An engineer sent me a screenshot: his Bluetooth module pulled 9 milliamps in sleep, and the datasheet promised 1.8. He’d designed the battery around the 1.8. The chip wasn’t broken. The number was just collected in a way he’d never run it.
The datasheet’s best day
That 1.8 milliamp figure comes from the radio off, the CPU at its lowest state, every peripheral clock gated, measured in a lab at twenty-five degrees. It’s real. It’s also a condition your product will hit for maybe two seconds a day. We stopped citing it to customers who were sizing batteries.
What you actually leave on
The Bluetooth module keeps a timer running, a watchdog, and usually a serial line that never fully sleeps because the host pokes it. Add those and sleep climbs to four or five milliamps before you’ve done anything wrong. The datasheet can’t list your firmware, so it doesn’t.
Temperature lies downward
At minus twenty, the same Bluetooth module needed eleven milliamps to stay alive, because the internal regulator works harder when it’s cold. The datasheet number was taken at room temperature. If your device lives in a freezer or a truck in winter, the spec is not your spec. We re-ran the test at the customer’s real low and tripled the battery.
How we measure now
We solder the Bluetooth module into the actual board, run the customer’s real firmware, and log current over a full day on a bench supply with a shunt. The graph that matters is the average, not the floor. One design averaged 6.3 milliamps over twenty-four hours, not the 1.8 anyone had planned around. We sent that number back before they ordered ten thousand cells.
What I ask before quoting a battery
What’s the real firmware doing at 3 a.m.? How cold does it get? Is the host awake most of the time? A Bluetooth module rated at 1.8 is honest about its floor, not your life. The gap between the two is where batteries die, and we’ve watched it happen more than once.
