The datasheet for the TPS7A4101 shows an example circuit for driving an LED string (enclosed below. I have previously inquired whether the regulator must have a cap on the IN pin for stability reasons. The answer was yes - the datasheet omitted the cap on the IN pin in the illustration. If there is an input cap (Cmin >= 1uF) then, when the board is powered up and power is applied across the LED chain there will be a large current spike through the LEDs as the input capacitor is charging. This current spike can be tens of amps - but of course short duration. This violates pretty much every LED specification out there. The LED datasheets will generally specify a Maximum current rating, such as, Imax = 100mA, t < 10us.
We used the TPS7A4101 as shown in the datasheet to drive LEDs. Now, one of our LED chains destructed.
If TI insists that there must be an cap on the IN pin for the stability of the device, how does TI recommend to get around the current spike that results from charging up the input cap so that the LED specifications are not violated?