• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 63 subscribers
  • Views 495 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

LP8860-Q1: Unexpected Thermal Shutdown

Martin Schad
Prodigy 60 points
Part Number: LP8860-Q1
Other Parts Discussed in Thread: CSD18543Q3A

Hello Team,

We have designed a backlight controller based on the LP8860-Q1, and we are quite
satisfied with most of the features of the device currently. But we have observed a
problem with "thermal shutdown" probably.

We have programmed the Device EEPROM according to our requirements, which is:

- 1 LED string (1+2+3+4) with 150 mA each (600 mA total)

- Adaptive mode enabled (startup voltage 24.0V)

- Hybrid dimming enabled, PWM dimming threshold is 12.5%

- No external NTC is used

- Thermal de-rating function disabled (EEPROM Reg. 24, INT_TEMP_LIM[1:0] = '00')

The supply voltage of the boost regulator is 12.0V and VDD is 5.0V, Switching frequency
is set to 800 kHz (spread spectrum enabled).

The programmed values are (from register 0x60 to register 0x78):

    0xF1,0xFF,0xDA,0xFB,0xDF,0x04,0xF2,0x77,
    0x77,0x71,0x3F,0xB7,0x17,0xEF,0xB0,0x8F,
    0xCE,0x85,0xE5,0xD0,0x35,0x06,0xD8,0x88,
    0x3E

Some of the EEPROM registers have "reserved" bits, and we don't mask these bits during
programming. We have derived the proper values of these bits from your LP8860-Q1EVM
development platform and your LP8860 programming software (which can generate
a EEP-File). That's why some of the reserved bits are programmed with '0' and others
are programmed with '1'. After programming, we use a verification function to make sure
that the data is written correctly (which also includes the reserved bits).

We mounted our backlight controller PCBA into our HMI device (which also includes
the 15-Inch LCD and a ARM-based compute unit), and have put it into our climatic chamber.

With a serial debug interface, we have monitored the on-die temperature of the LP8860-Q1
(Temp MSB/Temp LSB registers), and the status of the fault registers.

Below 69°C Ambient temperature, the backlight circuit works quite well. But at about
69°C (LP8860 die temperature about 91°C), the backlight switches off for some reason.
If we reduce the ambient temperature slightly, the backlight switches on again. No fault
was detected.

But we have set INT_TEMP_LIM[1:0] in EEPROM register 24 (0x78) to '00', therefore the
thermal de-rating function should be disabled, and we have clearly not reached the
critical upper on-die temperature limit of about 165°C.

Do you have a glue why the backlight will switch off in our case? May it have something
to do with the setting of the reserved EEPROM bits?

Best Regards,

Martin

  • 0
    TI__Genius 15776 points

    Could you offer your schematic

  • 0 in reply to Frank Qiu
    Prodigy 60 points

    Hello Frank,

    Attached you will find the schematics.

    The LCD backlight is connected to LED-P (Anode) and LED-N (Cathode). We have measured about 18.3 V from LED-P to PGND, which is well above the
    minimum output voltage of the LP8860-Q1 (16.0V). The value was measured at about 69°C ambient temperature (climatic chamber).

    BL-PWM-CY is connected to pin 18 of LP8860 (external PWM input), but it is not used (tied to GND). We use the I2C-bus to control analog/PWM dimming.

    The supply for the gate drive of T2 (CSD18543Q3A) is 5V (5V0-FC), and the charge pump of the gate drive is disabled.

    An external resistor set is provided to adjust the LED current, but we use the internal setting (150 mA per LED current source).

    "JP1" is a field of 8 vias on the PCB, which connects the GND layer and the PGND layer of the PCBA. It is placed near to the current sense resistor R50
    and near to the LP8860. The thermal pad of the LP8860 is connected to GND.

    Best Regards,

    Martin.

  • 0 in reply to Martin Schad
    TI__Genius 15776 points

    Hi Martin,

    From your describe, it seem like thermal shutdown, but you as you say the temp doesn't reach the thermal limit, could you help me to read the resistor 0x0F when the switch stop working. I want to make sure the issue first.

  • 0 in reply to Frank Qiu
    Prodigy 60 points

    Hi Frank,

    After some measurements and Software debugging, we found the reason for the "thermal shutdown".

    It was not related to the LP8860.

    We use an opto-isolated undervoltage circuitry to disable the backlight when the main supply voltage
    (generated by an isolated power supply, which is not shown in the schematics) drops below a certain
    value.

    The output transistor of the opto coupler (collector pin + pullup resistor) are connected to the gate of
    a N-MOSFET, and the MOSFET drain pin (and second pullup resistor) was connected to a GPIO Input
    of a Microcontroller, which should detect the power fail condition and should send the appropriate
    commands to the LP8860 via I2C bus.

    But at higher temperature, the V(CE) voltage of the opto coupler has increased significantly. It has
    reached Vgs(th) of the MOSFET, which has turned on inadvertently. The Microcontroller has detected
    a false power fail condition in that case, and has sent the wrong commands to the LP8860.

    The issue was therefore related to the CTR of our opto coupler. An increased optocoupler LED
    current  has solved the problem, and keeps V(CE) well below V(GS,th) of the MOSFET, even at
    high ambient temperatures.

    Thank you for your assistance,

    Best regards,

    Martin.