• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 4 answers
  • Subscribers 62 subscribers
  • Views 492 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.

BQ27Z561-R2: Delay from deep sleep to Sleep is 1 second extra

Marien Vermeulen1
Prodigy 230 points
Part Number: BQ27Z561-R2


I'm testing the BQ27Z561-R2 in details..  the chip performs wel.
I did static charge tests, and now i'm busy doing dynamic CC test.   results are quite good.

But the switch from DeepSleep to Sleep mode takes 1 second extra!
In another tread I already had the answere that the switch/exit  from Sleep to Normal is 4  seconds.  I got this estimated from TI support. 
My test show that the CC is working good during Sleep and Normal.

I have added my scope picture.
Above the battery current.  is flowing through a load  during 6 seconds.
Down the operating current of the BQ.  First Deep sleep, than Sleep, than Normal.  (= yellow circles)

In Deep Sleep the first blue arrow samples the current, when the battery has load.  this meets:  |wake comparator current| ≤ Deep Sleep Current.
So the BQ should switch to Sleep.   it does not...
The second  blue arrow samples again the load current.   Now the BQ goes to Sleep.   So it takes 1 extra second.
During  Deep Sleep de CC does not count the real (load) current.  This results in a lower  mAh.

Changing "Wake Check Time"  does not help   value 0  disables the exit from DeepSleep to Sleep.  value 2 and higher increase the exit delay.

It there a way to shorten the DeepSleep to Sleep delay ?

  • 0
    Prodigy 230 points

    it seems my copy paste picture is gone?
    now as attachment..

  • 0 in reply to Marien Vermeulen1
    TI__Guru** 106130 points

    Hello Marien,

    The gauge measures the current periodically in deep sleep not continuously, so it will take time for the sample to be taken and wake the gauge up from deep sleep.

    Sincerely,

    Wyatt Keller 

  • 0 in reply to Wyatt Keller
    Prodigy 230 points

    Hi Wyatt,  Yes  Deep sleep samples the current each second.
    But as you see it goes from Deep Sleep to   Sleep at the second Current sample.  Not immediately at the first.
    The blue arrows show these two sample moments..
    Due to this delay the CC starts rather late with continuous measuring...

  • 0 in reply to Marien Vermeulen1
    TI__Guru** 106130 points

    Hello Marien,

    The gauge must take a measurement and then wait at least 1 second before exiting deep sleep.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 230 points

    Hi Wyatt,

    Thanks for your answere,  than I have set up the BQ correctly.

    Is it possible to shorten those reacion times of the device:
     the 1 second for DeepSleep to  Sleep
     the  4 seconds for Sleep to Normal

    in a new version  -R3 ?

    It will improve the accuracy significant!  for short time operating loads.

    (I do not understand whoy the device should wait for changing modes.  I would also expect the device to be able to switch from Deep Sleep to Normal, without passing the Sleep.
    My BQ current consumption mesurements show:
         Deep sleep  = 10.8uA         datasheet  = 9uA
         Sleep  = 18.4uA         datasheet  = 11uA
         Normaleep sleep  = 60uA         datasheet  = 66uA
    it is no option to keep the device in Sleep instead of Deep sleep.  18.4 is to much.  would be 68% of the application.

  • 0 in reply to Marien Vermeulen1
    TI__Guru** 106130 points

    Hello Marien,

    Unpredictable pulsed loads are hard to capture when trying to conserve energy, if you know when the pulse will be you can wake the gauge from DEEP SLEEP to SLEEP where the CC is active all the time and will capture shorter pulses.

    The conversion time from SLEEP to NORMAL should not affect accuracy significantly because the gauge will be running the CC in SLEEP.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 230 points

    Hi Wyatt,  thanks for your answers, help us a lot.
    We let the BQ working stand allone, the current is the only value that signals the start of the oprational time of our application.
    This works fine.   Its only that de CC starts rather late...
    We did test this  on a 6 second opartional time of our app, each  20 seconds cycle.
    The CC gave a 80% result of the real capacity.
    For longer operational phases the errror gets less
    For shorter operational phased is will get worse..

  • 0 in reply to Marien Vermeulen1
    TI__Guru** 106130 points

    Hello Marien,

    I see the issue you're running into, it's a challenging application that you can't stay in SLEEP mode to be always capturing current measurements. I think your best 2 options are to either wake up from DEEP SLEEP manually if possible or remaining in SLEEP.

    Sincerely,

    Wyatt Keller

  • 0 in reply to Wyatt Keller
    Prodigy 230 points

    Hi Wyatt,

    We take th eSleep mode as an option.  My operating current from the BQ chip showed a 18uA consumption, more than the datasheet.
    But this seems to be caused bij floating IIC lines..  If IIC lines float between 0.5 an 1.0 Volt. The chip draws 7 uA more.
    When IIC islow, hen it is 13.5 uA in  our design.
    So We can go to Sleep.  Costs some energy, but we take it as an improvement.

    battery gauging  can be a difficult target...

    Thanks,  we solve this issue..