Current consumption measurement

Status
Not open for further replies.

EngIntoHW

Member
Hi,
I'd appreciate your inputs

I need to measure current consumption of a battery operated embedded device.
The device is operated by software, I use an application (GUI) to send it commands to perform various tasks.
The current measurement is done by a Fluke at the output of the battery.

Therefore it is direct current (not AC) that is measured.

My idea is to write a list of all the different tasks and states of the device, and a diagram that shows the transitions between the tasks and states.
Then measure the current at each transition, each task and each state.

Does it sound good to you?

Thanks a lot
 
Which fluke and what is the voltage and current range you expect to measure. A fluke meter in amps mode can give significant errors for low voltage, high current measurements.

 
The problem can be if the current varies a lot. If anything is pulsed, you may have trouble measuring the average current.

I recently has to measure total battery use for a device with various modes. The current was sampled 8 times a second, the average of 80 samples was recorded every 10 seconds, and that went on for over 2 day.

That might have been more than was needed, but in most modes the a Fluke was giving a very variable reading.

Also I used a sensing circuit that had virtually no voltage drop (at the expense of needing power to run the sensing circuit)
 

My issue is that it's fairly unusual to get any sensible current readings on a meter for most electronic devices - the current consumption commonly changes all the time, often over a large range. A better option would probably be an old mechanical analogue meter, which would tend to 'average' it out somewhat.

Or better still, monitor the current with a storage scope, and get a graph of how the current changes.
 

your method will work fine if you can adjust the code to monitor repeated measurements of semsor1 to get a good reading of amperage. Then repeat for each task (keep repeating task many times for several seconds so you can ge a solid reading. Repeat as needed.

then essentially calculate, as you suggest, the time for each sensor as you cycle through your instructions. And then make totals of timexAmps to get average mA/second over a whole cycle (including idle/standby time). Again, it all assumes your code is accessible and repeated measurements are a close approximation of actual circuit usage.
Good luck.
 
This **broken link removed** article shows how to calculate power with a scope.
Maybe you could set up scope triggers to capture a "cycle" of a particular task. You would have to integrate the results.

It's not too difficult to make a ZRA (Zero Resistance Ammeter) or Feedback ammeter which drops less than a mV when measuring currents. I made a feedback ammeter that was capable to +-100 mA and had 4 lower ranges. The output was +-10V at full scale.

You could possibly build an energy meter.

Another possibility is a battery Gas gauge IC https://www.ti.com/power-management/battery-management/fuel-gauges/overview.html

Once you had a "gas guage" in place, you could then create "typical" energy use.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…