TPS61022: Schematic and layout check in close proximity to BQ25895

Part Number: TPS61022
Other Parts Discussed in Thread: BQ25895, , BQ25890, TPS2121

Tool/software:

Hi, 

I am using the TPS61022 for 5V 3A boost from the SYS output of my BQ25895. I wanted to get a schematic check of the whole circuit and would love feedback on how to optimize on space and component size. I have two main questions, 1) since both my inductors are close how bad will the EMI be?

2) Previously, when no battery is connected and I was using the PMID 5V 3.1A built into the BQ25895, I would sometimes get a low voltage error on my CM5 when under heavy load, this disappears when a battery is attached. How can this be rectified? More capacitance? Would love some insights!

Schematic: 

I have not made the ground pour visible for easier readability but there will be bottom and top ground pours. This is a 6 layer PCB with 2 additional internal ground planes and a +5V plane.

  • Just another note, the Low Voltage warning on my Pi (which is being powered by the PMID output) seems to only come up once power draw exceeds 5-6W or 5V 1A ish. 

  • Hi,

    For TPS61022, this schematic looks ok to me. Putting two inductors together may have some impact. However, in order to minimize the problem caused by EMI, you can try to use shielded inductors, and reserve a subber circuit for later adjustment if there is space.

    For BQ25895, this device belongs to another product line, so I will forward this thread to them. Thanks.

    Aurora

  • Dear Aurora,

    thanks! Yes both inductors are indeed shielded, realistically I expect maximum current through either at any time to not exceed 4A, typically rest between 1-2A.

  • Sorry an additional question, does the inductor choice for the TPS61022 seem sufficient? https://www.lcsc.com/product-detail/C19188671.html
    The device is unlikely to draw more than 15-20W I.e. not more than 5V at 3-4A.

  • Hi,

    For BQ25895, could you please help to clarify "get a low voltage error on my CM5 when under heavy load"?

    Thanks,

    Ning.

  • Hi Ning,

    Using a USB power meter to keep track of power draw, I run geekbench 6 on CM5 in Pi OS, typically it’s ok under single core loads and power draw from USB tops at around 5W, (5V 1A) however once it reaches multi core this increases to up to 10W but usually around the 6-7W mark Pi OS starting giving a low voltage error. It then subsequently shuts down at times when it reaches above 8W atleast from the USB meter. I have a feeling some of the spikes aren’t registered by the power meter so it probably reaches closer to 10W but regardless the CM5 isn’t happy and pi os displays the low voltage error. 

    However this only ever comes up when no battery is connected and that it’s directly and only powered by USB. When a battery is connected there are no low voltage issues.

  •   

    sorry just another query, I just came across the BQ25890 with its reverse boost over VBUS. I’m wondering about the maximum input current limit when I2C is not configured and battery isn’t connected. The datasheet seems to suggest 1.5A default (I have D+ D- shorted in my case), where as I think the BQ25895 has a limit of 2A? 

    additionally are the two ICs pin compatible so I can just replace part number and use the BQ258950 to access the reverse OTG over VBUS rather than over PMID?

    On a side note, the USB C VBUS input currently uses 5.1k resistors on the CC lines which should negotiate up to 5V 3A but I had the idea of connecting the CM5's built in CC1 and CC2 lines instead. This is able to negotiate upto 5V 5A and I believe is USB PD. Would this cause issues with the BQ25895 (it already is set to max ILIM)? Its just I feel using the Pi's CC lines are a bit more robust and allows for simple PD implementation.

  • Really really sorry for so many comments and additions but regarding TPS61022 is it acceptable to replace 732k resistor with 750k to get closer to 5.1V rather than a fixed 5V?

  • Hi, 

    For BQ25890,

    For Q1, Is charge enabled?

    For Q2, in charge mode, the max VBUS current is 3.25A, in OTG mode, if VBUS needs to be OTG output, please use BQ25890. The default VBUS OTG output current is 1.4A.

    It's better to start another e2e thread to discuss TPS61022 since this thread is assigned to me and Aurora in another product line might not notice the TPS61022 questions.

    Thanks,

    Ning

  • Hi! No I’m specifically interested in charge not enabled which would be the case since no battery is attached, is there a different input limit when a battery isnt attached? Do you think it would be worth using something like the TPS2121 to bypass the BQ25895 when on USB 5V anyways? 

  •   just a quick update I managed to fix the low power and undervoltage issue by slapping on a 47uF capacitor on the output 5V. Could you confirm if the capacitor values in the datasheets take into account DC bias or not? This otherwise matches up with ignoring DC bias reduction in capacitance which makes a lot more sense now. 

    additionally, would it still be possible to get a layout review? Thanks!

  • Hi,

    The capacitor values on the d/s are before derating. For more details about part selections, please refer to the corresponding EVM design BOM on the EVM User's Guide.

    Please visit https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/874557/faq-bq25890-faq-bq2589x-schematic-review-and-pcb-layout-design-tips for BQ2589X Schematic Checklist and PCB Layout Design Tips. 

    For the layout guidelines, please refer to https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/790169/bq25895-4-layers-board-review---agnd-gnd-and-power-planes  

    Thanks,

    Ning.