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.

BQ24600: is too hot to touch but charging 4S 1850 @ 1A well from 20V supply

Part Number: BQ24600

We have a design using BQ24600 which is working well except the chips is getting surprisingly hot (yes, we have thermal pad with vias and 4 layer PCB with copper ground plan on mid-layer 1 plus copper on other layers).  From the block diagram and sample circuits, we can tell that the BQ24600 has no internal power transistors for "Power Conversion" (except gate drivers to drive external MOSFETs which do the power conversion) so we feel there is no reason the BQ24600 should get so hot.  We are using a very 'standard' configuration seen in sample circuits provided by TI in datasheets and we had additional support from Steven Bangdiwala (Thanks Steve for helping out Carlos a few weeks ago).  We think that the 20V supply we are using, to charge 4S 18650 Li-Ion, requires the BQ24600 internal 6V and 3.3V linear regulators to have significant voltage drop and therefore significant head dissipation.  Our design will operate in a harsh environment (even during charging at times...) so even though we are happy with the performance on the bench we cannot justify (to our client) the surprisingly high heat the BQ24600 is dissipating when charging at room temperature and ideal conditions.  We have TI LED drivers running 4A @ 8+ Volts into LEDs which run cool to the touch and these LED drivers have internal MOSFETs.  I think this should help make our point...  This makes driving the gates of the external MOSFETs at higher voltages seem reasonable (hmmm,  too bad those internal LDO's don't have a pin to adjust their voltages)...

We are wiling to use external linear regulator (or SMPS for that matter) to drop to 9V in an effort to cut the power dissipated in the BQ24600 down by half or take other recommendations from TI's team.  We have a 10 Ohm Resistor  with capacitor (RC Circuit) in place for VCC which is shown in example schematics.  If we knew more about the expected current draw for VCC we could better speculate what we would need to design to improve heat dissipation without causing other problems.

Any recommendations?  Or is this typical?

  • Root,

    The power dissipation on the IC is mainly resulted from LDO, gate driver to drive the FETs with high frequency. PCB needs to have enough copper areas for the heat to be dissipated. The power loss from the power stage also needs copper areas to dissipate. Reducing the input voltage can help reduce both charger IC and power stage (MOSFETs) switching power loss. We don't have quantified data for this as the results are also related to the MOSFETs' gate charge. Experimental tests to check the current to the VCC and the efficiency of the power stage with different VCC and input voltage on your board should be the most straightforward way to find the optimal point.