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.

is this project based on MSP430 feasible? :)

Guru 18605 points
Other Parts Discussed in Thread: MSP430FR5969

Good morning,

I usually work with BLE and NFC so my technical knowledge about certains MSP430 is limited. We will be using MSP430FR5969 for our project.

1) Theoretically, if I run the MSP430 @ 1MHz, it should present approximately 100 uA of current consumption. Am I right?

2) However, when accessing code in FRAM, my application will require more current, up to a total of 370 uA for code execution. Am I right?

Both questions above consider the datasheet values, page 18.

3) Theoretically, I can set the MSP430 in LPM3 with the ACLK being feed by the 32 kHz oscillator. While in this mode, I want to do ADC conversions. I think I spotted the value in the datasheet for this, and I would require, according to the datasheet, around 180 uA (section apartat 5.12.5.4). Amb I right?

4) I have not been able to identify how much power I will require to write to FRAM the previous value from ADC. We want to do some log. Can anybody help providing this value?

5) Would the value in 4) be lower if DMA is employed? Is there a value for this situation? This way I would be able to calculate the energy savings, My plan is to do incremental DMA addressing, indicated as possible in the family user guide, page 330, to do the value log.

I'm asking all this because we wanna know if MSP430 powered by a harvesting source providing aprox. 110 uA will be able to run the code above. So final question:

6) What is your opinion? :)

Have the nicest day!

  • I did not verify the power values you got form the datasheet. Also, much of this depends on the exact circumstances. So the CPU power consumption @1MHz depends on the cache hit ratio. If you execute linear code, cache hit ratio is 0% and operating current can be as high as 400µA. the datasheet assumes 75% hit rate for 100µA, which means code with many small loops. Using LPM0 wherever possible will drop the current significantly. But much of it is the DCO, so entering LPM2 where possible will cause even more savings.

    I can say that writing to FRAM does not consume more energy than reading from FRAM, as any FRAM read is destructive and followed by a write. So executing code from FRAM is as much write as read. Writing to FRAM (by DMA) does not consume more than reading an instruction from FRAM. This should answer question 4.

    For the ADC operating current, don’t forget to add the operating current of the reference, when using a different reference than VCC.

    Now if you only power the MSP up occasionally and let it rest in LPM2 or deeper for most of the time while accumulating energy from the harvesting source, you should be able to do the project. But it won’t be easy with so tight supply margins.

**Attention** This is a public forum