Repeatable LAN test worksheet

A single speed result cannot identify a faulty router. Use this worksheet to compare repeated measurements under the same conditions, then change one thing and repeat. It calculates numbers you enter; it does not run a network test, probe devices, or upload these inputs.

1. Define the path before measuring

Record the client and server, which device sends, Ethernet or Wi-Fi on each end, link rate, room or access point, test duration, number of streams, and whether a VPN is active. Keep the endpoints and test settings fixed. Pause cloud backups and other large transfers, keep both devices awake, and note anything you cannot control.

Run the same direction three times. For a Wi-Fi comparison, first use a wired server so a second wireless hop does not confound the result. Test the reverse direction separately: a fast receive result does not establish fast sending performance. Save each raw result alongside these notes.

2. Enter three results in Mbps

Run 1Run 2Run 3File size in decimal GB

Calculate comparison

Enter measurements to calculate.

Print worksheet

3. Read the comparison without overdiagnosing

For fictional runs of 400, 450 and 500 Mbps, the median is 450 Mbps and the range is 100 Mbps. That median equals 56.25 decimal MB/s. A 100 GB transfer at that continuous rate would take about 29.63 minutes before overhead. These are arithmetic examples, not PulseLAN benchmark results or a promise about your equipment.

A wide range is a reason to repeat under better controlled conditions, not proof of interference. Compare the same device wired and wireless, then compare another client on the same path. A difference can involve the client, radio conditions, driver, power mode or server; the worksheet cannot distinguish them. Avoid replacing hardware based on one comparison.

4. Separate network capacity from file-copy performance

A generated-data throughput test avoids much of the storage work involved in copying a file. If that test is fast but copies are slow, check a local copy on the destination and then a second client before blaming the network. Many small files, encryption, security scanning, caches and disk performance can change the outcome. Keep units consistent: Mbps are megabits per second; MB/s are megabytes per second.

See a recorded experiment

Our SSH measurement notebook shows an actual run and why it does not establish a PulseLAN or Wi-Fi benchmark.

Sources and method

Our repeat-run worksheet is an editorial comparison aid, not an industry certification. The ESnet iperf3 reference explains reverse direction and parallel streams. Microsoft’s file-copy troubleshooting guide explains why storage and client comparisons matter. See LAN versus internet tests and our editorial standards.