SCSI bus design: What is required at the ends of a SCSI chain? When installing SCSI devices on a shared bus, which component is used to terminate the bus at each physical end to prevent reflections?

Introduction / Context:
SCSI (Small Computer System Interface) uses a multi-drop bus. Proper termination at both physical ends is mandatory to prevent signal reflections that corrupt data. Understanding the nature of termination avoids flaky, intermittent behavior.

Given Data / Assumptions:

• Parallel SCSI flavors require end-of-bus termination.
• Termination may be active or passive but is fundamentally resistive network matching the bus impedance.
• Some devices have onboard terminators or auto-termination jumpers.

Concept / Approach:

Termination presents the characteristic impedance to traveling signals so that they are absorbed rather than reflected. Although physical enablers (jumpers, DIP switches) turn termination on/off, the actual electrical element is a resistor network.

Step-by-Step Solution:

Verification / Alternative check:

Scopes show reduced ringing and stable data eyes with proper terminators attached; removing them yields reflections and bus errors.

Why Other Options Are Wrong:

• BNC: a connector type used for coaxial networks (e.g., 10BASE2), not SCSI termination itself.
• Dip switch: a control to enable termination; it is not the terminating component.
• All of the above: incorrect because only the resistor network is the terminator.

Common Pitfalls:

Leaving three terminators active (both ends plus a mid-device) or mixing active/passive terminators inappropriately. Always check SCSI standard for the bus type (SE, LVD, HVD).

Final Answer:

Resistor