Modbus Modes and Timing Workbench

Compare RTU, TCP, and ASCII timing with character gaps, frame expansion, and polling throughput.

Learner-ready Modbus modes and timing workbench with RTU T1.5/T3.5 calculations, ASCII expansion, TCP latency model, Play/Step timeline, mode presets, and mobile-safe quick reference.

RTU gaps ASCII expansion TCP latency Polling budget

Modbus Modes and Timing Workbench

Compare the same read request in RTU, ASCII, and TCP. The timeline shows which part of the cycle is bytes on the wire, required silence, device turnaround, or network latency.

ModeRTU serial

Cycle time22.3 ms

RTU T3.54.01 ms

Throughput44.8 polls/s

Stage 1: select the transport mode

RTU and ASCII use serial character timing. TCP uses an MBAP header and depends more on network latency than byte time.

RTU

Binary serial Compact frames with CRC-16 and strict silent intervals for frame boundaries.

TCP

Ethernet/IP MBAP header replaces serial timing and CRC; throughput is usually latency-limited.

ASCII

Hex text serial Human-readable colon and CR/LF frame, but each binary byte becomes two hex characters.

RTU poll cycle Request 8 B, response 11 B

Character time1.146 ms

T1.5 / T3.51.72 / 4.01 ms

Wire bytes/chars19 bytes

Useful payload31.6%

RTU is efficient, but the silence matters At 9600 baud, the required RTU frame boundary gap is a visible part of the poll cycle.

What to notice

Timing controls

Change baud rate, payload size, device delay, and network latency. The timeline rescales to the total cycle.

Mode RTU serial TCP Ethernet ASCII serial

Serial baud rate 1200 2400 4800 9600 19200 38400 57600 115200

Registers read

3

Device turnaround

3 ms

TCP round trip

6 ms

Devices polled

20

Core idea

RTU frame boundaries are timed silence, not bytes. ASCII frame boundaries are visible delimiter characters. TCP frame boundaries come from MBAP length over a stream.

Common trap

Raising baud rate helps serial byte time, but slow device turnaround and required gaps can still dominate small polls.

Planning use

Multiply one poll cycle by the number of devices to estimate scan time before tuning timeout values or polling rates.

Quick reference and technical notes

RTU character timeThis workbench uses 11 serial bits per character. Character time = 11 / baud.

RTU low baud gapsAt 19200 bit/s or below, T1.5 and T3.5 are 1.5 and 3.5 character times.

RTU high baud gapsAbove 19200 bit/s, the serial-line guide recommends fixed 750 us and 1.750 ms timers.

ASCII expansionAddress, PDU, and LRC bytes are sent as two ASCII hex characters, plus colon and CR/LF delimiters.

TCP timingTCP removes Modbus CRC and silent-gap timing, but network latency and socket behavior affect poll rate.

Efficiency shownUseful payload is response register data divided by the modeled request plus response wire size.

Numbers are planning estimates for a read-holding-registers transaction. Real systems also need timeout margins, retries, gateway delay, and device-specific processing time.

Modbus over Serial Line v1.02 Modbus Application Protocol v1.1b3 Modbus frame structure workbench