4
\$\begingroup\$

Time to trip of most of PTCs is about 0.1 seconds. In case of short between VCC and GND the current will be very high (maximum a power supply can provide) for at least 0.1s. Isn't it enough to burn out one of the components usually present in series with PTC: voltage regulators, power FETs, ferrite beads, etc?

To be specific: I have a device which is powered from USB bus via 500mA hold current PTC, 5V->3.3V LDO and ferrite bead. In case of short the current will be about 20-25A (this is what PC power supply can provide) for a period of 0.1s, afterwards a PTC of PC side or PTC on USB device will trip off power. Does LDO and ferrite bead survive 0.1 s of high current (and also does PC survive)?

\$\endgroup\$
  • 4
    \$\begingroup\$ Fuses (of any type) aren't really meant to protect sensitive electronics from overcurrent or overvoltage, but to protect against fire. \$\endgroup\$ – Nick T Nov 12 '10 at 13:39
  • \$\begingroup\$ I think the 0.1 seconds rating is the time to maximum off resistance. It will reduce the current in a significantly shorter time, both due to the resistance of the PTC and the device's characteristics. Also, the resistance of the PTC will also limit the current to less than 25A. \$\endgroup\$ – Thomas O Nov 12 '10 at 22:10
  • \$\begingroup\$ You will never, ever, get 20A if the device has a USB cable anywhere in the power connections. \$\endgroup\$ – Connor Wolf Nov 13 '10 at 0:29
2
\$\begingroup\$

I believe at 20-25A time for a PTC to trip is significantly less than 0.1s. From what I recall, 0.1s is generally the time to trip at 7A or less. In either case, voltage regulators usually have over current protection, and for the most part, a ferrite bead is just a wire. You should be fine for the most part unless you have some real sensitive components.

\$\endgroup\$
2
\$\begingroup\$

There are hundreds if not thousands of different PTC resettable fuses with a wide array of Time-Current Characteristics. Consider the Littelfuse 60R010:

  • It has a Hold-Current rating of 0.1A which is the "Maximum current the PTC will pass without tripping in 20C still air".
  • It has a Trip-Current rating of 0.2A which is the "Minimum current at which the PTC will trip in 20C still air".
  • The PTC's datasheet provides an Average Time-Current Curve which provides an estimated "time-to-trip" for different currents (all at 20C still air).

Also, fuses ARE indeed meant to protect sensitive electronics:

  • Cheap Chinese consumables may be designed to only prevent a fire, but these are throw-away devices. They are not exactly a "role model" for best practice design.
  • An embedded design engineer may utilize multiple resettable or non-resettable fuses within a design to protect every trace, passive component, IC, etc. from an overcurrent condition.
  • For aerospace designs and/or life-critical devices, fuses are employed to protect everything. A single wisp of semiconductor smoke can ground a plane (or a fleet of planes). A fuse may open a faulty circuit while allowing other parts of the circuit to continue functioning.

USB devices are not intended to pass 20A; however, current is always: I = V/R. If the voltage is high enough, or if the cable resistance is low enough, large currents can pass. USB cable is typically comprised of relatively small wire (i.e. 28AWG). At approximately 65milliohms per foot, a 1-foot 28AWG wire shorted to ground could result in 75A when connected to a 5VDC power supply.

NOW TO ANSWER THE ORIGINAL QUESTION: Will the LDO and ferrite bead survive the 0.1sec high current pulse?

  • Short answer: Yes, the LDO and ferrite bead will likely survive the pulse if the PTC trip-current is specified correctly

  • Longer Answer: The LDO may have an internal overcurrent protection circuit that protects it (check datasheet). The total input impedance of your circuit will likely limit the short-circuit current from exceeding 20A. Component manufacturers often provide pulse-withstanding data for their devices. They understand that inrush currents and short-circuit currents exist, and therefore test their devices to determine their maximums. If your components do not offer this data, try another manufacturer. A proper design should be able to use one or more resettable or non-resettable fuses to protect every component. Finally, a PC's USB port is already protected, as others here have suggested.

\$\endgroup\$
  • \$\begingroup\$ No, fuses are never meant to protect sensitive electronics. Fuses are slow, mechanical (/physical), cheap devices that are never designed to protect against a tight-tolerance failure windows. They are just there to prevent catastrophe. If you want to protect sensitive electronics, you need an active current measurement with the proper bandwidth and precision to protect against a fault. \$\endgroup\$ – user36129 Nov 25 '14 at 13:10
0
\$\begingroup\$

USB ports should be protected, so 25A even for short periods is unlikely.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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