1
\$\begingroup\$

I'd like to implement a safety mechanism in my circuit to avoid overvoltages/overcurrents in the event of a short, but am unsure as to what what the most robust or appropriate design approach would be. The primary rail is +15V and I want to limit the power to 15W.

After looking around, it seems that this should be pretty straightforward with discretes (e.g.link here), but would this be robust enough? Are there tricks/pitfalls in scaling as necessary?

I thought that there may be ICs for what I'd imagine is a common problem, but I'm unsure of what the appropriate one would be: inrush current limiter? surge suppression IC? power NTC thermistor? Something else entirely?

Or perhaps the simplest solution is a fuse, but that raises other questions as well. Will it trigger quickly enough? Does it need to be resettable?

Any help steering me in the right direction would be greatly appreciated.

\$\endgroup\$
2
  • \$\begingroup\$ Can you tell us more about the load? Would help to know what kind of protection is more or less appropriate. \$\endgroup\$ Commented May 6, 2015 at 6:28
  • \$\begingroup\$ Definitely - the load is a power button switch and an array of LEDs. \$\endgroup\$
    – EE_padowan
    Commented May 6, 2015 at 15:13

2 Answers 2

1
\$\begingroup\$

My advice is to use a smart high side switch whenever you can... search for that on google... It really saves time and simplify your design. It's basically a mosfet that is current limited and usually fully protected (over-temperature,short circuit, loss of GND, ESD protected,over voltage,reverse polarity...)

It's surely robust enough if that's your concern... you can check the datasheet and look for the "protection timing diagram" for knowing the exact behavior in case of overload or short circuit

\$\endgroup\$
0
\$\begingroup\$

The link seems satisfactory. The main thing to look out for is power dissipation of the current limit circuit during a fault. A non foldback limit will dissipate the full 15W in the event of a short. A value of 0.6 ohms is OK, if you parallel two 1 ohm resistors you will get 0.5 ohms which is near enough. Theses funny values are normal for current protection.

\$\endgroup\$
2
  • \$\begingroup\$ Hello, and welcome to the stack. This is not the sort of answer we typically recommend giving - it's quite simply far too short. What you have written is more of a comment, and doesn't go any further into the question's other concerns. When writing an answer, take the initiative to be a teacher - the purpose of the stack is to generate and store detailed knowledge that will be useful to lots of people. Take a look at a few of our top questions and answers to see what I mean. \$\endgroup\$
    – user39962
    Commented May 6, 2015 at 3:23
  • \$\begingroup\$ Despite seeming straightforward, I'm not quite sure how to implement this. Running the numbers with 0.6V for the diode turn on voltage and a max current of 1A, I'm left with a curious Rsense value of 0.6ohms. What's the best way to go about increasing this to a reasonable value? Would adding a third diode between the BJT base and the output/sense resistor raise this ceiling? Also, I don't understand the role of the Reference Diode. \$\endgroup\$
    – EE_padowan
    Commented May 6, 2015 at 16:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.