Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 64 subscribers
  • Views 505 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.

LP5907: Partially on state

Brent Allen
Prodigy 100 points
Part Number: LP5907

I have an LP5907MFX-3.3 deployed on a design and am experiencing odd behavior where the regulator turns slight on with an odd waveform, and after a period of time turns on fully, regardless of the voltage on the enable pin.

The power is supplied by a single cell LiPo battery, charged by an MCP73811 charge controller via USB. The regulator is connected directly to the Vbatt net due to the lack of power-path management. The enable line of the regulator is driven by a latching power circuit detailed here: mosaic-industries.com. A schematic snippet detailing the power circuitry is shown below:

The rest of the system includes a microcontroller, bluetooth module, and I2C IMU. The odd behavior occurs after the battery charges fully, and results in a ~1.7V stepped waveform on the 3V3 net that results in brownout of the MCU. The enable pin is consistently held low at an average voltage of 140mV, and adding an additional 10kΩ pulldown (or even shorting to ground with a jumper wire) didn't change anything. A scope capture of the 3V3 net is shown below:

When the enable line is held low constantly what would cause the output to rise so high and fluctuate so much? Shouldn't the 230Ω internal discharge resistor on the output, activated when the enable line is low, prevent this kind of behavior?

*EDIT* Additional scope captures follow below:

Another scope capture is below to show the three most relevant signals: Enable (yellow), Vbatt (fuschia), and 3V3 (cyan). Additionally, the nPWROFF signal has been physically disconnected from the MCU to ensure it's not influencing the circuit.

Another scope capture, this time showing the shutdown behavior when the power switch is held. Signals are Enable (yellow), Vbatt (fuschia), and 3V3 (cyan).

  • 0
    TI__Guru 55940 points
    Hi Brent,

    Thank you for the detailed description and scopeshots. In particular the last two scopeshots showing all of the pertinent rails are very important for us to help debug remotely.

    I agree that it is odd that when you drive the EN signal low that you are seeing a signal on the output. The loads that you mentioned often require multiple input rails. Are you able to isolate the LP5907 from the loads in your application and repeat the last two scopeshots? This will help check if another voltage is leaking onto the output.

    Very Respectfully,
    Ryan
  • 0 in reply to Ryan Eslinger
    Prodigy 100 points

    Hi Ryan,

    Thanks for the reply. Your mention of 'voltage leaking onto the output' got me thinking about which nodes have the potential to be powered when the latching power circuit is switched off. The only two I could think of were the battery voltage monitor (connected to an MCU analog input through a 1/2x voltage divider) and the charge status monitor STAT1 in the schematic.

    I should've thought of the latter sooner when I had determined the issue occurred only when charging terminated - note the STAT1 output in table 5-1 shows 'H' for 'Charge Complete - Standby'. When plugged in via USB and charging completes, that pin goes to Vbus (~5.2V). After slicing that trace, the system behaves perfectly.

    Beyond causing this undesirable behavior, connecting STAT1 directly to the MCU is also potentially damaging for the MCU as it's not rated as being 5V tolerant.

    I imagine the best course of action would be moving the LDO upstream of the latching power circuit. Another option might be adding a diode between the MCU and battery charger such that the charger can sink current from the MCU to pull the pin low, but cannot pass current back into the MCU's pin.

    Thanks again for setting me on the right track.