What I understand is that the S7V7F5 voltage regulator doesn’t have any reverse voltage protection and I would like to know if you can recommend a solution to protect from a mistake in the connection?
Would you add a simple diode before the regulator on the vin pin?
If so, do you have a specific diode to recommend?
The problem with diodes is that you lose significant efficiency because of the forward voltage drop.
Instead, the best way to protect against reverse voltage is to use a p-channel power MOSFET, connected in the forward direction, with the gate connected to ground. When power is properly applied, the grounded gate will make the MOSFET conduct with very low resistance (you can find P-channel MOSFETs with < 10 milliohm resistance.) When connected reverse, the gate will be positive bias and not make the MOSFET conduct, and the body diode of the MOSFET will be reverse biased and thus not conduct.
So: + input goes to DRAIN of P-channel MOSFET.
SOURCE of P-channel MOSFET goes to the rest of your circuit, as "+."
GATE of P-channel MOSFET is connected to ground connector, which also goes to the rest of your circuit as “-.”
A good discrete power MOSFET in TO-220 form factor is the Vishay SUP75P03: digikey.com/product-detail/e … ND/2623113
It can take up to 20V input on the gate/source pair, can regulate up to 30V on source/drain, and can take up to 75A with proper heatsinking. It has a resistance of 10 milliohms at 4.5V, and 7 milliohms at 10V. The only draw-back is the price: $2.90 in singles!
I’m looking to use this regulator to power an Arduino derivative board (Ardupilot). The board will occasionally be plugged into and powered from a USB connection, while the battery is not connected, so there will be back-voltage on the regulator.
Can it handle this? I don’t want to put a diode in front of it because I want 5V Vcc.
Also, can this regulator handle a large capacitor in front of it? Something like 4700uF?
That is not something we have tested extensively, but applying a voltage to the output is probably ok as long as you do not have anything powering the input at the same time. (I suggest you cautiously test it yourself before counting on it to work that way.)
I think it will handle that okay. My main concern is that it could look like a short circuit on the output and trigger a protection shutdown, but it should not damage the regulator. Like with voltage on the output, I suggest you cautiously test it yourself.