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.

  • Resolved

TPS7A49: Output Voltage is High

Prodigy 20 points

Replies: 3

Views: 59

Part Number: TPS7A49

Hi,

We are using the TPS7A49 to regulate 20V to 18V for a low noise supply. However, when testing the circuit on the PCB, the output voltage is too high (measured 19.3V). After reading other threads, we determined the cause to be from leakage current into the NR/SS pin. Our layout has the 20V input routed very close to the noise reduction capacitor. In this revision of our board, it's not possible to add guard traces to prevent the leakage current.

The datasheet mentions in the "Do's and Don'ts", to not resistively or inductively load the NR/SS pin. However, through experimentation, we noticed that adding a 10M resistor in parallel across the noise reduction capacitor brought the voltage on the output back to the expected amount.

I would like to know more details about what is meant by "Do not resistively or inductively load the NR/SS pin", and if our solution will work reliably.

Thanks,

Judd

  • Hi Judd,

    I am reviewing your question and will respond within 2 business days.

    Thanks,

    - Stephen

  • In reply to Stephen Ziel:

    Hi Judd,

    In modern linear regulators the internal noise at low frequencies is dominated by the internal reference voltage.
    This is because noise generated by the internal bandgap reference is gained up by the low frequency gain of the internal error amplifier.
    To improve this, a low pass filter is inserted between the bandgap reference and the internal error amplifier.
    The low pass filter is usually an RC filter, with an external capacitor called the NR/SS capacitor used as the filter cap.

    By loading this node down with an external resistor, you are generating a voltage divider and this could change the setpoint.
    The datasheet is silent on the internal series resistor, but using math I get 700k ohms series resistance based on your output going from 19.3V to 18V with a 10Meg pulldown.
    Adding inductive loops or inductance can result in noise on this node, which defeats the purpose of the filter and can significantly reduce the noise benefits of the LDO.

    I located the previous thread which suggested leakage as a possible culprit in higher than expected output voltage.
    https://e2e.ti.com/support/power-management/f/196/p/727848/2692410?tisearch=e2e-sitesearch&keymatch=TPS7A49#2692410

    In your application, have you confirmed that the PCB board is clean and free from contaminants?
    I was able to take a TPS7A49 EVM, saturate it with isopropyl alcohol, and confirm the output voltage was significantly higher than the designed setpoint.  As the alcohol dried over a minute or so of time, the setpoint fell until it locked onto the typical steady state voltage.  This demonstrates the impact that an unclean board can have on the operation of the device.

    Thanks,

    - Stephen

  • In reply to Stephen Ziel:

    Hi Stephen,

    After some more experimentation I can confirm there was some leakage on the board, either due to contaminants or from the nearby traces at +20V. After removing the original IC, lifting the EN pin on a new IC and soldering it to the cleaned PCB, the chip operates within spec. We will make the appropriate changes to the board layout in the next revision.

    Thanks,

    Judd

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.