Measuring very low currents - nA to uA

[Cicero]

I'm venturing into battery backed solutions, and was wondering how to actually measure such low currents of my (eventual) system?

I haven't started anything as yet, but hell, how is this even done?

Buy something like this maybe? https://www.adafruit.com/products/882

 

[koolguy]

Hello,
For measuring current there is shunt Resistor across it we measure Voltage, applying V=IR we get Current!

 

[Cicero]

Hello,
For measuring current there is shunt Resistor across it we measure Voltage, applying V=IR we get Current!

Dude, are you being serious?

Nanoamps to microamps brother.

 

[koolguy]

You can do parallel connection .
https://www.google.co.in/search?q=s...l5.3570j0j8&sourceid=chrome&es_sm=93&ie=UTF-8

 

[KeepItSimpleStupid]

I did measurements in the pA range using an electrometer. Usually, these sorts of measurements are done with a feedback ammeter. The other technique for measuring currents is the shunt ammeter(the voltage drop across a resistor). The box the OP pointed to is a shunt ammeter.

For the feedback ammeter, Ib and Vos are the dominant parameters. Vos is a strong function of temperature. I did design a 4-terminal I-V converter that had +-10V out in 4 decade ranges from 100 mA with bias and suppression.

If the is a uProcessor solution, you might be looking at an average current consumption. If the currents are tiny and you want an average, you can use coulomb meter and divide by time which I also did. The coulomb function is usually included in an electrometer.

 

[kubeek]

Do you mean measurement inside a built system, or measurement for evaluation?
I suggest you check out Dave Jonse's uCurrent for reference: https://www.eevblog.com/projects/ucurrent/ on the lowest settings it outputs 1mV/nA and the schematic is readily available.

 

[Cicero]

I did measurements in the pA range using an electrometer. Usually, these sorts of measurements are done with a feedback ammeter. The other technique for measuring currents is the shunt ammeter(the voltage drop across a resistor). The box the OP pointed to is a shunt ammeter.

For the feedback ammeter, Ib and Vos are the dominant parameters. Vos is a strong function of temperature. I did design a 4-terminal I-V converter that had +-10V out in 4 decade ranges from 100 mA with bias and suppression.

If the is a uProcessor solution, you might be looking at an average current consumption. If the currents are tiny and you want an average, you can use coulomb meter and divide by time which I also did. The coulomb function is usually included in an electrometer.

Awesome, this is what I was looking for. Will check it out. You're right, I am thinking more average current consumption of a microprocessor based system. Measuring sleep current consumption, and then active consumption etc.

Do you mean measurement inside a built system, or measurement for evaluation?
I suggest you check out Dave Jonse's uCurrent for reference: https://www.eevblog.com/projects/ucurrent/ on the lowest settings it outputs 1mV/nA and the schematic is readily available.

Thanks mate, measurement for evaluation. Also linked that same device in the OP

 

[KeepItSimpleStupid]

I need to say a few things:
1. A scope capable of averaging would be ideal.
2. RMS COULD work somewhat.
3. Be careful with supply caps and know what your really measuring. It might be better to break circuit after fillter cap and/or measure processor alone.
4. Without and special heroics, but a relatively complex design because of biasing, I managed 40 pA of offset with no means too null. Plan was to null with a D/A. 40 pA of DC offset was fine.
5. The output of an I-V converter is inverting. Vb-I*Rf. Vb is applied to non-inverting input.
6. A cap is usually placed across the FB resistor.
7. Shield and Gaurd concepts are introduced at low currents. Wire motion generates a current.
8. There are electrometer grade OP amps.
9. Low currents and conventional switches don't get along.

My goal was biasing within a mV; 4 terminal; Nominal bias 0.6 V and a trapezoid AC current waveform on top of a DC current.

 

[Mosaic]

Dude, are you being serious?

Nanoamps to microamps brother.

I backed this Kickstarter, although I have yet to put it to real use.
**broken link removed**

 

[kubeek]

third time's the charm Mosaic

 

[Cicero]

third time's the charm Mosaic

Hahaha

 

[DerStrom8]

I too was going to mention Dave's uCurrent solution, but looks like I was beaten to it (three times)