When characterizing IoT sleep states, testing high-impedance sensors, or diagnosing PCB leakage, very small electric currents are measured in microamps (µA, 10-6 A), nanoamps (nA, 10-9 A), and picoamps (pA, 10-12 A). Standard multimeters often lack the resolution or the low-burden-voltage shunts required to read these values without crashing the device under test (DUT). For 90% of hobbyist and bench work—like measuring ESP32 deep sleep current or CMOS logic standby draw—you need a DMM with a dedicated µA jack and 0.1 µA resolution. For sub-microamp leakage, you must step up to a dedicated picoammeter.
The Scale of Low-Current Measurement
To understand why standard meter ranges fail here, look at the math of battery life. A CR2032 coin cell has a capacity of roughly 220 mAh. If your circuit draws 10 mA, it dies in 22 hours. If it draws 10 µA, it runs for 2.5 years. Measuring that 10 µA draw requires a meter that can resolve down to 0.1 µA without introducing a massive voltage drop.
Current measurement at this scale is governed by three distinct tiers:
- Microamps (µA): Microcontroller sleep modes, LED indicator bleeds, and bias currents for standard op-amps.
- Nanoamps (nA): Photodiode dark currents, reverse-biased Schottky diode leakage, and CMOS gate standby currents.
- Picoamps (pA): Piezoelectric sensor outputs, ionization chamber readings, and high-value resistor network leakage (100MΩ+).
Meter Setup and Probe Placement for µA and nA
Measuring current requires breaking the circuit and inserting the meter in series. At the microamp level, how you insert the meter matters just as much as the meter itself.
Meter Setup Block
- Dial Position: Set to µA or mA (do not rely on Auto-Range if your meter defaults to the 10A shunt first; the 10A shunt has a low resolution and high burden voltage).
- Lead Jacks: Black lead to COM. Red lead to the dedicated µA/mA jack (never the 10A jack for low-current work).
- Range Selection: Manually set to the 200 µA or 2 mA range to force the meter to use its highest-sensitivity shunt resistor immediately upon connection.
Probe Placement Procedure
- Power Down: De-energize the DUT. Disconnect the battery or USB power.
- Break the Circuit: Lift the VCC pin of the IC, desolder a jumper wire, or unplug the positive battery lead. Always break the high side (VCC) rather than the low side (GND) to avoid ground-loop errors if your test setup shares a common ground with oscilloscopes or bench supplies.
- Insert Probes: Place the Red probe on the power source side of the break. Place the Black probe on the DUT side of the break.
- Power Up and Settle: Re-apply power. Wait 5–10 seconds for decoupling capacitors to charge and the DUT to enter its target state (e.g., deep sleep) before recording the reading.
Expected Readings and Parasitic Mistakes
Knowing what a 'good' reading looks like prevents you from chasing ghosts. Below is a reference table for common low-current test points.
| Test Point / Component | Expected 'Good' Reading | 'Bad' Reading (Indicates Fault) | Common Culprit for Bad Reading |
|---|---|---|---|
| ESP32 Deep Sleep (No peripherals) | 10 µA – 150 µA | > 2.0 mA | GPIO pins left floating or pulled high externally; USB-to-UART chip awake. |
| CMOS Logic IC (e.g., 74HC series standby) | < 1.0 µA | > 50 µA | Input pins floating (causing internal shoot-through current). |
| Reverse-Biased Schottky Diode (e.g., BAT54) | < 50 nA (at 25°C) | > 1.0 µA | Thermal runaway, damaged junction, or dirty PCB flux creating a parallel path. |
| Op-Amp Input Bias Current (e.g., LM358) | 20 nA – 50 nA | > 200 nA | ESD damage to input stage or common-mode voltage violation. |
Mistakes That Give Misleading Readings
If your readings are erratic or the DUT keeps resetting, you are likely falling victim to one of these three bench errors:
- Burden Voltage Brownouts: Every ammeter uses an internal shunt resistor to measure current. On the µA range, this resistor might be 10Ω to 100Ω. If your circuit draws 500 µA across a 100Ω shunt, the meter drops 50 mV. If your DUT is running on a 3.0V coin cell and requires 2.9V to stay out of brownout, this voltage drop will crash the microcontroller, causing the current to drop to zero, the voltage to recover, and the MCU to reboot in an endless loop. Fix: Use a meter with a lower burden voltage, or power the DUT from a bench supply set 100 mV higher than nominal. For a deep dive on this, see the All About Circuits guide on ammeter burden.
- Triboelectric Noise: When measuring in the nA or pA range, simply moving the test cables generates static charge via the triboelectric effect inside the cable dielectric. Fix: Use rigid, low-noise triaxial cables and let the setup sit completely still for 30 seconds before logging data.
- Surface Leakage Paths: Finger oils, humidity, and leftover RMA solder flux create high-impedance parallel resistors across your test points. A 100 MΩ flux path across a 3.3V rail will leak 33 nA, completely ruining a nanoamp measurement. Fix: Scrub the PCB with 99% isopropyl alcohol and a stiff brush, then bake at 50°C for an hour before testing.
Low-current measurements are almost exclusively performed on low-voltage DC systems (CAT I or CAT II environments, like PCBs, battery packs, and 12V/24V automotive). Never probe mains voltage (120V/240V AC) with the red lead in the µA/mA jack. The µA jack is protected by a low-interrupting-capacity fuse (often 0.5A or 1A). Hitting mains voltage will instantly vaporize the fuse, and if the meter lacks proper internal blast shields, it can cause an arc flash. Always verify your meter's CAT rating and fuse specifications before testing, and physically move the red lead back to the V/Ω jack when finished with current tests.
Decision Tree: Choosing the Right Low-Current Tool
Don't guess which tool to buy. Use this decision matrix to select the exact instrument for your required resolution and budget.
| If your target current is... | And your application is... | Buy this exact tool (Concrete Pick) | Why this pick? |
|---|---|---|---|
| > 1.0 µA | IoT sleep current, hobby robotics, general bench work (Budget < $300) | Brymen BM859s | Offers 0.01 µA (10 nA) resolution on the µA range with a very low burden voltage. Best value for makers. |
| > 1.0 µA | Industrial troubleshooting, harsh environments, requires CAT III safety (Budget < $500) | Fluke 87V | Industry standard. 0.1 µA resolution. Rugged, reliable, and holds calibration in the field. |
| 1 nA to 1 µA | Photodiode dark current, sensor leakage, material testing (Budget > $2,000) | Keithley 6485 Picoammeter | Dedicated 5.5-digit picoammeter. Resolves down to 10 fA. Eliminates DMM shunt-switching glitches. |
| < 1 nA | Piezo sensors, ionization chambers, high-impedance electrometry (Budget > $5,000) | Keithley 6517B Electrometer | Includes built-in voltage source for I-V sweeps and triaxial connections to guard against leakage. |
If you only own a basic $20 DMM but need to measure 5 µA without burden voltage crashes, skip the ammeter function entirely. Place a precise 10.0Ω shunt resistor in series with the VCC line. Use your DMM's DC Volts function (which has 10MΩ input impedance) to measure the voltage drop across the resistor. Apply Ohm's Law (I = V / R): a reading of 50 µV across a 10Ω resistor equals exactly 5 µA. The burden voltage is limited to a harmless 50 µV, keeping your DUT happy.
Summary and Next Steps
Measuring very small electric currents requires respecting the physics of the measurement loop. By selecting the correct µA jack, breaking the high-side VCC path, mitigating burden voltage, and cleaning your PCB of parasitic flux paths, you can confidently characterize nanoamp and microamp circuits. For sleep-current debugging on modern microcontrollers, pick up the Brymen BM859s; for serious lab-grade leakage analysis, invest in the Keithley 6485. Always return your red lead to the voltage jack when finished to protect your meter—and yourself—from the next high-voltage test.
