Measuring firmware power draw without a lab bench
A contributor's method for getting usable sleep-current measurements from battery-powered firmware using commonly available equipment.
- Author
- A. ContributorSample external contributor record
- Published
- Reading time
- 1 min
Sleep current determines battery life in most low-power designs, and it is the measurement people skip, because the equipment usually recommended for it costs more than the rest of the project. A multimeter in series will not do it: the burden voltage at microamp ranges is enough to reset the device under test, and the meter's sampling rate hides the wake-up spikes entirely.
Shunt and oscilloscope
A 10-ohm shunt on the low side with an oscilloscope across it gets you most of the way. It gives a continuous trace rather than an average, which is what you actually need — the average is dominated by whichever transient you cannot see.
- Place the shunt between the device ground and supply ground, not in the high side.
- Set the scope to DC coupling and the smallest vertical range that does not clip the wake transient.
- Trigger on the rising edge of the wake event and capture at least one full duty cycle.
- Integrate the trace over the cycle rather than reading a peak.
Related
Analysis
What edge runtimes actually change about request handling
A practical account of the constraints edge runtimes impose, and which familiar server-side assumptions stop holding when you move there.
Build DocumentationPrototype
Building a mains power quality logger for unstable grid supply
An isolated voltage and frequency logger built to characterise supply quality at sites where equipment was failing without explanation.