1
\$\begingroup\$

As we know that, everything is marketed as Low Power or Ultra low power. I understood that we can classify the controllers based on their application and clock cycle for the above mentioned sections. But what makes me little confused is what contributes to these facts of Low Power and Ultra Low Power consumption because the application basically demands the average power consumption. Then how MCU can be classified into these categories. Speaking from architectural point, are there any significant differences? I know it is a very basic question and can get points from Internet, but I didn't find sufficient explanation towards a customers point of view. It would be helpful if experts can give me some ideas about difference between Low Power & Ultra Low Power designs.

\$\endgroup\$

closed as primarily opinion-based by pipe, Voltage Spike, Dave Tweed Sep 30 '17 at 14:30

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Low and ultra-low power are usually designate MCUs that are capable of low power modes such as "sleep" or "deep sleep" modes. \$\endgroup\$ – Eugene Sh. Sep 14 '17 at 15:16
  • \$\begingroup\$ @EugeneSh. Please avoid answering questions in comments. \$\endgroup\$ – pipe Sep 14 '17 at 15:25
  • 3
    \$\begingroup\$ @pipe Not sure if this comment is comprehensive enough to be an answer. \$\endgroup\$ – Eugene Sh. Sep 14 '17 at 15:27
  • 1
    \$\begingroup\$ @EugeneSh. It is, it's just not a good answer, and would be downvoted if you tried to pass it on as one. By writing it as a comment, you get your chance to "I'll just put my 2 cents here" without anyone being able to vote on it or accept it as an answer that solved the problem. Stack Exchange was designed to avoid such comments. \$\endgroup\$ – pipe Sep 14 '17 at 17:59
5
\$\begingroup\$

The power consumed by an application/device can be divided in aspects:

a) the power needed to do the required amount of computation

b) the power required by the chip in its lowest power mode (sleep, deep sleep, hibernation, etc.)

c) the power required by the chip while doing the barest minimum of activities

Depending on the application, one or more of these aspects will dominate the overall power consumption. Low power chips generally have a very low power mode, reducing the power for aspect b), and ways to have the hardware take care of c) (responding to keypress, timeout, A/D conversion completed, etc.) without involving the CPU (thus reducing power).

Low Power or Ultra Low Power are just marketing terms without any agreed-upon kwantitative meaning. A manufacturer has to come up with a new term for the chips that do better than its previous range of Low Power chips... But an engineer doesn't care about the name. The datasheet (and often info beyond that) is what realy counts.

\$\endgroup\$
  • \$\begingroup\$ What all should be considered when picking best available Microcontroller Based on Low-Power Specifications with 802.15.4 ? Can you give some important points I should keep consider? \$\endgroup\$ – Thomas Sep 15 '17 at 14:34
  • \$\begingroup\$ 'the best micro-controller' is always based on too many factors. But quite often choosing the uC is not a key choice. What are your application characteristics: duty cycle, computation power, energy source, const-of-fauilure, etc.? Maybe the best uC choice is dictaed by available chips with the relevent radio hardware built-in? \$\endgroup\$ – Wouter van Ooijen Sep 15 '17 at 15:52
-1
\$\begingroup\$

For a given clock cycle and a given instruction execution, the underlying hardware makes a lot of difference in the amount of power consumed by the uC. So also does the state of the uC (Active / Sleep / Low Power state, etc).

Based on those parameters, manufacturers classify their products into buckets called low power and ultra low power. There is no specific standard.

As an application developer, if you have a requirement where power consumption is a huge criteria, then you can start selection in the ultra-low power series of uC's. Again, your application power requirement can even be met by a "normal" uC. So those classifications only serve as a starting point for your search.

\$\endgroup\$
  • \$\begingroup\$ Is there any differences from the architectural point of view? \$\endgroup\$ – Thomas Sep 14 '17 at 15:37
  • \$\begingroup\$ Power consumption is independent of architecture per se. However you can see certain architectures consuming more power on average, for example x86 chips compared to ARM based ones. \$\endgroup\$ – Raghavendra Kumar Sep 14 '17 at 15:40
  • \$\begingroup\$ some of that has to do with the foundry and process as well... \$\endgroup\$ – old_timer Sep 15 '17 at 1:53
  • \$\begingroup\$ the marketing departments for these companies are going to strive to put these parts in those buckets, so sometimes if one particular mode/feature/test shows a low power number they call that good enough. dont be fooled by how they are advertised. check their specs, assume they are wrong by some percent, do whatever testing you can, then decide on the part for your product... \$\endgroup\$ – old_timer Sep 15 '17 at 1:55
  • \$\begingroup\$ note testing one part is not good enough, you would really need thousands to get a good range, but if your test is out of their specs then go after them about it and see if you get better numbers, all processes have variation and there is a range of power consumption within that process that the vendor will consider acceptable. buy and test two parts and you might find one consumes twice as much power and that might be perfectly fine within their process and screening. \$\endgroup\$ – old_timer Sep 15 '17 at 1:57

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