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.

BQ51013A gets way too hot under moderate load

Other Parts Discussed in Thread: BQ51013, BQ51013A, BQ51013B, BQ51013BEVM-764, TIDA-00318

We're using a BQ51013 to charge a lithium battery.  At 500mA coming out of the BQ51, we measured the chip at over 140F.  The BQ51 is on a small device and it's heat soaking everything, including the battery.  This has been causing batteries to bulge.

Dropping the charge current down to 250mA reduces the BQ51013's temperature to 125F, but that's still too hot for the battery.

What can be done to reduce the BQ51013 temperature?  It's heating up our entire device.

  • Curtis,

    There is no reason the bq51013A should create that much heat.  What package are you using?

    For unrelated reasons, I would suggest moving to the bq51013B.  The bq51013B meets the WPC v1.1 standard (versus the bq51013A meeting v1.0).  The key difference is how the devices handle foreign objects.  They are functionally equivalent and in the same footprint.

    How big is the board you are using?  The bq51013BEVM-764 is a large PCB which has excellent heat sinking capability.

    I've taken thermal images on multiple boards and have not seen 140F (60C) at 500mA.  See TIDA-00329 for an example of a very small (<30mm2) with poor thermal capability.  At 300mA, the IC temperature is 46.9C (116.4F).  The TIDA-00318 is a larger board (5mm x 15mm) and has much better thermal characteristics.

    The bq51013B datasheet section 11 reviews board layout recommendations, including thermal performance.

    Measure the voltage at the RECT pin and at the OUT pin,  This will be the area with the majority of the power dissipation.  

    Regards,

    Dick

  • And that's what I get for net being able to take accurate measurements. We had a transistor sitting right next to the bq that was misbehaving and dumping a metric crapton of power into the board. We got a thermal camera and that showed the bq getting absolutely soaked from the transistor.

    oops.
  • Curtis,

    Thank you for the reply. Glad to hear you have solved the heating issue - and glad it's not the bq51013B.

    Regards,

    Dick
  • Scratch that.  We dropped the charge current down some more, and the BQ quickly became the hottest component on the board at 125F, and it's heat soaking the battery.  The BQ current output is 100mA.  

    Is there a more efficient low current charging chip available?

  • Curtis,
    What package are you using? The Thermal information in the bq51013B datasheet, section 7.4, may be helpful.
    If you are using the RHL (VQFN) package the Junciton-to-ambient thermal resistance is 37.7C/W (99.9F/W). At 100mA output current, the estimated power dissipation would be (VRECT-VOUT) * 100mA. Depending on your charge current setting via ILIM, the RECT voltage could be as high as 7.1V (see figure 3 in the bq51013B datasheet). If the battery voltage is 3.5V, then you get (7.1-3.5)*0.1 = 0.36W. Using the 100F/W from above that's a rise in temperature of 36F. If you're using the CSP, it would be higher.
    Copper for heat dissipation is always a good idea.
    If you're seeing 125F, there may be other issues going on. The Cs capacitors (AC1) could also be a heat source.
    Regards,
    Dick