accurately measuring integrated power consumption

Win, is this something that has to live in the field or on you lab bench.

If it's on the lab bench then some A/D converter that plugs into your computer. Like a labjack.

George H.

I forgot to mention LoRa transmitter currents. We'd like to setup the instrument, and make a measurement over a complete cycle of activity.

On the bench. Thanks for the suggestion, George, but with rapidly-changing, and pulsing currents, we need an integrating ADC, rather than SAR type, with millions of samples, some missing the pulses.

Measuring power in circuits with low frequency crest factors requires filtering of the rms measurement values. Something like a Yokogawa will give a wideband analog power-related output which can be filtered and logged. Data that doesn't filter will risk aliasing and spurious error.

All analog processing adds another layer of tolerance and error - but avoids aliasing, or picket fenceing, which seems to be what you're concerned about.

RL

An older Fluke 8922a offers analog output, but is hampered by autoranging, if engaged. Autoranging is likely to be useless when you're looking at LF measurements with large crest factor - the errors of the largest measurement range needed without clipping will dominate the average.

RL

On a sunny day (31 May 2019 07:00:23 -0700) it happened Winfield Hill wrote in :

Video FLASH ADC record to disk, process later? Or just add a PIC to that ADC do the processing in real time.

That is how I would do it...

Not clear if you need measurements over very short intervals, or just how much it average out over a day. If it's more along the lines of the latter, how about a Kill-a-Watt meter? Depends on how accurate you need it and over what interval.

Is the supply voltage constant? If so you just need to measure average current. Seems like any DVM that you can acquire samples of would work; suck out a string of measurements and average. Maybe lowpass the current sensor to improve your statistics.

Well just an integrator.. low pass filter. I guess that's obvious, so you must want to see the peak heights too?

GH

ISTR a flyer from Keysight about dedicated instruments for measuring such PSU currents.

Googling for "dynamic current measurements" yielded

"For IoT applications, you can measure the current profile of low power device including both sleep and active states with up to 100 dB wide dynamic range, without having to use multiple instruments or make multiple measurements.

"For mobile applications, the wide bandwidth and low noise performance enable current profile measurements of PDN (Power Delivery Network or Power Distribution Network) including both large inrush current and small sleep current with the same instrument."

You need to borrow a coulometer from an electrochemist :-). Usual mode is to set a constant output voltage and then integrate the current. Some brief googling turned up this general purpose machine as an example:

Cheapest and fastest would be an eval board for one of numerous coulomb cou nter ICs. They all seem to use a voltage-frequency-converter (VFC) with eac h pulse representing a discrete unit of charge, intended to interrupt an ex ternal uC for counting. The eval board software looks good, time stamping t he VFC pulses, storing, displaying a running total, communicating with the board via USB. Th most expensive component required is a cheap laptop. The coulomb counters seem to be limited to around 60dB dynamic range.

Power it from a supercap and measure the before and after voltages!

Ones that I tried had self-discharge time constants of months.

Sounds good, have a part number, to get me started?

The eval board and software for the DS2740K looks like the simplest and most self-contained, but as usual Maxim fails on ordering information. Analog sells the eval board for their LTC4150 as p/n DC756 in stock at $50, but it's an I2C interface requiring a DC590B USB contoller for another $50 with no availability until July.

You can use the Silabs evaluation board energy profiler:

Cheers

Klaus

This one seems more advanced and easier to use with any target:

Cheers

Klaus

What's wrong with the coulomb meter in the book?

You don't need a coulomb meter, use 2 electrodes in suitable solution. Yes, it's old school wetware, but you've probably got the bits in stock & can do it now.

NT