# Short to GND protection

#### krakajax

I am designing a 500mA source circuit using a npn and pnp transistors. How can I protect the pnp if the output is shorted to gnd. Both transistors are in common collector configuration.

Related Electrical Engineering News on Phys.org

#### Rogue Physicist

A fuse is a good start, in series with the output leads, to prevent excessive current flow:
Look up the speed of the fuse, and do some tests too: Sometimes its embarrassing for the special circuit to cleverly protect your fuse by hari kari.

If you can't get a fast enough reaction time, you can try fast electronic switches which require a voltage to stay on (like a transistor in series with your output). If the lead is grounded, then the transistor or switch shuts off, protecting the rest of the power supply from current damage.

#### berkeman

Mentor
krakajax said:
I am designing a 500mA source circuit using a npn and pnp transistors. How can I protect the pnp if the output is shorted to gnd. Both transistors are in common collector configuration.
The way you usually do short circuit protection is with a low-value current sensing resistor and feedback from that sensing resistor to spoil the input drive to your output circuit. You want to use a small enough value resistor so that it doesn't rob much power from your output capability, but large enough that the voltage across the resistor in overcurrent conditions is large enough to sense reliably. In fact, depending on the topology and complexity of your circuit, you will usually use a diode drop voltage across the sense resistor as the threshold where you start feeding back a signal to spoil the input drive.

So if you have a high-side PNP transistor that is sourcing current to a ground-referenced load, you will put a resistor in series with the collector output of that PNP transistor, and sense the voltage drop across it to tell you what the current is. Or even better (since the collector voltage will move around), you can put the current sensing resistor between the emitter of the PNP and the +V supply. Size the resistor so that you get a 0.6V drop across it when the current limit should kick in, and use that diode drop voltage to turn on another PNP transistor that has its emitter connected directly to +V and base connected to the junction between the current sense resistor and the emitter of your main output PNP. The collector current from the sensing PNP can then be used to spoil the drive for the main output PNP.

This simple diode-drop technique can only be used if it works for you over temperature. The base-emitter diode turn-on voltage will vary at -2.1mV/degreeC, so you have to size the current sense resistor to turn on the overcurent protection PNP not too early at high temperature, but at low temperature it has to turn on in time to protect the output PNP. If you need a tighter window on your current limit, you can add temperature compensation into the Vbe sense circuit, or you can go with a full-blown diffamp implementation of the current sensing where you are only left with the tempco on the sense resistor.

I hope that helps. -Mike-

Last edited:

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving