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.

TPS7A49: Soft-start time: how is it determined and tolerance

Part Number: TPS7A49
Other Parts Discussed in Thread: TPS7A30,

What is the tolerance of the soft-start time? I am working on specifications for design verification purposes and I need some reference (even a ballpark) for minimum and maximum. For TPS7A30 too, for that matter. The data sheet only has the formula tss(ms)= 1.4 * C(nF).

Also, what is the criterion for tss? Per graph on figure 20 of data sheet, the waveform is RC like. The graph is for 1.2V output and, based on it, the reference point would be 83.3% of Vout (1V). Is this always true?

In my design, output voltage is 14V for Vin 15.5V, Css = 33nF (so tss=46ms) and total capacitance on the output is around 300uF (distributed around two boards). Load current is of a few mA.

  • Hi Elder,

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

    Thanks,

    - Stephen

  • Hello, Stephen,

    Any update on this?

    Thanks,

    Elder.

  • Hi Elder,

    The NR pin is a "noise reduction" pin which also features as a soft start for the linear regulator.
    This function is common across other linear regulators as well.
    The concept is to reduce the noise of the linear regulator by filtering the internal band gap reference.
    In the PSRR plots the noise of the bandgap can be dominant at the lower frequencies, especially when it is amplified by the internal error amplifier.
    So TI places an RC filter between the bandgap and error amplifier to filter this noise.
    This RC filter slows down the turn on of the linear regulator as a result. 
    So the user has to make a judgement call on how much noise is acceptable vs how fast can the linear regulator turn on.

    For the TPS7A49:
    Figure 19 does not really show the RC nature because there is no external NR cap attached, and the ramp rate is very fast.
    There is no internal current source to increase the soft start time and transform this into more of a ramp, as opposed to an exponential turn on.
    Other components employ this additional current source.  See for example the TPS7A4701, which calls it a "fast charge" in the block diagram.

    Doing some quick math, and assuming 5 time constants gets us to roughly steady state on Vout, the linear regulator must be dividing down the reference by 23.5%.  1.2V * 23.5% * 5 time constants = 1.4.

    For the TPS7A30:
    Doing similar math: 1.2V * 15% * 5 time constants = 0.9.

    The tolerances have not been measured or characterized for the startup in these components.

    Thanks,

    - Stephen

  • Hello, Stephen.

    Thank you for your answer. I was aware of the noise reduction function but your explanation puts it into the SS context and makes it making more sense. I will probably use two or three time constants for the specification (facilitates measurements for an RC behavior). Also I will guestimate +/-30% for tolerance. It provides a ballpark to determine the circuit is within the expected limits. In my experience, most of the time the components are very close to the typical figures (though I always try to design for the worst case).

    Thanks.

    Elder.