• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 62 subscribers
  • Views 326 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.

BQ24075: Shorting IN and OUT Together

Donald Holden
Prodigy 10 points
Part Number: BQ24075

Hello,

I am currently evaluating the bq24075 and plan to integrate it into a product soon. I am prototyping with a Seeeduino XIAO. I was mislead to believe this device had some sort of advanced power routing features onboard, as it has a separate VIN pad on the bottom of the board which is marketed to be "perfect for powering the board with a LiPo battery." I assumed this implied that it was somehow isolated from the 5V pad. The 5V pad is either input or output. When powered from the USB port, the 5V pad is just a pass-through. Alternatively, you can supply the 5V pad from an external supply, which won't back-feed through the USB port thanks to the inline diode. What I didn't know is that the VIN pad on the bottom of the board is ALSO tied directly to the 5V pad. So essentially there's nothing special about this board whatsoever in terms of power management. It's no different from any other Arduino-inspired board with a 3.3V regulator. Schematic is here. files.seeedstudio.com/.../Seeeduino-XIAO-v1.0-SCH-191112.pdf

Sorry for the backstory. Onto the bq24075. Because of my incorrect assumption, I've essentially been prototyping for three weeks with the IN and OUT pins SHORTED TOGETHER. I was feeding the bq24075's IN from the 5V pad on my XIAO, so I can charge my LiPo when a USB cable is connected. Then I had the OUT of the bq24075 connected to the VIN pad on my XIAO. It seemed like a solid plan and it's been working perfectly for weeks now. But today I finally discovered on accident that the 5V and VIN pads are actually directly connected. I've been studying the bq24075's datasheet to try and figure out why I haven't seen any white smoke. Shorting the IN and OUT together seems incredibly unorthodox after all. However, after looking over the datasheet for a while I've concluded that I think it's actually perfectly okay to do this thanks to the bq24075's Power Path Management. But I really wanted confirmation from TI before I integrate it into my product in this configuration. Is there an adverse effect I'm unaware of - a reason I shouldn't be using the bq24075 in this way?

From what I can tell, Q1 will turn on and effectively short VIN to OUT when VIN is driven by 5V anyway since it would be greater than VIN(DT) + VBAT. But WITHOUT an external supply connected, shorting the IN and OUT pins together externally as I currently have it configured, VIN would be equal to VBAT (since Q2 will be on, to power the circuit from the LiPo), and therefore NOT greater than VIN + VIN(DT), which means that Q1 will remain off anyway. So, even though the bq24075 is back-feeding itself from OUT to IN when running on the LiPo, the voltage will remain too low to activate Q1.

This makes sense to me, but it's so unorthodox that I just wanted to confirm with TI that this is an acceptable circuit with no adverse effects. I sincerely hope to hear from a TI engineer about this soon. I apologize for the lengthy post for what is probably a trivial, elementary question. I'm a self-taught engineer after all. I could avoid all of this simply by powering my circuit with an external 3.3v regulator, driven by the bq24075's OUT, rather than the XIAO's onboard regulator which is tied to the USB 5V rail creating this weird loop. But if I can operate it this way I would much prefer to do so to keep my BOM and PCB footprint smaller.


Thank you very much,

Don Holden

  • +1
    TI__Mastermind 23200 points

    Hi Don,

    Welcome to E2E! 

    If my understanding is correct, you've got the IN and OUT pins on the BQ24075 shorted is all, is that correct? This isn't entirely ideal but the reason you aren't seeing any damage on the device is most likely the Q1 FET. Although in the functional block diagram we have one FET, we actually have two in a back to back configuration. This will prevent any current going in the direction of OUT to IN. 

    We don't really recommend this operation though as this could also simply be bypassing the circuit all together and so although power is coming into the IN pin and charging the battery, it doesn't sound like the device/load that is connected to the OUT pin is actually being powered by the OUT voltage but rather the XIAO. 

    Best Regards,

    Anthony Pham

  • 0 in reply to Anthony P Pham
    Prodigy 10 points

    Thank you for your reply. Yes, the truncated version of what I was trying to say is that I am shorting the IN and OUT pins.

    I'm happy to know this configuration won't cause any damage. I understand why TI wouldn't necessarily recommend this configuration, as it doesn't fully take advantage of the BQ24075's Power Path Management features, correct? In my current configuration, the BQ24075 is acting as more of slave than a master in the power management topology. So yes, when USB power is connected it powers the XIAO's 3.3V onboard regulator which powers the rest of my circuit, but also powers the BQ24075 with 5V USB bus power and therefore charges the battery. I am okay with this operation. When USB power is removed, Q2 turns on and the battery takes over. It's unfortunate because I'm wasting the BQ24075's Power Path Management functionality when operating on USB bus power and using the XIAO's rather rudimentary power management instead... which is literally just a diode.

    So basically, if you think about it, all I'm REALLY giving up in this configuration is the BQ24075's Battery Supplement Mode, right? I'm okay with that, since my circuit's power requirements will never come close to saturating even a 500mA USB supply, with plenty of power leftover to charge the LiPo as well (honestly the device won't be in use while charging anyway).

    Does this all sound correct? Not an OPTIMAL design by any means, but not certainly not damaging? Again, thanks so much for your prompt reply and your helpful advice. TI has never disappointed me.

    Don

  • 0 in reply to Anthony P Pham
    Prodigy 10 points

    I'm just waiting for confirmation that my configuration won't have any adverse effects to the BQ24075, even though it's not an optimal design as it bypasses the chip's Power Path Management features. I just want to be certain I won't cause any harm or reliability issues down the road, and then I will mark this issue as resolved.

    Thanks again, Anthony.

  • +1 in reply to Donald Holden
    TI__Mastermind 23200 points

    Hi Donald,

    I don't believe there would be any effects on the BQ24075 but this is not a guarantee on my end as this is not an application that we test as a use case method.

    You're correct in that this configuration wouldn't allow for supplement mode. 

    Best Regards,

    Anthony Pham