• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 315 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TPS92515: Repeatability error on a design using Vadj

Led design01
Prodigy 40 points
Part Number: TPS92515

Hello,

I am observing an output current  repeatability error, only on a few units (around 2%) on a production, using Vadj to control the output current.

The production test I am doing is setting Vadj from 0 to 1V which give 650 mA output current +/- 50 mA depending on board. Usually output current is consistent all the time on a SAME BOARD within 1 mA.

Sometimes, on a few units gives either for instance 650 mA or 620 mA. There seems to have only two scenarios possible 650 or 620 mA for the SAME BOARD. Settting Vadj to to 0v and back to 1V will switch to 650 or 620 randomly.

- I checked Vadj which is exactly the same with no ripple in both scenario

- Vin is the same in both scenario

- Load is the same in both scenario

- Switching frequency is a little bit different

Am I clear enough in my explanations ?

Any idea where to look at ? 

Thanks !

  • 0
    Prodigy 40 points

    I'm thinking it can be related to "turn-on calibration " but do not find information on the way it works ...

  • 0 in reply to Led design01
    TI__Mastermind 28655 points

    Hello,

    Are you power cycling when seeing different measurements?  Basically turn on, measure, turn off and back on and measure?

    Best Regards,

  • 0 in reply to Irwin Nederbragt
    Prodigy 40 points

    Yes, I confirm I am power cycling as the IC is powered by a boost converter that is turned off when light set to 0% for power saving but main MCU stills ON. Is there a delay to respect between power ON and adjust Viadj that will ensure repeatable calibration ?

  • 0 in reply to Led design01
    TI__Mastermind 28655 points

    Hello,

    The TPS92515 has a power on calibration that allows tighter set-point regulation.  It has 5 mV adjustment steps at the error amplifier so if a TPS92515 is close to one of the calibration points it can power up in different states.  At Viadj = 1 V the peak current threshold will be 100 mV, this is what can change by 5 mV if close to a calibration step during power cycling.  If this is an issue look at using the TPS92515A however it's set-point accuracy is not as good.  I created a table of the set-point differences.  If you are using 1V for a reference the accuracy will be worse by 2.2V/1V on the table below.  Also increasing your Viadj and using a higher current sense resistor value will reduce the step difference.

        2.4V clamp percent 2.2V reference percent
        mV   mV  
    TPS92515A minimum 222 92.50% 205 93.18%
      typical 240 100.00% 220 100.00%
      maximum 257 107.08% 234 106.36%
               
    TPS92515 minimum 224 93.33% 211.5 96.14%
      typical 240 100.00% 220 100.00%
      maximum 251 104.58% 223.5 101.59%

    Best Regards,

  • 0 in reply to Irwin Nederbragt
    Prodigy 40 points

    Thanks a lot for your reply. I have few questions more : 

    - Does something can be done to have repeatable initialisation ? Does the state of Viadj when power up impacts the calibration ? Is there any recommendation to help improving having the same calibration at each power up ?

    - I understand it hesitates between two calibration points and sometimes it start with one, sometimes the other. Am I right ? Does this behaviour depends on some fabrication tolerances ? I'd like to know if not seeing this phenomemon on a particular board / IC can give confidence that it will not happen, while those showing  the issue are, due to manufacturing tolerances, not "as good" ?

    thanks

  • 0 in reply to Led design01
    TI__Mastermind 28655 points

    Hello,

    1)  The TPS92515A solves this issues for power on calibration since it is disabled.  Viadj does not have anything to do with power on calibration, it's something done when measuring error amplifier offset internal to the device.  Nothing can be done during power-on to avoid this and is why the TPS92515A was created.

    2)  If it is near a calibration point any slight change during measurement can move it from one state to another.  A slight temperature change, any slight noise, etc.  It doesn't have to do with manufacturing tolerances, it has to do with how close to a calibration step the TPS92515 is at.  If one device is near the center of a calibration step it won't change during power cycling while a device that happens to be close to a calibration step may change between calibration steps.  It just depends on where the offset of the internal error amplifier is.  There are multiple calibration steps.  Again, the TPS92515A does not have this calibration during power on and really the only way around this issue.  Using a higher value of Viadj will reduce this affect but for deep dimming the TPS92515A should be used if repeatability is needed.

    Best Regards,

  • 0 in reply to Irwin Nederbragt
    Prodigy 40 points

    Very clear, thanks, though quite unconfortable to discover that during production ...

    TPS92515AHV-Q1 seems to be the only reference having the 'A' suffix (I don't need HV type in  my application). Can you confirm ? Then availability is another issue ... let me know if you have any way to get those parts please (10 Ku )

    thanks,

  • 0 in reply to Led design01
    TI__Mastermind 28655 points

    Hello,

    The AHV is the only one available.  This was created for the changing set-point issue you are seeing during deep analog dimming.  As far as availability, unfortunately this is one of many parts that are hard to find which is an industry problem.  Sorry about the unpleasant news.

    Best Regards,