Choosing a Multiplexer's Break-Before-Make Behaviour

1. Define the two paths involved in the transition

A selection change involves a previously conducting path and a different path that is to conduct next. Define both roles before comparing switching terminology. The important requirement here is their order during the handover. A statement describing the selected connection after the change does not, by itself, describe whether the old and new paths were connected together during the transition.

TI's TMUX1208/TMUX1209 descriptions provide a named context for that question. The relevant comparison concerns how a device moves from one selected path to another. It differs from merely confirming that the required inputs can be routed. Two candidates might offer the same broad routing function while leaving a different or unspecified handover behavior between selections.

Consider a measurement route that changes between two signal sources. The selection requirement should distinguish the source previously connected from the source intended next, without prescribing an address pattern here. That makes the non-overlap question understandable to engineering. It also prevents a general multiple-input label from being treated as evidence for the transition behavior the circuit actually requires.

2. Explain the documented break-before-make interval

TI defines break-before-make as disconnecting the on-state switch before making the connection with the next on-state switch. The intervening time is the break-before-make delay. This definition contains two distinct events and their order. Retaining them explains what the feature addresses without needing a numerical interval or an interpretation of the manufacturer's measurement figure.

The interval is therefore not simply another name for the time taken to arrive at the new selected connection. It concerns a period between breaking and making paths. If an application requires non-overlap, a generic transition-speed claim leaves that requirement unanswered. The exact device's documented order is the relevant evidence before any separately established timing acceptance condition can be assessed.

Use this definition at the component boundary. It supports the named device's between-path switching description under its applicable conditions. It does not demonstrate every consequence for external sources or receivers. A sourcing comparison should preserve the required order and identify whether the candidate documents it, while leaving application-specific timing and interaction questions to their own evidence.

3. Separate address transition from enable timing

Address control selects the route, while enable control determines whether the selected connection is allowed to operate. TI discusses turn-on and turn-off timing associated with enable separately from break-before-make. These descriptions use different controlling events. An enable-response statement should therefore not replace evidence about the handover from one selected path to another, even though both affect when a connection conducts.

Bidirectional operation is another independent characteristic in the named product description. It concerns conduction through a selected signal path in either direction. It does not define which path opens first during a selection change. A requirement for bidirectional routing can coexist with a non-overlap requirement, but documenting the first does not establish the second or supply the transition behavior by implication.

Control-input fail-safe behavior addresses the logic-input and supply relationship. It is not a statement that the analog signal paths have every form of powered-off protection. Keeping these distinctions explicit avoids assigning one feature to the wrong connection or power condition. The comparison here supplies no address sequence, physical pin instruction or power-sequencing procedure; it identifies which functional description answers each question.

4. Keep path isolation and settled measurement evidence distinct

Opening the previous route before connecting the next prevents overlap in the documented switching description. It does not establish when the receiving measurement has settled after the new connection. Switching order and settled observation describe different events. A receiver can have a new selected path while the wider signal chain still requires its own applicable evidence before a measurement is accepted.

This is especially important beside an ADC requirement. Channel architecture concerns whether observations are simultaneous or selected in sequence. SAR acquisition concerns the input chain during its acquisition interval, while delta-sigma filter latency concerns retained measurement history. Break-before-make answers none of those complete measurement questions. It contributes a specific routing condition that should remain visible alongside them rather than being promoted into a universal settling assurance.

Finish the candidate comparison by identifying the required old-path/new-path order, the separate use of enable and the measurement condition needing further assessment. Ask engineering whether the documented switching description covers the intended route change. That produces a focused technical request without deriving a wait time, declaring an ADC result settled or claiming protection of a signal path after its switch supply has disappeared.

Sources and further reading

• Texas Instruments TMUX1208/TMUX1209 SCDS389C — break-before-make and functional descriptions

TMUX1208/TMUX1209 saved text on physical pages 17 and 20 was reread for break-before-make, enable and functional-role definitions. Figures and numeric timing were not interpreted. No address recipe, powered-off signal guarantee, whole-datasheet review or measured settling result is supplied.