to access and control the instruction register. Only one register can be accessed at a time.
Test-Logic-Reset
The device powers up in the T est-Logic-Reset state. In the stable Test-Logic-Reset state, the test logic is reset
and is disabled so that the normal logic function of the device is performed. The instruction register is reset to
an opcode that selects the optional IDCODE instruction, if supported, or the BYP ASS instruction. Certain data
registers also can be reset to their power-up values.
The state machine is constructed such that the T AP controller returns to the Test-Logic-Reset state in no more
than five TCK cycles if TMS is left high. The TMS pin has an internal pullup resistor that forces it high if left
unconnected or if a board defect causes it to be open circuited.
For the SN54ABT18646, the instruction register is reset to the binary value 10000001, which selects the
IDCODE instruction. Bits 87– 84 in the boundary-scan register are reset to logic 0, ensuring that these cells,
which control A-port and B-port outputs, are set to benign values (i.e., if test mode were invoked, the outputs
would be at high impedance state). Reset values of other bits in the boundary-scan register should be
considered indeterminate. The boundary-control register is reset to the binary value 000000000000000000010,
which selects the PSA test operation with no input masking.
Run-T est/Idle
The T AP controller must pass through the Run-T est/Idle state (from T est-Logic-Reset) before executing any test
operations. The Run-Test/Idle state also can be entered following data-register or instruction-register scans.
Run-Test/Idle is a stable state in which the test logic can be actively running a test or can be idle.
The test operations selected by the boundary-control register are performed while the T AP controller is in the
Run-Test/Idle state.
Select-DR-Scan, Select-lR-Scan
No specific function is performed in the Select-DR-Scan and Select-lR-Scan states, and the T AP controller exits
either of these states on the next TCK cycle. These states allow the selection of either data-register scan or
instruction-register scan.
Capture-DR
When a data-register scan is selected, the TAP controller must pass through the Capture-DR state. In the
Capture-DR state, the selected data register can capture a data value as specified by the current instruction.
Such capture operations occur on the rising edge of TCK upon which the T AP controller exits the Capture-DR
state.
Shift-DR
Upon entry to the Shift-DR state, the data register is placed in the scan path between TDI and TDO and, on the
first falling edge of TCK, TDO goes from the high-impedance state to an active state. TDO enables to the logic
level present in the least significant bit of the selected data register.