| JTAG is an industry standard method of serially accessing signals or test points. Originally JTAG was designed as a replacement for "bed of nails" circuit testers, as shrinking board geometries made such testers increasingly fragile and expensive. More recently, JTAG is being incorporated in large chip designs as a way to independently test the chip subsystems by sensing and controlling signals at each subsystem boundary. | |
| The JTAG specification does not limit its use to boundary connections, however. Design-specific JTAG loops can access internal memories, override control signals or even initiate BIST sequences. | |
| The JTAG interface provided on the tester baseboard consists of 5 signals: | |
| TCK - Test Clock. The source for TCK will be either the adjustable onboard oscillator or a user-provided source. TCK need not be synchronous with the D.U.T. clock, although it may be simplest to just use one clock. | |||
| TMS - Test Mode Select. The single control bit used to select the JTAG function. | |||
| TDI - Test Data In. A single data bit into the D.U.T. | |||
| TDO - Test Data Out. A single data bit out of the D.U.T. | |||
| TRST - Test Reset. This signal places the JTAG controller into a known initial state. | |||
| A white paper on JTAG by Sun Microsystems is available; also an online tutorial by Asset InterTech has additional information. | |
| ------------------------------------------------------------------------------------------------------------ | |
| The state diagram for JTAG control is shown here: | |
| ------------------------------------------------------------------------------------------------------------ | |
![]() |
|