• State
  • Locked Locked
  • Replies 18 replies
  • Subscribers 32 subscribers
  • Views 1705 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.

SLOW Clock Output on TM4C123

Bruno Saraiva
Guru 13040 points
Other Parts Discussed in Thread: TM4C123GH6PZ

Gentlemen,

Working on a new project that uses TM4C123GH6PZ. This board project has a Bluetooth module, which requires a 32768Hz clock into it (square wave, 1.8V). 

On our previous boards, we have a 16MHz on OSC0/OSC1, and a 32768Hz on XOSC1/XOSC2. This board will use RTC, hibernate, etc, so it is likely that the crystal arrangement will remain the same.

Is there a pin on the uC that can be used as an output for the 32768Hz signal? Or is it convenient that I electrically tap the crystal signal into a opamp and "duplicate" its clock? I guess what I'm looking for is a suggestion for a clean design that can use the 32768 clock for both purposes.

Further, I don't believe it is a good idea to use a software clock output (but maybe I can configure one of the timers to run from the slow crystal and generate a 50% PWM output at that rate?).

Cheers,

Bruno

  • 0
    TI__Guru**** 244400 points
    Hello Bruno,

    Not on the TM4C123x devices. You can use a HW timer to generate a 50% output clock from the main system clock and then signal condition the same for the 1.8V square waveform.

    Regards
    Amit
  • 0
    Guru 27665 points
    Maybe use an oscillator instead of a crystal?

    Robert
  • 0 in reply to Robert Adsett
    Guru 13040 points

    Thanks Robert and Amit,

    Indeed, after posting, I was researching a bit more, and came across this oscillator: SIT1533 series from SiTime.

    I guess I can use such an oscillator to generate the clock... I'd power it from 3V3, and divert it to the two devices:

    - directly into the XOSC0 pin of the uC, and
    - via a level shifter into the 1.8V USB clock input.

    Now it is just a matter of designing the power control of this circuit, in order to obtain the best possible controllable low power arrangement...

    If anyone is willing to discuss the ideas, the help is much appreciated! Thanks again.

    Bruno

  • 0 in reply to Bruno Saraiva
    TI__Guru**** 244400 points
    Hello Bruno,

    I would look at a CDC device (Clock distribution circuit) as well.

    Regards
    Amit
  • 0 in reply to Bruno Saraiva
    Guru 47900 points

    Bruno Saraiva said:
    Now it is just a matter of designing the power control of this circuit

    IIRC - famed (dead) General Custer similarly proclaimed, "Now it is just a matter of...few Indians!"     And - unknown to the good General - there were many - and others lurking.

    Moral of this story - boyz run fast/far when boss/client/wife announces, "Just, Simple, Quick!"    Almost always - such proves untrue - or it's the (unanticipated & improperly considered/planned for) often "unknown unknowns" - which cause great mayhem & destruction...

    Power control may appear simple - doing it fully/completely - at greatest efficiency - and quickly - perhaps, Not so much!

    Thread's title, "Slowed Clock Output" confounded our staff - thus we skipped.    Would not, "32KHz clock generation/distribution" proved better?

  • 0 in reply to cb1
    Guru 13040 points
    You tend to be pessimistic, oh Great cb1...
    Hadn't we proclaimed that a power control is a matter of just a few thoughts, our unknowing customers would judge we knew nothing of the theme... And maybe we'd better start a bakery business, which is just a matter of putting some flour, a few eggs and butter together and heat it up for some minutes.
    No, I still believe we shall face these challenges with elegance and might, and put on a face that shows we can make them happen! And oh, actually make them happen!
    Then again, there are so many millions of products out there with such an embarassing, naive, childish, arduinish project - and even those somehow work - that it comes to a time when we should give up searching for the next better alternative, and approve the one we have at hand! Myself I tend, at times, to spiral around the targeted optimum solution too much - only to look at my competitors products buzzing their noise already on the market... :(
    We need to send spare parts to our station in Mars? We already have the parts on Earth, and we already know where Mars is supposed to be... now it is just a matter of building the cart!
    Otherwise, always pleasant to read your comments! ;) My quote of philosophy for the week is gone.

    Amit, I looked at CDC. Thanks!
  • 0 in reply to Bruno Saraiva
    Guru 47900 points

    Bruno Saraiva said:
    My quote of philosophy for the week is gone.

    Staff/I shed tears - surely you've got, "at least one more!"   Really - yours was an inspired writing - not ALL tech must be "by the book" dull/boring zzzzz.

    Note my "attack" ("if" my writing so qualifies) was confined to a "single word" - did not tread anywhere near "personal."  

    I stand by, "Raising the gates - manning the moats" each/every time client/boss/wife announces, "Just!"    It is NEVER "Just!"

    As to "Great cb1" that award unsought by this reporter - just trying to honor, "Work Release" by (sometimes) thoughtful appearance here.  (surely beats, "chain-gang")

  • 0 in reply to cb1
    Guru 13040 points

    Folks,
    To share my further researches:
    I decided to abandon de use of the TM4C123 as the RTC base of the project. Modern dedicated RTC's can be extremelly low power, some to the order of 15nA when running on battery backup (time keeping only).
    So at least for our scenario, the best way is to fully shut down the uC when not used, and rely on a separate chip for RTC. Those IC's typically have a square wave output, which will supply the initially described 32768Hz signal for the BT module.
    If a periodical uC wake-up is required on the project (not my case, but quite common on remote sensors and IoT stuff), there are similar low power RTC's with an actual programmable switch to perform that task.
    Thanks again for sharing your suggestions - it is not with all those ideas that the final figure comes to existance.
    Cheers.
    Bruno

  • 0 in reply to Bruno Saraiva
    Guru 27665 points
    Loop nicely closed Bruno.

    Robert
  • 0 in reply to Robert Adsett
    Guru 47900 points

    Better than that - loop & philosophy - combined. (thus 2 "Likes")

    Might it be that Bruno - here/now - has subscribed to, "Right (i.e. BEST) Tool for the Job!"     And sometimes (often) that is NOT the MCU!

  • 0 in reply to Bruno Saraiva
    TI__Guru**** 244400 points
    Hello Bruno,

    When you mention the uC to be fully shutdown, did you mean Hibernate or Power OFF?

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Guru 47900 points
    Pardonnez-moi Amit - but as the RTC , "does so much - and so well" in satisfying poster's objective - does this (fine MCU detail) really matter?
  • 0 in reply to cb1
    TI__Guru**** 244400 points

    Hello cb1,

    I agree, but the next statement prompted the question

    Bruno Saraiva said:
    If a periodical uC wake-up is required on the project (not my case, but quite common on remote sensors and IoT stuff), there are similar low power RTC's with an actual programmable switch to perform that task.

    It is suggestive not definitive. Not want another poster to use this statement as an implementation without knowing how.

    Regards

    Amit

  • 0 in reply to cb1
    Guru 13040 points
    Mes amis,
    I mean REALLY FULLY POWERED OFF, as in "unplug it from the mains". Or in my case, disconnect the device's 4xAA batteries and leave only the backup Lithium battery supplying power to the RTC.
    This is comparing I_HIB_RTC from the datasheet, at 1.4uA (not bad, really), to the same condition on the external RTC chip, at less than 20nA (impressive).
    I got particularly happy with this concept: the RTC IC version with a programmable switch would allow, for example, to wake up the uC (turn on the main 3V3) once every XX minutes, do something, and then turn it fully off (as opposed to leaving it on hibernate mode).
    Je suis quite tired today to make sense, sorry if the explanation is less than ideal.
    Cheers.
  • 0 in reply to Bruno Saraiva
    TI__Guru**** 244400 points
    Hello Bruno

    Appreciate the post (in spite of being tired). That does clarify my doubts.

    Regards
    Amit
  • 0 in reply to Amit Ashara
    Guru 47900 points

    Loved everything - especially "Le Francais!"    (but you KNEW that!)

    So - which RTC did you choose?     Sounds super!    And @ 20nA that battery will (likely) "outlive" cb1!     (horrors that...)

    What tired - you got DEUX "LIKES" - did you not?

  • 0 in reply to cb1
    Guru 13040 points

    That's the Abracon AB18xx (with the additional switch), or AB08xx (without). Some vendors trade them as AmbiqMicro.
    I chose a very small rechargeable battery - saves me footprint and ensures cb1 outlives it, we wouldn't want otherwise!

    What tired? This ain't room for personal laments, but getting ready to move to another country, rearranging the company's partnership, covering for my previous partner who also left the country, keeping up with the technical projects, and trying to ensure family stays happy... Can't wait to see all these in the proper squares in a couple of months or so! Merci por les deux "j'aime"!

  • 0 in reply to Bruno Saraiva
    Guru 47900 points

    Merci mon ami. J'acheter dix - such Abracon w/switch - even (and especially) without [yet] fleeing the country!

    This reporter would settle for (maybe) 3 of the 5 "targets" you've listed. In our case - operating from a Back Room - behind a Back Room (but w/expansive windows & bright, unblockable views) helps hold prices down - enables great focus...

    We assume such a move does not involve, "crossing the pond?"    I'd "j'aime" - encore un fois (3) - but forum disallows such...