Josh Huang's Engineering Notes
The Scan Result That Looked Like Proof and Wasn't
A wireless controller kept disconnecting on the same in-game trigger, and a live network scan seemed to hand over the smoking gun — until the identical channel numbers and suspiciously sequential MAC addresses gave away that the scan data itself was fabricated.
This article is also available in Traditional Chinese: 中文版 — same content, just a language difference.
A wireless game controller reliably dropped to keyboard/mouse fallback mode after one specific in-game action, every time. Two physically-reasonable theories got chased and each made things no better or actively worse. Then a network scan turned up a strikingly clean list of "congested" nearby Wi-Fi networks — clean enough that it turned out to be the wrong kind of evidence entirely.
Two reasonable hardware theories, both wrong
The first theory was 2.4GHz Wi-Fi interference — plausible, since the controller's wireless dongle and the home router both operate in the same crowded band, and the disconnect was consistent enough to look environmental. Moving the dongle further from the router changed nothing. The second theory was a USB bus contention issue — also plausible on a machine with several USB devices sharing controllers — so the dongle was moved to a physically isolated USB bus. The disconnects didn't just persist, they got measurably worse, which at least ruled out "not enough USB bandwidth" as directly as a theory can be ruled out.
Unlock this article to keep reading, or subscribe for unlimited access to everything. See Pricing for details.