• State Resolved
  • Locked Locked
  • Replies 12 replies
  • Subscribers 67 subscribers
  • Views 656 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.

BQ2057C: Part is getting "Hot" 60C during charging. Is this normal?

Brian Barrett
Prodigy 10 points
Part Number: BQ2057C
Other Parts Discussed in Thread: BQ2057,

I am using the BQ2057 to charge a 10Ah battery at 225mA charge current. The MOSFET is not getting too hot, but the BQ2057C is getting to 60C. Since I'm trying to pass this product through UL985, I have a limit of 1/2 Tstorage (or Tj if listed) as my max IC temperature. In this case 62.5C. This leaves me with barely any headroom on temperature. 

UB2 is the BQ2057C and I've tried to add as much copper under it as possible.

DB1 is my input diode supplying 5V to the BQ2057. RB6 is my high side current sense resistor and PFB2 is my MOSFET.

Ignore RB1 & PFC3, they are not connected in this configuration.

Here is my schematic.

Here is my simplified schematic. I have a uC that will ground the TS pin to turn off charging, but that is working as normal.

I think I have implemented this very close to the datasheet example. 

I can't see why the BQ2057C is getting hot at all. What power is it dissipating? Is there something I can do, besides reducing my charge current to lower the temp of the BQ2057?

Here is what the PCB looks like at steady state with thermal camera.

  • 0
    Prodigy 10 points

    A note from a Ti Engineer saying the Absolute Maximum Tj for the BQ2057C is 150C (as it's not listed) would also solve this issue for me. There is a temperature cut-off at 160C with 20C hysteresis built into the IC. This implies a max Tj, but since it's not listed, I can't use that as max Tj for UL. They use 50% max Tj or 50% max Tstorage if no max Tj is listed. I have contacted UL and a note from a Ti engineer stating the max Tj should be sufficient for me. 

  • 0 in reply to Brian Barrett
    TI__Guru**** 153116 points

    Hi Brian,

    I would not expect the IC to get that hot.  Does an oscilloscope plot so any of the IC pins oscillating?

    Regards,

    Jeff

  • 0 in reply to Jeff F
    Prodigy 10 points

    Hi Jeff,

    I've checked each pin on the oscilloscope and see nothing out of the ordinary. There are no oscillating signals.

    The CS pin is directly connected to an uC pin set as an input. The value changes as expected to indicate charging finished. I am not reading it as 3 state pin, but it's acting as expected.

    BQ2057 is also charging battery to100%. Charging won't start if battery is 4.1V or over as expected. Besides the heat, there seems to be nothing wrong in the way the circuit is functioning. I can enable/disable the charging by grounding the TS pin. Compensation is not being used and COMP pin is tied to VCC. VCC is about 4.6V which seems proper. The MOSFET is SI6423DQ which is almost identical to the MOSFET used in example application circuit in the datasheet. I can find no errors in my schematic or implementation on the PCB. That isn't to say there isn't an error, only I can't see one. I'm using the BQ2057CDGKR version of the chip. 

    I should mention I have a number of these boards all with the same behavior. It's not a single board with the issue, it's all of them.

    This temperature recorded is with the PCB installed inside an enclosure, but as you can see from the thermal image, the heat is very localized to the IC. I have confirmed the uC pin connected to STAT is set as an input. I can lift the pin and repeat the test to ensure that isn't the cause. 

    Has anyone heard of fakes of these chips? 

    Pin 1 VCC (>4.2V < 5V as expected)

    Pin 2 TS (1/2 VCC as expected)

    Pin 3 Stat (VCC as it's charging. expected)

    Pin 4 GND

    Pin 5 CC (battery voltage about 3.3V, this looks reasonable)

    Pin 6 COMP (tied to VCC)

    pin 7 SNS (0.1V across sense resistor)

    Pin 8 Vbat (3.X Volts. Charing as normal)

  • 0 in reply to Brian Barrett
    Prodigy 10 points

    I asked if anyone knew of any fakes. I should have given a picture of the IC I'm using.

    I've also disconnected the STAT pin to see if that makes any difference. I'll know in an hour or two.

  • 0 in reply to Brian Barrett
    Prodigy 10 points

    With the STAT pin disconnected the temperature still reaches 60C or more. The issue is not power being used from STAT Pin.

  • 0 in reply to Brian Barrett
    TI__Mastermind 23220 points

    Hi Brian,

    I see that you have the COMP pin in what appears to be the low-side current sensing configuration but I see that the SNS pin is still pulled for what may be high side current sensing. Is it possible to  connect the VCC pin to the COMP pin? 

    And just to confirm, the battery is charging?

    Best Regards,

    Anthony Pham

  • 0 in reply to Anthony P Pham
    Prodigy 10 points

    Hi Anthony,

    I think it's just my schematic is a bit unclear. VCC and COMP pin are both connected together as they have the same net label 5V_Charge.

    I'll have another careful look though and make sure it's consistent with high-side current sensing configuration.

    Thanks for the suggestion.

  • 0 in reply to Brian Barrett
    Prodigy 10 points

    I can't find any error in the schematic. I received new chips today from an alternate supplier. The marking for pin 1 is different. The font might be different as well. Regardless, even after switching the ICs the temperature is still the same. Just about 60C.

    I'm still at a loss for why this chip is getting hot. Hot implies significant Pd but there should really only be the Pd from the gate resistor on the MOSFET? which is 1k. The stat pin is floating, so nothing going there. 

    If anyone can suggest any tests to perform, I'd be appreciate it.

    Also if there are any Ti Engineers that can comment, if we could confirm this chip has Absolute Maximum Tj of 150C (140C would be OK too), I can still pass my certification. The way the datasheet lists only Tstorage and states an automatic shutdown temperature of 160C I don't have a specific number for maximum Tj, so UL has to use the listed Tstorage instead of Tj making the temperature too high to pass or at best very marginally acceptable.  

  • 0 in reply to Brian Barrett
    TI__Mastermind 23220 points

    Hi Brian,

    Brian Barrett said:
    I received new chips today from an alternate supplier. The marking for pin 1 is different. The font might be different as well. Regardless, even after switching the ICs the temperature is still the same. Just about 60C.

    This is interesting, can you share what supplier you're getting these from? 

    Brian Barrett said:
    I think it's just my schematic is a bit unclear. VCC and COMP pin are both connected together as they have the same net label 5V_Charge.

    I see this now, should be no problem. Thanks for that mention.

    Brian Barrett said:

    I'm still at a loss for why this chip is getting hot. Hot implies significant Pd but there should really only be the Pd from the gate resistor on the MOSFET? which is 1k. The stat pin is floating, so nothing going there. 

    If anyone can suggest any tests to perform, I'd be appreciate it.

    I'll need to look into what tests are possible. I've reserved some time tomorrow to look into tests on our end to see if we see similar behavior

    Brian Barrett said:
    Also if there are any Ti Engineers that can comment, if we could confirm this chip has Absolute Maximum Tj of 150C (140C would be OK too), I can still pass my certification

    We are currently looking into this, apologies for the delay.

    Best Regards,

    Anthony Pham

  • 0 in reply to Anthony P Pham
    Prodigy 10 points

    Hi Anthony,

    Thanks. I'm looking forward to hearing your results.

    Regarding the supplier, due to chip shortages and long shipping times, I've been buying through various local sourcing agents. I did see today Mouser has some stock and will buy a few from there as well.

  • 0 in reply to Brian Barrett
    TI__Mastermind 23220 points

    Hi Brian,

    Unfortunately, we're not seeing the same behavior on our end so we're still exploring possibilities on this high temperature coming from your part. 

    On some of our test boards, the device (when configured for 250 mA) doesn't seem to warm up. Like you expected, the FET / BJT heats up a lot more than the BQ device.

    Would you be able to drive the TS pin with a voltage outside of the VTS1 and VTS2 range? This would disable charging and we can see without charging the battery how much power the device is consuming. 

    Best Regards,

    Anthony Pham

  • 0 in reply to Anthony P Pham
    Prodigy 10 points

    Hi Anthony,

    I found the problem. There is no issue with the BQ2057C. Diode DB1 has a voltage drop of 0.7V @ 288mA so it's dissipating about 200mW in an SOD123 package. The proximity of diode to the BQ2057C combined with the relatively low resolution of my thermal camera sent me down the wrong track. It was the diode getting hot and not the BQ2057C.

    Your test and confirmation that the BQ should not get hot was very helpful. I did test the power consumption with the TS grounded and it was in fact 0.

    I tested the current at the benchtop power supply and into the battery. They were identical (+/- error on the benchtop power supply current reading). It didn't make sense that the BQ was dissipating any power. A few more voltage readings and I saw the large voltage drop across the diode (MBR160) and it clicked.

    Thanks again for the help and I hope my simple error here is helpful to someone else in the future. 

    Regarding my UL certification, since it's the diode getting hot and not the BQ2057C the temperature is acceptable as the Tjmax of the diode is 150C. I might still swap it for a SOD123F or bigger but ideally the Vf is still 0.7V so the Pd in the MOSFET does not increase. 

    PCB together with only the BQ2057C, MOSFET and other required components.