I have an embedded system with a MCU, GPS module, GSM module and TI module. The project is largely depend on the total battery consumption. In order to characterize battery consumption, I have to measure current with respect to each module from the embedded PCB.

Does anyone know how to perform current measurements from the PCB with respect to the above mentioned modules?


Can you break the power connection to each module somehow? Generally current measurement will involve either inserting an ammeter (like the current mode on a multi meter) into the circuit between the source and the module power. Or if you can put a very low resistor between the source and the load you can measure the voltage and calculate the current.

Otherwise you could use a current probe for a scope but in that case you still need to break the connection and insert a little wire loop so you can put the probe around it. This is a nice way to see what your current profile really looks like. Same with the power resistor approach.

Usually you design in the ability to break the power connection to measure power. Also some new regulator controller parts from Linear will actually measure current for you and report it over a serial bus.

If you don't have the ability to break the current path you'll only be able to measure total system power.

----- Update to answer your questions below -------

Digital Mutli-Meter

Well first you don't have a DMM do you? Because just to start off that's going to be easier for you. Something like this:

Fluke 179

With that all you'll have to do is put it in series between your battery and your circuit board.


If you want to go the scope route you can but you'll need to order a 0.1 Ohm resistor. Make sure you get one that's the right Wattage, as in P = i^2 * R. So if you got yourself a 0.5 Watt 0.1 Ohm resistor you could run about 2 amps through it.

Now place that resistor in your circuit like so (ignore that I drew 0.01, I did that before I thought about how small your currents might be!):

enter image description here

Notice how you hook up the scope probes, you need to take a differential measurement, because you are trying to measure the voltage drop across the resistor. Normally I'd say use a differential probe but I'm guessing you don't have one. Instead you can clip the two GND leads together and then tie them to GND. Then place the first probe on one side of the resistor and the second on the other side. When you go to your scope you want to subtract those signals from each other. On an older scope that might mean you have to add and invert one channel. Alternatively if your scope is battery powered you might be able to float it (there's other more dangerous ways to float it too). I'd recommend the safer route though.

You chose that 0.1 Ohm resistor to make your life easier. Ohms law says V = I * R so if you draw 10mA your scope will move by 1mV. Hopefully you're not drawing only 10mA if you are you may need to get more serious (or use a bigger resistor to adjust your voltage output. Keep in mind your scope will pick up some noise as well so you need to play with that resistor till you find a happy point where you're getting more of your signal and less ambient noise.

Hope that helps

| improve this answer | |
  • \$\begingroup\$ I have the embedded pcb with me.Current probes are pretty expensive,so cant even think in those lines. Only option remaining is to use current sense resistor.When you say,breaking the power connection,what exactly do you mean? \$\endgroup\$ – john Oct 19 '12 at 2:59
  • \$\begingroup\$ He means, you need to have access to a trace you can cut to insert the current-sensing resistor. If all the power traces are on an inner layer, or something, you could have a hard time doing that. \$\endgroup\$ – The Photon Oct 19 '12 at 3:39
  • 1
    \$\begingroup\$ In that case, the best you can do is to find a way to shut off (or remove) the submodules one by one, while still operating the other modules in a "normal" way. Then if you measure the total system power under all those different conditions you can get an idea of how much power each module is responsible for. \$\endgroup\$ – The Photon Oct 19 '12 at 3:41
  • \$\begingroup\$ How do you find the total system power in this case? \$\endgroup\$ – john Oct 19 '12 at 16:34
  • \$\begingroup\$ Use one of the methods above but measure instead from between your source of power and your board. So maybe between your battery and the board. If you're using a DC adapter or bench supply to give it 5V or whatever you use just insert yourself between the supply and the board. \$\endgroup\$ – Some Hardware Guy Oct 19 '12 at 16:37

Another approach that can be used is to simply setup the unit with a current meter to measure the total draw from the battery. This setup is relatively simple to achieve because that connection is generally a single point and very localized.

Next evaluate the current draw when your unit is operating in its idle mode. Then adjust your software to place various modules in the circuit under differing operational duty cycles. Do this one module at a time whilst leaving the other modules in their idle state. Monitor the current changes for each duty cycle setting. After you have gathered all current readings you can directly see how changes in the operational behavior of your system change the total current draw. Using this data you can adjust the operating behavior of the system to achieve the battery energy usage profile that you need for the product.

Some times some surprising things can come out of an investigation like this. You may see that the increases in overall power for some modules hardly changes when their usage duty cycle is changed. On the other hand you may discover the need to add capabilities to your design to support dynamic power gating of certain modules to meet a power budget target. This may include realizing that you should be supporting a sleep and periodic wake up mode for the microcontroller as well.

One last comment. If you do design your board such that you can support power switching to various modules this adds a convenient single point at the switching FET where you can easily add in individual module current monitoring.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.