USB sourced circuit - how to protect from short ?

Thread Starter

2Hexornot2Hex

Joined Apr 16, 2020
54
Hey gurus,

Suppose I want to design a USB powered circuit (I actually do).
The circuit does XYZ (doesn't matter...some digital logic...caps...etc), it uses USB interface only as a power source (without data pins).

How should I properly protect the USB source from a short circuit ? (for example, due to some ceramic caps fail, bad design or even poor assembly [not intentionally of course...hehe] that eventually might lead to a short circuit)

USB interface - could be PC/laptop sourced...a short circuit might have a terrible and costly effect.

What's the best practice for this ? what's actually being implemented in the devices for those kind of failures ?

Thanks.
 

Deleted member 115935

Joined Dec 31, 1969
0
In theory, "ALL" USB power sources have current limiting protection, that by default is 500 mA at 5V.

But as is noted, cheap wall plug types might not be so compliant, and a PTC in your device is always a good idea, and very low cost, but its resistance might be a problem if your requirement is near 500 mA,
in which case a more active circuit with a MOSFET is called for
Many around,
such as
https://www.analog.com/en/products/ltc4085-3.html
 

Thread Starter

2Hexornot2Hex

Joined Apr 16, 2020
54
But as is noted, cheap wall plug types might not be so compliant, and a PTC in your device is always a good idea, and very low cost, but its resistance might be a problem if your requirement is near 500 mA,
in which case a more active circuit with a MOSFET is called for
Why its resistance might be a problem if the 500mA is the requirement ?
 

nsaspook

Joined Aug 27, 2009
8,897
In the paper, there's a schematic describing that all 4 USB wires are connected to the shield...a bit weird, no ?
Where else would you connect ESD Suppressors to? The shield is the drain point for ESD Suppressors as it is assumed to be a equal-potential (energy dumped into the shield won't be induced directly into the conductors signal path) point for non-signal energy.
 

cmartinez

Joined Jan 17, 2007
7,513
Most likely the source USB device will have internal short circuit protection for bad connections but you can always add your own for external device connections.
"most likely" ... yes, you're right ... but a few years ago I burned one of my laptop's USB ports (an HP laptop, btw) when I accidentally shorted it for about 10 seconds testing one of my circuits...

That particular incident confirmed one of my favorite quotes: "Assumption is the mother of all f-ups" ...
 

crutschow

Joined Mar 14, 2008
28,158
I quite like Jeri's circuit here
One possibly problem with that circuit (or any that uses the BJT base-emitter voltage as the current-limit point) is the series (shunt) resistor that must drop about 0.6V at the limit point.
That may too much of a drop for a 5V circuit to operate properly at lower currents.
 
Top