Define LVDS Receiver Output for Each Idle and Fault State

A stable output with an unplugged cable does not prove that an LVDS receiver behaves identically when its inputs are terminated or shorted. Fail-safe features are specified for particular conditions. Review each condition that the product can encounter and keep unsupported fault assumptions visible.

Read the named conditions

TI’s DS90LV012A/DS90LT012A datasheet explicitly discusses open, shorted and terminated-input fail-safe behavior. That coverage belongs to those devices and their stated conditions. It should not be inferred for another receiver merely because both accept LVDS signals.

Create a state table for normal drive, transmitter disabled, transmitter unpowered, cable disconnected at each end and relevant conductor faults. Record the expected receiver output and the source of that expectation. Use an unknown entry when the datasheet does not establish the behavior.

Keep the termination in the drawing

Show whether termination is internal or external and where it sits relative to the connector. Unplugging a cable can leave the receiver terminated in one design and open in another. A test report saying cable removed is therefore incomplete without the connection arrangement.

Also record the common-mode conditions and any external bias network. A fault that drives one conductor toward a supply is not equivalent to simply removing the differential signal. Check the device’s permitted input conditions before designing or performing a fault test.

Define the system meaning of idle

Choose what the downstream circuit should do when no valid stream is present. A fail-safe high or low output is an electrical state, not proof that a decoded data word is valid. The receiving logic may need framing, timeout or other protocol evidence before accepting information.

For a clock input, determine how the rest of the system recognizes loss of transitions. For a data input, identify whether a static output could resemble a meaningful command. Keep this behavior in the interface requirement rather than assuming the receiver component resolves the application decision.

Verify representative transitions

Capture the receiver output while entering and leaving the relevant idle states. Use the installed cable, termination and loading arrangement. Record any pulses observed during transitions separately from the final static state, because downstream logic may react to the transition even when the eventual level is correct.

The completed review should identify supported fail-safe cases, measured behavior and any remaining unsupported fault condition. If the receiver or termination changes, revisit the table. This makes the missing-signal behavior a defined interface property rather than a conclusion drawn from one unplugged-cable demonstration.