Battery-Powered Node Brownouts Under Motor Start
A remote monitoring node runs from a 12 V 7 Ah sealed lead-acid battery through a buck converter set to 5 V. The node's radio draws 1.8 A peaks for 40 ms during transmit; idle draw is 35 mA. A pump on the same battery bus draws 4 A for 2 s at start. During pump starts the node resets, and the 5 V rail dips to 3.9 V on a scope with a 20 MHz bandwidth limit. The buck converter is rated 2 A continuous with no specified transient headroom, and the battery is two years old with an unknown state of health. The site is 40 km from the workshop.
1. Diagnose the reset and propose the most reliable on-site remedy, justifying the choice against alternatives.
Read the worked answer
The reset is a supply-rail brownout, not a firmware fault: the pump start pulls the battery terminal voltage down through its internal resistance and cabling, and the 2 A buck has no headroom for the 1.8 A radio burst on top of idle and pump loading. Options: (1) add bulk capacitance at the 5 V rail, which helps the 40 ms burst but does nothing for a 2 s pump sag; (2) fit a larger buck with transient headroom and a low-ESR input capacitor bank; (3) separate the radio supply with its own regulator and a diode-OR from a small local supercapacitor; (4) replace the battery. The most reliable single action is to give the radio a dedicated regulator fed through a Schottky diode from a supercapacitor, so radio bursts are decoupled from the pump. Also measure battery internal resistance and replace it if it exceeds the manufacturer's limit, since a two-year-old battery at an unknown state of health may be the root cause.
2. State three measurements to take on the next site visit and what each would prove.
Read the worked answer
- Battery terminal voltage during a pump start, logged at the battery posts, proves whether the sag originates in the battery or in the cabling (compare with voltage at the buck input). 2) Battery internal resistance via a step-load test, proving whether the battery needs replacement. 3) 5 V rail with a wide-bandwidth probe (no 20 MHz limit) to see fast transients the limited scope may be hiding. Optionally, radio current profile to confirm the 1.8 A figure. Each measurement isolates one element of the power chain so the fix targets the true cause rather than a symptom.