Measuring supply current in low power ATtiny84 application

I am experimenting with a low power battery design ATtiny84 and I am using a Pololu USB AVR ISP Programmer. Pololu USB AVR Programmer v2.1

I read Nick Gammon's excellent articles on the procedures to reduce the power: Gammon Forum : Electronics : Microprocessors : Power saving techniques for microprocessors and I now want to monitor the effects of the measures I take but I cannot measure currents below about 2 micro amp with my Fluke 79 II meter.

I looked at the device https://www.eevblog.com/product/ucurrentgold/ but it appears to be out of production and a bit expensive for me anyway. I have obtained, on loan, a Keithley 485 Picoammeter http://www.tunl.duke.edu/documents/public/electronics/Keithley/keithley_485_picoammeter_manual.pdf which appears to offer the same features as the ucurrentgold.

The problem I have is if I leave the Pololu ISP plug attached to my circuit the Keithley reads over range (whereas my Fluke does not) so I have to remove the ISP plug after each programme download which is a bit of a faf.

If anybody has experience of measuring this type of low current I would appreciate advice to remove the need to unplug the ISP plug each time.
Maybe I am not connecting the Picoammeter in the correct way, I have it in the negative supply line of my power supply, I tried it in the positive supply but it was no better.

so I have to remove the ISP plug after each programme download which is a bit of a faf.

Of course you do, unless you want to measure the current drained or supplied by the ISP.

However, if you want detailed help, you will have to provide more information about how you connected everything.

But consider that Nick Gammon's article showed that the data sheet is correct on power usage in the various modes. Do you expect your results to differ from what is in the data sheet?

Its simple, you just need a bit of wire and a standard multimeter, or build a switch thingy to make it easier.

Details here;

http://www.loratracker.uk/?p=843

You can supply through a forward biased diode and measure the voltage across the diode. You need to calibrate with known
currents and apply the ideal diode equation, but this will work down to below 1µA with a standard 10M
impedance DMM.

This has the advantage over a large-value shunt resistance in that the diode will handle much larger currents
too, but is more rough and ready.

OK all, thanks for the advice, I didn't want to blindly accept the data sheet values of current as my circuit has other components on it as well which will also effect the total current drawn.

I intend to make an isolating switch between my circuit and the Pololu ISP plug, a bit of a faf but I realize I would not need to do it often, only in the final stages of the design.

as my circuit has other components on it as well which will also effect the total current drawn.

Those currents can be measured separately, and added to the total.

This wont give you the exact current , but its good for comparing alternative a and b when testing.
Just connect a big cap(500uf+ something) and measure the time before it fails.
Not sure what you are making but this have been very helpful for me when i make my wireless sensors.
Best of luck

If you have control over the circuit, you shoudln't need to unplug the ISP every time or use a diode (diodes have reverse leakage current anyway that will throw off your reading). You just need to insert the ammeter into the circuit in the right place so that all of your project's current goes though it, and only your project's current. That means you can't just run the power supply line through the ammeter. You need to connect only the specific branches of your circuit that you want to measure to it, and make the other branches (like the ISP's power line) bypass it.

jremington:
But consider that Nick Gammon's article showed that the data sheet is correct on power usage in the various modes. Do you expect your results to differ from what is in the data sheet?

I can only speak for myself, but I sometimes make mistakes. If the power modes were important to my project, I'd want to verify that my project was going into the power mode I want, at the times I want it. If I got some detail wrong in the program or circuit design, it might not do that.

You must be pretty good if you can make your program function perfectly without any sort of testing or verification. :wink:

OK Jiggy-Ninja,

You just need to insert the ammeter into the circuit in the right place so that all of your project's current goes though it, and only your project's current.

That was a great suggestion.
The Pololu needs to be connected to the VCC of the processor to sense the voltage for programming and it was this which was drawing the extra current. By rearranging the power supply to my circuit so that the Picoameter was connected in the VCC connection to my ATtiny84 after the Pololu VCC connection solved the problem and allows me to measure the total circuit current with the Pololu still connected.
Also this method allows me to isolate the ATtiny84, and the other components on my board whilst leaving the Pololu ISP plugged in.
This is made easier because my circuit is built on a Proto Board.