Winfield Hill wrote:
You may have a better chance of surviving something like this with the NPN-PNP Vbe compensated buffer because the 317 current drops to zero and there is no stored charge in the output transistors. If the LM385 was a bit faster, something like this would survive- a variation on the standard precision current source with current source pulldown that is cut-off with shorted load- the PNP power can be saved with enough heat sink to carry it through the x msecs under non-short but abrupt drop in output voltage.
View in a fixed-width font such as Courier.
. .. . .. .. Vin >--+ .. | .. +------------+ .. | Vin | Vout = Vleds+3.3V .. | sw |---+--L1-+-----+----------+----------+----+ .. | | | | | | | _|_+ .. | | - | | | 3.9 --- .. | | D1^ | | LM385 1W | .. | | - | | ---/ | GND .. | | | 2.7K | // \---------+ .. | | GND | | --- | .. | | | | | |/e .. | LM2574-ADJ | | | | ,-| .. | etc | | | | |/e |\ 2N5195 .. | | | | +----| | .. | | | | | |\ | .. | | | | | '---+ .. | | | \| | BC856 | .. | | | BC846 |-------------------+ .. | | | e/| | | .. | | | | | | .. | | | | | | .. | | | | | | .. | | e\| | | D2 | .. | | BC856 |---+--------------|>|----+-->>--, .. | | /| | | 1N4148 | .. | | 1.23V | | | | .. | fb |---+-----' | |/ - .. | gnd | | +-10K-+--| BC846 LEDs V .. +------------+ 1K | | |\e - .. | | | \| | | .. | | | BC846|--+ | .. | | 4.7K e/| | | .. | | | | 560 | .. | | | | | | .. | | | | | | .. GND------+----------+-----------+-----+----+------------->>--' . ..