Measuring very low current

Question

I have asked here a month ago regarding measurement of very low current. I was told that Amper-Meters today has this resistance in series that will cause a voltage drop, hence you can't measure a very low current.

So what happened with my Fluke is a very strange thing. If I put it on the scale of 2mA, or even 20mA, and try to measure current in my PCB, I get many resets of the MCU because the voltage is dropping.

But I discovered something else, if I put the scale on the Ampere meter to be in 20A, then the PCB is working and I can measure the current (!).

---

So my questions are:

1. Why is it working only with a scale of 20A ?

2. Are the measurements good? Can you count on them? I can tell that the LCD shows a curent of 0.03 in this scale, that means its about 30uA? (that's what I expect)

Answer 1

Most meters will drop about 200-300mV at full scale on their current ranges. So the resistance will be higher on the lower current ranges.

As Wouter mentions you need to have good decoupling on the MCU so that when the current pulses the voltage drop caused by the meter is not excessive.

You may need to add more capacitance than you will need in your final application.

Also be aware that if you are supplying power from a battery their internal resistance will go up as they become discharged so failing with a meter in series may be an indication that you may have a problem even without the meter in the circuit as the battery discharges.

It is often difficult to measure the quiescent current of an MCU if the software is designed so that the MCU is in a low current state for much of the time with just occasional bursts of activity.

Depending upon the circuit you may need several hundred microfarads of capacitance on the power rail to supply the energy for these current pulses.

Answer 2

You can check your meters documentation, it most probably states the series resistance of the mA/A settings. It is most probably higher for the lower-current settings, and lowest for the 20A connection, which might be why this setting does not have a bad effect on your circuit.

That said, I doubt whether your problem is the voltage drop. Do you have a capacitor on the MCU side of the meter? An MCU will draw current ins pikes, which will not show on your meter (at best, they will be averaged). You need a capacitor at the MCU pins to supply this spiky currents.

Answer 3

You can use a very precicion resistor (0.1 or 0.01) in series with the circuit. After, verify the voltage over the resistor with an osciloscope. I hadnt tested this method yet... but I think it works.