There has been a great deal of value to me in my involvement with the project. Would like to share it with the family planning gold coast team so they can also read it and implement something new.

Just a quick question about hookup. Is there any risk of damage to the device if it is hooked up the wrong way around by accident? I.e. connect the load to the OUT jacks?

I haven't tried that :-)
But I would not recommend it. I am sure it can be damaged that way.

Noted :)

Just received my CurrentRanger. Great job on the board and firmware. Everything feels pretty schmick.

The only downside (but I'll change this in the firmware myself) is that you can't play Mary had a Little Lamb straight away since the buzzer doesn't sound on a button when it's active and you press it again.

Thanks for the feedback, Hey maybe I should add a little MENU where you can turn "always buzz" ON-OFF to get your 3 key piano (-:

I'm also interested as Dan is in having some kind of GPIO output to indicate range. It looks like, from your reply, you broke out the "SPI+D3" header for nothing in particular, but just because the IO was there and a user might come up with something for it?

I'm imagining I'll implement a version of the firmware that outputs a range indication on each of those pins. For instance, pin 1 high in nA range, pin 2 high in uA range, and pin 3 high in mA range. That way, on a mixed signal scope (I'm thinking Digilent Analog Discovery), I can set up a Math trace that is something like (A1*D1*1e-9)+(A1*D2*1e-6)+(A1*D3*1e-3)=current (Amps). In so doing, I can then use whatever waveform analysis to do the integration/cursors/etc.

Are you aware of anyone having done this work already, or have you done it yourself? For measuring things like microcontrollers swinging from sleep to LED lightshow, it seems like it'd be the most useful option on the market. There are a few devices ranging from attainable (Atmel atpowerdebugger, SiLabs STK "Advanced Energy Monitor"/Energy Profiler) to unobtainable (I think J-Link makes a J-Trace with energy profiling that costs $$$), but those devices still don't have the dynamic range of the CurrentRanger - the atpowerdebugger has two separate channels, one for 100mA and one for 1A, so you MIGHT be able to use that depending on your target hardware but I'm not sure if you can, say, put the high-drain stuff on one channel and the MCU on the other, simultaneously. But I know the AEM on the SiLabs EFM32 STK boards tops out at 100mA. 

It'd be killer to have a solution that does (nA-A)-to-trace for easy post analysis. 

Sorry, the SiLabs AEM actually does .1uA to 50mA, so a bit more limiting. Great dynamic range, but not enough to precisely profile an entire system. 

Alex,

That GPIO header (it's actually the SPI pins) is left there intentionally to offer ways of experimentation and augmentation to folks like you who don't fear to try fiddling with coding and extending firmware functionality. I'm not aware of anyone doing anything yet with the GPIO. But always a good idea to check in the to see what current users are up to, or any neat customizations.

https://lowpowerlab.com/forum/currentranger/

Ahh, a forum link is JUST what I was hoping for, I suppose. Thanks!

Interesting point, I haven't taken that into account.

Is this a situation where somehow the potential of the board vs the potential of external load are very different, and some inrush current causes the board MCU to see a brown-out condition?

Very neat.  This looks like a useful device, Felix.  Instruments like this are definitely useful when designing ultra-low power devices.  I took a look at your schematic and a brief look at the code (but not in depth) and have a couple of questions.

1. How fast are you sampling the ADC when autoranging?  Is it just as fast as the main loop executes (and then 1024 averaged samples at a time?)?

2. In your own testing would it ever have been useful to also output the auto-range to the scope (e.g. did you do testing where the power levels changed enough - say a micro sleeping to waking and then keying on a RF output stage - that during a trace capture the ranges changed)?

Also have you ever experimented with connecting a power supply sense to the output of the current monitor as a way to avoid voltage droops (and perhaps more accurately capture power consumption by keeping a steady input voltage)?

Thank you, the "official" firmware ADC is a middle of the road approach - not the fastest possible. It gives a reasonably accurate reading for a 12bit, and it allows reasonably fast switching.

You will see the switching on the scope by very short spikes. If you need digital signalling that's possible by the extra broken out GPIO headers (SPI+D3) which could also be used with SPI devices for whatever purpose.

I guess I don't really understand your last question - but I don't think I tried that.

Thank you for the information.  The third question was a way to deal with the voltage drop that occurs across the sense resistor(s).  It probably isn't too big of an issue for a lot of situations but I have seen micro-controllers execute a brown-out reset while I am measuring their input current using a shunt resistor and they suddenly require a lot of current.  One solution, which I haven't tried, would be to see if connecting a lab power supply remote sense positive input on the output side of the shunt resistor would deal with that.  Not really directly related to your project but I wondered if you'd ever experimented with that since you have done a lot of current monitoring.