• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 2377 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.

Zigbee End Device battery power design and selection

Mosi
Intellectual 735 points
Other Parts Discussed in Thread: CC2530, TPS62730, TPS62740

For a application some Zigbee End Device modules (based on CC2530) are used as simple Switches and they should be powered by batteries. Now I was wondering are there some considerations in hardware design phase to get the best use out of a battery? 

Like I have seen people put a capacitor in parallel with the battery to sustain battery life! Should I calculate this capacitor or any pretty large capacitor would be fine? (I know there is the leakage)

Plus which kind and how many batteries are usually used in applications like mine? (CC2530 operating supply is 2~3.6V)

I was thinking about using 2 (in case of powering directly from battery) or 3 (in case of using 3.3V LDO after battery) AAA 1.5V NiMH batteries. What do you think? Is there any better options? (of course cost and space is very important!)

I have another question. Should I put a LDO regulator after the battery or is it better to power the module directly from battery and without LDO?

Considering that CC2530 works between 2 ~ 3.6V and if I put a 3.3V LDO then battery does not power the module when it fells down to 3.5V and will waist a lot of its capacity (LDO does not work correctly anymore)

On the other hand I know if I use LDO after battery and get a lower voltage then current consumption of the CC2530 would decrease considering its operating current is a function of its input voltage.

I'm afraid if I use a battery directly then there would be some sort of instability in the performance of the module. I'm not sure though. 

So to sum it up I need advice on:

1- Which kind and how many batteries

2- Using or not using LDO

3- Parallel capacitor with the battery

for a long run time and good performance.

  • 0
    TI__Guru**** 150820 points
    I moved your post to this forum which is better suited to your type of request.

    You can see an in-depth analysis with the TPS62730 here: www.ti.com/.../swra365b.pdf

    But since we released that device, we made the TPS62740 which is more efficient.

    Most customers use 2x AA/AAA batteries, but this usually depends on your application and use case.
  • 0 in reply to Chris Glaser
    Intellectual 735 points

    Thank you so much Chris.
    Using 2x AA batteries seems reasonable. But does it matter which kind? Ni-CD, Ni-MH, Li-ion ?
    I mean what is the routine that other people do for these kind of applications?
    Do they use non-rechargeable batteries because they are likely to last longer?

    One other thing. I was planning on using TPS780xx presented in slyt356.pdf (http://www.ti.com/lit/an/slyt356/slyt356.pdf) application note.
    Is it better to use that LDO or TPS62740 which is a Step-Down Converter?
    I noticed TPS62740 has 350nA IQ vs TPS780xx which has 500nA and I think that matters too much. Don't you think so? But does it have other advantages over low IQ LDOs?

    I'm kind of confused. If you could answer me you'd really be helping me out man!

  • 0 in reply to Mosi
    TI__Guru**** 150820 points
    Yes, the choice of battery is determined by whether you want to design a charger for your product or just want the user to just buy alkalines when they are empty. Among batteries, there is a clear tradeoff of cost, size, power density, etc.

    The app note I sent should address your questions of whether to use an LDO or SMPS and the tradeoffs. Modern, LPRF optimized SMPS give longer battery life.
  • 0 in reply to Chris Glaser
    Intellectual 735 points

    Thanks again.

    No I don't want it to be chargeable. It's just not worth it and I think it's not just the right thing for me.

    I have those limitations as well. Space and cost are very important along with capacity.

    I want to use 2x AAA Alkaline batteries (for they have good characteristics and they are pretty small and cheap) but it seems that all of them are likely to give 1.5V each (3.0V altogether) and 3.0V is not good enough for this application.

    Guess I have to go with a 3.7V Li-ion (or two of them in parallel)

    Do you have a better suggestion? 

  • 0 in reply to Mosi
    TI__Guru**** 150820 points
    Yes, AAA batteries give between 1V and 1.5V as they discharge. You can find example curves in their data sheets.

    Did you read the app note already?

    A buck converter or LDO would go between the batteries and give the voltage directly to the CC device.
  • 0 in reply to Chris Glaser
    Intellectual 735 points
    No right now I'm sort of busy with something else. I guess I'm just gonna have to read it tomorrow.
    Yeah I know it would be placed between the batteries and the CC. That's what I was talking about. When you have a 3V and you put a LDO or SMPS then you're not working in the entire CC voltage range (2.0V ~ 3.6V according to datasheet) which would make me use a 3.7V Li-ion battery so with a LDO or SMPS I can get 3.6V down to 2.0V
    That is if I can find a suitable AAA sized Li-ion battery with a good capacity. I don't know I just figured I'd have much better options for power and battery part in my design.
  • 0 in reply to Chris Glaser
    Intellectual 735 points
    I've been reading that application note you suggested but there is one thing about it that really confuses me. This app claims that TPS62730 reduces battery current consumption by delivering less voltage to CC254x in active mode (2.1V apparently) and delivering the original battery voltage (2.0V ~ 3.6V) in sleep or bypass mode.
    While the slyt356.pdf document (extend battery life in MSP430 applications) says that you have to deliver a high voltage to SoC in active mode (like TX) and deliver a minimum voltage (like 2.0V) in sleep mode to reduce sleep current consumption and extend battery life (considering SoC's current consumption is a function of its voltage) and apparently TPS780xx does this job very well.
    The second case makes more sense to me. SoC working with a higher voltage and sleeping with a lower voltage.
    What do you think? Why is there two different perceptive here? Which one is more optimized and better way?
  • 0 in reply to Mosi
    TI__Guru**** 150820 points
    Both app notes have very similar ideas to save power in a low power MCU system. Let me point out a few differences:

    The TPS780xx app note was written many years ago before the TPS6273/40 were released.

    The app note discusses an MSP430, whereas you have a CC254x.

    We now have the TPS62740 which is likely more efficient than either the TPS62730 or an LDO.

    The app note on the TPS62730 really does a good job of explaining how a SMPS saves power and yet maintains the RF performance.



    Clearly, your MCU consumes less power if it is supplied with a lower voltage. Either an LDO or SMPS can be used to make this lower voltage. With an LDO, the current drawn from the battery equals the Iq + the MCU's current. For a SMPS, the battery current drawn equals the Iq + the MCU's current * the efficiency and a Vout/Vin factor. Clearly, this is less input current due to the conversion factor of Vout/Vin.

    The TPS62730 was innovative in that it cleanly converted to the lower voltage during Tx--thus, giving the benefit of the Vout/Vin conversion factor. In this higher current mode, the Iq is not important. So, it is more efficient than an LDO. In the low power mode, it is in bypass mode drawing very very little Iq--even lower than an LDO. Yes, the MCU current will be higher but there is no Iq loss. This is a batter tradeoff than running with the lower volatge but adding the Iq of the TPS62730. Overall, it is a very good tradeoff which results in longer runtime for the system.

    The TPS62740 is better because it operates the MCU at the lower voltage all the time, while drawing even less Iq than the LDO. So, both modes have clear efficiency improvements.

    And your MCU has certain requirements for its supply voltage for various operating modes. I believe the CC254x just requires ~2.1V for all modes. For some higher frequency modes, the MSP430 in the example required a higher voltage. The TPS62740 supports this too with its Vsel pins. Of course, you need to check the voltage you need in your application.