RS-232 Cable Design: Capacitance, Edge Timing, Grounding, and Validation

Define the Complete Point-to-Point Link

Record both RS-232 transceiver part numbers, guaranteed loaded output levels, receiver thresholds and input impedance, load-capacitance rating at the intended data rate, UART framing, cable capacitance per length, connector and protection capacitance, and endpoint ground conditions. RS-232 is single-ended and point-to-point; do not treat a multi-drop RS-485 topology as interchangeable. The Serial Interface Guide introduces the cable budget; this page follows it through electrical validation.

Compute an Electrical Screening Length

Select a supported total capacitive load from the actual driver datasheet under the relevant data-rate and supply conditions. Subtract non-cable load, then divide by the chosen cable capacitance per meter. The RS-232 Cable Length Calculator performs this first-pass estimate and reports propagation-delay references. A traditional 2500 pF figure is a useful historical comparison, not a blanket rating for every transceiver or baud rate.

Using a 2500 pF reference budget, 300 pF fixed load, and a 50 pF/m cable gives (2500 − 300)/50 = 44 m. Choosing a 100 pF/m cable halves the screening result to 22 m. Neither number proves the loaded swing, timing, noise margin, or common-ground suitability of that installed link. Recheck both directions if the endpoint transceivers or wire pairs differ.

Relate Loaded Edges to UART Sampling

Cable capacitance slows voltage transitions. The receiver must see a stable mark or space at its sample point, not merely a crossing somewhere inside the bit. A high baud rate reduces the available settling interval; driver slew behavior, receiver thresholds, oscillator mismatch, and cable noise determine the useful margin. The UART Timing Calculator checks frame and sampling timing, but it does not model an analog RS-232 waveform.

At 115200 baud, one bit lasts about 8.68 µs. At 44 m and a cable velocity of 2×10⁸ m/s, one-way flight time is about 0.22 µs, or 2.53% of the bit interval. This is not a 2.53% baud error: a fixed delay shifts the whole waveform, while clock mismatch accumulates across a frame. Slow edges and ground-referenced noise may be more important than flight time here.

Preserve Voltage and Ground Margin

RS-232 uses voltage relative to a reference conductor. Ground differences, common noise, ESD currents, and cable shield connections can alter the voltage observed at the far-end receiver. Check the chosen receiver's valid input thresholds and absolute limits in every expected ground and power state. Isolation can be appropriate where ground potential or surge conditions are beyond the selected devices, but an isolation decision requires a system safety and installation review. Do not infer safety or EMC compliance from a length calculation.

Verify connector pinout and whether a null-modem crossover is required. Check hardware handshake lines if the application uses them; a working TX/RX pair does not prove RTS/CTS behavior. Avoid copying RS-485-style 120 Ω termination across an RS-232 output without a device-specific design.

Validation Plan

Scope the analog line at the far-end receiver with a suitable probe. A logic-analyzer decode can look clean while voltage or edge margin is small. Count framing, parity, and application errors over meaningful patterns and duration, then repeat with the intended cable, connectors, enclosure, and supply variation. Compare both transmit directions.

Common Mistakes

• Treating a historical 15 m figure as a universal maximum.
• Using 2500 pF without checking the selected driver datasheet.
• Ignoring receiver, connector, PCB, and ESD capacitance.
• Assuming all cable types have the same pF/m.
• Equating capacitance screening length with guaranteed operation.
• Confusing one-way propagation delay with UART baud mismatch.
• Checking baud rate but not the loaded far-end waveform.
• Ignoring the single-ended ground-reference path.
• Applying RS-485 termination rules to RS-232.
• Forgetting handshake wiring or crossover requirements.
• Testing only one transmission direction.
• Claiming isolation or regulatory compliance from a resistor/cable calculation.

Primary Engineering Sources

For load and cabling context, see Analog Devices serial-interface guidance and the TI TRS3122E datasheet application section. For any build, use the exact transceiver and cable datasheets rather than transferring example ratings across devices.