• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 3 answers
  • Subscribers 80 subscribers
  • Views 5182 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

MSP430FR6989: What would you recommend to measure the current from the MSP430 microcontrollers and other?

Sara Souza
Expert 1450 points
Part Number: MSP430FR6989
Other Parts Discussed in Thread: ADS1224, , INA233, INA223, ENERGYTRACE

Hi, thanks for coming help!

I am looking for a solution to measure energy of low power embedded systems, including MSP430 mcus.

I know about ETT, but it is not a nice option for me because it only gives me graphical results, and I actually need numerical results in order to manipulate.

Any suggestions?

For while, I am thinking about use an OpAmp with a shunt resistor and a ADS1224.

  • 0
    TI__Intellectual 1370 points

    Hi Sara,

    What calculations are you hoping to do with the data? The Energy Trace tool in CCS should be able to give numerical average current values for your application. Depending on the accuracy needed for your calculations, you could also use the cursors on the graphical results of Energy Trace to get rough numerical power and time values from the plotted data. All the options for Energy Trace are described in the CCS User's Guide.

    There is also test equipment available that can act as a power source to your board and simultaneously take high precision measurements of voltage and current.

    Do you want to take measurements for benchmarking your design or do you want to be able to take measurements in your finalized system? For a lower cost option or solution that is embedded directly into your system, you could use a shunt resistor and amplifier as you said. Depending on the precision of the measurements that you need, you could directly use the MSP430FR6989 ADC to take differential voltage measurements and store them in FRAM to readout later or output directly to another device. This would increase current consumption of your system, however. 

    For higher resolution measurements, you could use the ADS1224 as you said, but you would need to output the data to another device as samples are taken because the ADS1224 has no built-in memory to store data.

    If you want the measurement tool to be separated from your system, you can use another MSP430 device with a built-in ADC to gather and store data. MSP430 parts can be compared here.

    Regards,

    Ryan

  • 0
    Genius 5435 points

    Hi Sara,


    I suggest to look at TI's INA233. It has 16-bit accuracy, voltage, current, power, corresponding accumulators. It can measure and accumulate power consumption even when your device is in reset or power down mode.

    Regards,

    Alexey

  • 0 in reply to Ryan Meredith
    Expert 1450 points

    Hi Ryan,

    1. I hope to get energy consumption information in order to compare codes execution consumption, make tests, predict how long a battery or other supply source is going to last , scheduling, and other variables that demands energy information.
    About the numbers,  I can study CCS's User Guide for checking it, but It would limit my measurement to TI's devices supported by ETT. And our goal it to have a tool for different MCUs.

    2. What test equipments do you refer?

    3. We want to have an isolated system to measure energy, even because an integrated measuring systems would also waste energy, then the results would be affected by the own system.

    First I thought about using my own MSP4306989 and its differential ADC as you said to read a shunt and send me to a PC via serial (where it will finally be used) . The configuration I thought was this one I saw in a paper (It's attached), but I have just discovered I couldn't use the differential ADC from the MSP430fr6989 because of 2 things: First because the Vin+ would be bigger than the Vref+ because of the use of the OpAmp, and second because in the differential mode I would loose the bits resolution in half (because of the negative and positive values).

    4. For while I thought about using the ADC1224 with some interface to send it for a PC.

    5. I liked the idea from Alexey, but is INA223  able to measure low currents (those at the range of LPM modes) ? From what I saw it seems also to use shunt resistor, ok?

    Best Regards,

    Sara MonteiroVol8No1_A8.pdf

  • 0 in reply to Alexey Bagaev
    Expert 1450 points
    I enjoyed the idea , but is INA223 able to measure low currents (those at the range of LPM modes) ? From what I saw it seems also to use shunt resistor, is it right?

    best regards,

    Sara
  • 0 in reply to Sara Souza
    Genius 5435 points
    Let to be correct, I was talking about INA233, and not about INA223. The value of shunt resistor can be selected my Ohm's law. So, for low maximum currents (like 100mA) it can be even 1 Ohm. The shunt probe can be even on the input of voltage regulator, so even power losses in conversion can be included to measurements.

    Alexey
  • 0 in reply to Sara Souza
    Genius 5435 points
    Your scheme is wrong. Take a course for analog amplifiers first. It is available from TI for free.
  • 0 in reply to Alexey Bagaev
    Expert 1450 points

    Actually this is not my scheme. It was suggested in a paper that was published at IEEE journal. (Attached).

    But thanks anyway!

    4101.Vol8No1_A8.pdf

  • 0 in reply to Sara Souza
    Genius 5435 points
    IEEE is a very respected organization, I know. And it still didn't change my point of view about that diagram. Nor by this document, nor by IEEE's reputation. But let I keep my thoughts with me in respect to IEEE. :)
  • 0
    Prodigy 50 points

    Hi Sara,
    I had similar problem in my last low power project. Very low current (few uA) most time and regular and unregular current peaks (about 10mA). Determining number of years of battery life was difficult to calculate or to measure using typical measurement instruments. Finally I tried EnergyTrace supported by any of MSP430FR launchpad (look for promotional one for $4,30 ;)
    I wrote my own application in Visual Studio C# that commucates with EnergyTrace module via msp430.dll. My application gets measured energy value for defined time period and transforms it to average current in that period. Result is accurate enough to my needs (TI says 2%+0.5uA). Teoretical resolution is 0.1uA. I use periods adequate to what I need to measure and simulate (sometimes 2 seconds, sometimes 30 second). This module measures energy not current so it is accurate.
    EnergyTrace module doesn't have to be connected to destination microcontroller programming interface. It's enough to connect VCC and GND only, so we can test not only dedicated microcontrollers. In the future I plan to make this application more complex (graphs, charts etc).

  • 0 in reply to Andrzej Sierzega
    Expert 1450 points
    Hi Andrzej!

    Cool! I've tried a code in C avaiable from Carrot Industry, they used a library from TI also, but I just got results from one MCU using also a specific board.
    Is your code avaiable or is it private? I'll search a little more about msp430.dll.
    Thanks

    best regards!
  • 0 in reply to Sara Souza
    Prodigy 50 points

    Hi Sara!

    I plan to make this application as non-profit open source code. Some improvements are needed before and also there are TI law properties to msp430.dll used in the project.

    If you are interested in my C# source code please contact me on private. Also you must know that this is very beta version and it uses windows console.

**Attention** This is a public forum