• State TI Thinks Resolved
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 66 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

Original question:

TPS1HB16-Q1: questions

TPS1HB16-Q1: Strange high side swtich behaviour...

Dimitar Kolev
Prodigy 50 points
Part Number: TPS1HB16-Q1
Other Parts Discussed in Thread: HSS-MOTHERBOARDEVM

Tool/software:

Hello TI team, 

have a strange behaviour of RPS1HB16-Q1 part. Will attach schematics if needed. I will double check but I think mine is with -B suffix

So, pin connections

Pin 1 / GND

Pin 2 SNS / 1k  PUD to GND and 15k series - not connected to MCU / Calculated for 5V MCU

Pin 3  LATCH / 15k PUD to GND

Pin 4 EN / In series 15k to 5V logic (inittially and always enable, until logic goes low)

Pin 5 ILIM / 27k to VBB (12V) - Max. load for now set to 4.4A so this should be 25k max... my fault

Pin 6, 7, 8 - VOUT

Pin 9,10, 11, 12, 13, 15 - NC (I can see in some variations with suffix D - pin 15 is SEL2 and PUD with 15k to GND?, but cannot find such datasheet)

Pin 14 SEL - PUD 15k

Pin 16 DIAG_EN - PUP to 5V 15k

Pin 17 IC tab - VBB 12V

Switch is intended to drive inductive load (relay), but just for initial experiment a small LED is connected to Pin 6,7,8 - consumed current is like 1mA

On power up (5V on logic, 12V on input) - LED was off initially, VOUT around 1V. EN signal was HIGH, 5V on init. Suspicition fall on ILIM incorrect value of 27k, so values below 25k and aboive 5k were tested, with same bad result. If ILIM is shorted to 12V - LED is on, but measurement shows 5.8V or similar voltage on VOUT - no way to reach 12V. Shorting LATCH to GND does not change much. Will further try to disable diagnostics... but behaviour is really strange, No excessive high current on 5V or 12V, no heat generated. I wonder if UVLO protection can kick in and EN pin should be enabled later - but I try to swtich on 12V first and then 5V - resulta was the same. I am a bit out of ideas what can be the reason. Do you intend to add figure on Figure 9-3. TPS1HB16-Q1 Inductive Discharge (5 mH), page 22? Relays intended is 3W /6W, will double check inductance if that can cause problems.

Thanks in advance!

  • 0
    TI__Mastermind 25055 points

    Dimitar,

    Thank you for the detailed explanation here.

    One note here is that for all of the digital pins, they do have weak internal pull-downs to ground so no need to explicitly pull them down (with the exception of SEL1, that needs to have a path to ground to enable FET turn on during reverse battery).

    For the ILIM pin- make sure for the new resistor value that it is as close to the pin as possible with little capacitance on the trace. We do have a good tool that will let you know the variance of the current limit behavior here https://www.ti.com/tool/download/SLVRBG2 .

    An important thing to note here is that this device has an integrated pull-up from VOUT to VBB that is used for short-to-battery and open load detection. When DIA_EN is high, this pull-up is enabled which probably explains the voltage you are seeing on the output. Is it possible that you are triggering the open load detection with such a small load current? 

    In the off-state, when DIA_EN is high there is a constant comparator that is compared (VBB - VOUT) to the open load detection threshold. If the difference is less than the open load threshold, the device detects an open load condition and signals the FAULT pin. For testing, could you rerun this test with DIA_EN low?

    Best Regards,
    Tim

  • 0 in reply to Timothy Logan
    Prodigy 50 points

    Hello Tim,

    finally have time to try. Unfortunately result is the same - unloaded output (1mA LED only) is staying on 1.5V, but it should go 12V. Will retry from beginning - I don't think anything is burned, it is just some bad configuration or something wrong in schematic or implementation. It is really simple, but it somehow is failing. 

  • 0 in reply to Timothy Logan
    Prodigy 50 points

    By the way, after chasing ghost I think I found the problem and will try to solve when possible. For some unknown reason VBB pad (and it is the only one) was covered with solder mask (but not the thermal vias). So there is no solder paste opening, and no solder possible - solder via thermal vias, even they are not plugged and not isolated turns out not possible - they are just 0.3mm. And erratic behavior and some output enough to power 1mA LED, but not enough for more - is actually leakage from the 5V logic (my opinion). Please correct me, if I think in a wrong direction. Solution will be next board, but for those I have already is just removing the component, scrapping the isolated pad and re-solder. Or I think so. 

  • 0 in reply to Dimitar Kolev
    TI__Mastermind 25055 points

    Dimitar,

    I apologize about the late response- was a national holiday in USA last week.

    This sounds like it would be the issue. If needed, it is possible to test on the HSS-MOTHERBOARDEVM to verify functionality on your load before actually building the new board. 

    Best Regards,
    Tim

  • 0 in reply to Timothy Logan
    Prodigy 50 points

    Hi Tim,

    Happy 4th of July, 4 days later ;)

    Fail was found and corrected - for unknown reason VBB vias under the chip was as they should - without mask, but soldering pad paste and isolation aperture was closed - so VBB was not soldered and not connected. Desolder, scrap insulation and solder back - now it is working as it should. Faulty measured voltages was leakage from logic inputs...I guess. Thanks for support and guidance.

    Have a great week,

    Dimitar