• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 52 subscribers
  • Views 2041 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

CC2564 Power Consumption

Charles Gervasi
Expert 2270 points
Other Parts Discussed in Thread: CC2541, CC2564, CC2540

I'm new to Bluetooth Low-Energy, and I'm trying to work out power budget.  Application Note 092 does a great job of showing the current draw in various stages of a connection.  

You can work out that this event uses 22.0 uAs of charge.  So if we had one of these per second, we'd be pulling 22uA average.  

Elsewhere in the app note, it says PM3 uses only 0.4uA and PM2 uses 0.9uA.  The difference is PM2 allows a timer to run; PM3 requires a hardware interrupt to wakeup. 

I'm hoping someone knowledgeable about Bluetooth can let me know how many transactios of how much data are required to establish a connection.  I imagine all transactions will require about the same time and current in each stage except for the RX and TX stages. 

My application will have infrequent transactions, so current in idle (e.g. PM3) may predominate in my power budget.  (The application requires a battery powered device with a tight power budget to talk to a line-powered device.)  I'm trying to compare the CC2541 SoC (which has an on-board 8051 capable of running the stack) and CC2564 (which is "just" the Bluetooth controller and radio). 

CC2564's datasheet talks about "deep-sleep" and and "shut-down". 

I can't find info about what deep-sleep mode is about.  It doesn't make intuitive sense that these modes should have a higher current draw than PM2, which leaves a counter running, on a SoC.  I would expect the sleep numbers to be lower since CC2564 is not a SoC with a processor running the stack.  Is CC2541 SoC just amazingly good in terms of current draw during sleep compared to CC2564? 

Can someone help me understand how much power budget I can except the battery-powered side need to establish a connction and the purpose of CC2564's "deep sleep"?

Is there also an app note that goes over the basics of establishing and dissolving a Bluetooth LE link in an application the requires infrequent data exchange?

  • 0
    TI__Guru 57865 points

    Hi Charles,

    I cannot answer for CC2564 and giving you a fair figure for connection establishment will be hard as it depends on several parameters. What I can tell you is that using CC2541 as one chip solution will significantly improve power consumption compared to a dual chip (CC2564) solution as no processing due to serial communication is needed. 

    I believe you'll find some useful information on our wiki (www.ti.com/ble-wiki).

    I suggest you clone this post to the Bluetooth forum to get some answers on the CC2564.

    Best Regards

  • 0
    TI__Expert 4060 points

    Charles,

    In regards to your question about comparing the deep-sleep/shut-down current between CC2564 and CC2540/41, remember that these are very different devices created in different silicon technologies, so while it may seem non-intuitive, it is indeed true that the sleep current of CC2540/41 is much lower than CC2564.

    Best regards,

    Karl

  • 0 in reply to Karl T.
    Expert 2270 points

    Karl T. said:

    In regards to your question about comparing the deep-sleep/shut-down current between CC2564 and CC2540/41, remember that these are very different devices created in different silicon technologies, so while it may seem non-intuitive, it is indeed true that the sleep current of CC2540/41 is much lower than CC2564.

    Do you know the practical difference between deep-sleep and shut-down?  Our application does not require us to stay links.  We transmits bursts of data that are triggered by other events.  The battery-powered device does not need to be able to be linked and ready to receive data. 

    In shut-down CC2564 pulls a max of 3uA at room temp.  Our power budget is around 400uA average, so the 3uA wouldn't rule out this chip, unless putting it in this mode requires more charge (current * time) to get it back in a mode in which in can communicate. 

  • 0 in reply to Charles Gervasi
    TI__Expert 4060 points

    Hi Charles,

    Unfortunately, I'm not familiar enough with the CC2564 to tell you the difference between deep-sleep and shut-down. As Nick suggested, you are more likely to get an answer to that on the "classic" Bluetooth forum.

    I'm pretty sure that the CC2564's shut-down mode is pretty comparable to the CC2540/1's PM3 mode, except that I do believe that the CC2564 does not have any retention in this mode (all memory and registers need to be re-loaded on wake-up) while the CC2540/41 has retention for RAM and registers in all power modes. This may have some effect on start-up time. Again, I'm not familiar enough with the CC2564 to say how much energy it would spend waking up. If you get this information on the other forum, then it should be easy to compare with the CC2540/41 data from the appnote you've already found.

    Best regards,

    Karl