🛠 Profiler la consommation : µA réels, DC/DC et Power Profiler Kit

Why the nRF52833-DK is a measurement bench

The kit's User Guide documents dedicated current measurement pins: P22 (VDD domain) and P23 (VDDH domain), with solder bridges SB40/SB41 to cut and 10 Ω series resistor footprints — the interface intended for Nordic Power Profiler Kit / PPK II. The Nordic current measurement guide actually uses this DK as its reference platform. Add the SW6 "nRF ONLY" switch, which isolates the J-Link OB interface MCU — without it, you measure the debugger's consumption, not the SoC's — and the SW9 switch, which selects the power source (VDD / Li-Po / USB).

Reference figures (Product Spec v1.7, 3 V)

• 0.6 µA in System OFF (RAM retention disabled);
• 1.5 µA in System ON with RTC wake-up;
• 4.9 mA peak during transmission (0 dBm), 4.6 mA during reception;
• 52 µA/MHz in CoreMark from flash — at 64 MHz, this indicates the CPU's cost.

The SoC integrates DC/DC + LDO regulators (1.7–5.5 V supply): enabling the DC/DC in the firmware is often the first saving on an intensive radio profile. Operating temperature −40 to +105 °C: the device is qualified well beyond a standard application MCU.

Step 1 — Prepare the kit

1. Cut the SB40/SB41 solder bridges (or use pins P22/P23 with probes) to insert the ammeter in series on the desired domain — VDD for core/logic, VDDH if you are powering at the upper end of the range.
2. Set SW6 to nRF ONLY to exclude the J-Link OB interface from the measurement.
3. Disconnect the serial terminal: the User Guide (§9) warns that measurements are inaccurate as long as the USB-to-serial converter is connected.
4. Power source: USB 5 V, Li-Po connector J6/P27 (2.5–5.0 V), VIN on P20 (3–5 V), external supply on P21 (1.7–3.6 V), or the CR2032 holder included in the box — handy for validating a real battery budget.

Step 2 — Connect the PPK II

The PPK II replaces the power supply with its programmable source and continuously samples the current (up to 100 kS/s), with CSV export to integrate the profile into average load calculations. Without a PPK: a µA-meter on the measurement pins is sufficient for idle points (System OFF / ON+RTC), but not for capturing radio peaks lasting a few ms.

Step 3 — Build a useful profile

Measure in this order: (1) System OFF, (2) System ON idle with RTC wake-up, (3) full radio cycle — advertising at the target interval, connection, exchanges — while adjusting the 4 power levels (−20 to +8 dBm in 4 dB steps) and data rates (2 Mbps vs 125 kbps Long Range: fewer retries, shorter peaks). The key formula: µA·s per cycle × number of cycles = Ah/day.

Pitfalls

• Forgetting to set SW6 to the interface position → you measure ~mA instead of ~µA and draw incorrect conclusions.
• Serial terminal connected → distorted measurements (documented).
• The onboard 5 V boost and its reverse-protection diode cause a voltage drop: with a Li-Po battery, also measure under the battery's actual voltage.
• Silicon anomalies (errata) can affect the RTC — a drifting wake-up invalidates the entire budget; see the errata tutorial.