LTC3774

In LTC3774 datasheet page 16 regarding Inductor DCR sensing if we disable the SNSD+ pin will we:

  • Need a separate sense resistor?
  • Or A simple R1 & C1 RC filter on SNSA+ pin suffice?

Parents
  • Hello,

    If the SNSD+ pin is shorted to ground, a separate sense resistor is not required. An RC filter tied to the SNSA+ pin (R2&C2 in Figure 4) is all that is required. For applications using DCR sensing with SNSD+ disabled, keep in mind the following:

    • Vsense(max) will be 5x larger than the specified value. The inductor chosen should have a DCR high enough to provide an accurate current limit.
    • The RC time constant of the filter should now be set equal to L/DCR instead of L/DCR/5.

    Best regards,

    Mike

  • Hi Mike,

    In the burst Mode Operation, It is mentioned that Inductor Peak current is limited to 1/3 of maximum sense voltage irrespective even if voltage of Ilim indicates a lower value. Will this be over and above the above condition of Vsense max being 5 times the specified value?

    Regards,

    Karan

  • Hi,

    Can anyone please answer my query above regarding burst mode operation.

  • Hello Karan,

    The answer to your question is no. Here is the explanation:

    The peak current sense voltage during Burst Mode operation is clamped to approximately one-third of the maximum sense threshold even though the voltage on the ITH pin (not ILIM pin) indicates a lower value.

    If the DC path is disabled and the ILIM pin is floated, for instance, then the maximum current sense threshold will be equal to 20mV * 5 which comes to 100mV. During burst mode operation, the peak current sense voltage will then be equal to 100mV / 3 or 33mV. Once the load rises high enough, the converter will leave burst mode. When this happens, the converter will operate in continuous mode and the maximum current sense threshold will return to 100mV.

    Let me know if you have anymore questions.

    Best regards,

    Mike

  • Thanks Mike.

    Regarding 

    ''During burst mode operation, the peak current sense voltage will then be equal to 100mV / 3 or 33mV. Once the load rises high enough, the converter will leave burst mode. When this happens, the converter will operate in continuous mode and the maximum current sense threshold will return to 100mV.''

    I could not find this information in the datasheet. Can you point where the same is mentioned.

    Thanks

    Karan

  • Hi Mike,

    My output is having an oscillation of 6 Khz signal. Please see below waveform.

    Can you list the possible causes for this oscillation & where to look?

    Regards,

    Karan

  • Hello Karan,

    Here are some possible causes for the 6kHz oscillation.

    1. The 6kHz noise could be due to Burst Mode operation provided the MODE pin is floating and the load is light. If the LTC3774 is setup for Burst Mode, tie the MODE pin to ground. This will place the converter in forced continuous mode and should clear up the ripple. To answer your previous email, Burst Mode operation is described on page 11 of the data sheet.
    2. Noise is getting into the loop and causing the instability. Here are a couple of suggestions to make the design less susceptible to noise:
      1. Adjust the ITH compensation by increasing CTHP and reducing RTH. This will reduce the high frequency gain. Make sure CTH>=CTHP * 3 for adequate phase margin. The ITH compensation network is shown below.
      1. Choose an inductor with a smaller inductor value or higher DCR and re-calculate R2 and C2 in the DCR sense filter for the new inductor values. This will increase the current sense ripple voltage and improve the signal to noise ratio.

    Thank you,

    Mike

Reply
  • Hello Karan,

    Here are some possible causes for the 6kHz oscillation.

    1. The 6kHz noise could be due to Burst Mode operation provided the MODE pin is floating and the load is light. If the LTC3774 is setup for Burst Mode, tie the MODE pin to ground. This will place the converter in forced continuous mode and should clear up the ripple. To answer your previous email, Burst Mode operation is described on page 11 of the data sheet.
    2. Noise is getting into the loop and causing the instability. Here are a couple of suggestions to make the design less susceptible to noise:
      1. Adjust the ITH compensation by increasing CTHP and reducing RTH. This will reduce the high frequency gain. Make sure CTH>=CTHP * 3 for adequate phase margin. The ITH compensation network is shown below.
      1. Choose an inductor with a smaller inductor value or higher DCR and re-calculate R2 and C2 in the DCR sense filter for the new inductor values. This will increase the current sense ripple voltage and improve the signal to noise ratio.

    Thank you,

    Mike

Children
No Data