Instability in aluminium housing
While developing a device based around LoPy (development version, not OEM) I noticed very interesting behavior. LoPy is installed on custom developed carrier board which is fixed within aluminium housing.
Some of the LoPy boards I tried crash within few seconds after booting if WiFi AP is initialized. I believe some sort of resonance/standing wave is occurring within the aluminium housing saturating RF frontend of WiFi interface. I don't understand why is it causing complete crash of entire micro-controller. Watchdog will trigger reset but it will get stuck directly in the boot phase. If I remove one panel of the housing the instability disappears.
Other LoPy boards which will not crash during first few seconds after boot seems to be able to run long-term.
Internal antenna was used. The purpose of WiFi interface was to provide web configuration interface accessed from close proximity to the device and it worked surprisingly well with internal antenna transmitting through the aluminium housing. In case of LoPy boards which did not crash of course :).
Do you use internal or external wifi antenna?