AN 100: In-System Programmability Guidelines

ID 683546
Date 9/22/2014
Public
Document Table of Contents

1.2.1. TCK Signal

A noisy TCK signal causes most in-system programming failures. Noisy transitions on rising or falling edges can cause incorrect clocking of the IEEE Std. 1149.1 TAP controller. Incorrect clocking can cause the state machine to transition to an unknown state, leading to in-system programming failures.

Because the TCK signal must drive all IEEE Std. 1149.1 devices in the chain in parallel, the signal may have a high fan-out. Like any other high fan-out user-mode clock, you must manage a clock tree to maintain signal integrity. Typical errors that result from clock integrity problems are invalid ID messages, blank-check errors, or verification errors.

Altera recommends pulling the TCK signal low through the internal weak pull-down resistor or an external 1-kΩ resistor.

Fast TCK edges combined with board inductance can cause overshoot problems. When this combination occurs, you must either reduce inductance on the trace or reduce the switching rate by selecting a transistor-to-transistor logic (TTL) driver chip with a slower slew rate. You must not use resistor and capacitor networks to slow down edge rates, because resistor and capacitor networks can violate the input specifications of the device. Use a driver chip to prevent the edge rate from being too slow. Altera recommends using driver chips that do not glitch after power up.