Post by thread:[PIC] survivability of short circuited output pins

Matt Rhys-Roberts wrote: > I can't find anything in the datasheets about whether a PIC (say 18F > series) can withstand a short circuit on any of its pins when configured > as outputs. Clearly this would draw more than the 25mA spec limit, but > would it survive? Is there any known internal protection? it is not guaranteed to survive. People have reported chips to survive such an experiment, but no guarantees. Of course, there is no guarantee either that it will not survive. Or in which manner it will die. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

In fact the datasheet does indicate what will happen when (if) you exceed "Absolute Maximum Ratings" in this case 25ma / IO pin either sourced or sunk: *NOTICE: Stresses above those listed under “Absolute Maximum Ratings” may cause permanent damage to the device. This is a stress rating only and functional operation of the device at those or any other conditions above those indicated in the operation listings of this specification is not implied. Exposure to maximum rating conditions for extended periods may affect device reliability.* 2010/1/25 Matt Rhys-Roberts <spam_OUTmattTakeThisOuTnu-ins.com> > I can't find anything in the datasheets about whether a PIC (say 18F > series) can withstand a short circuit on any of its pins when configured > as outputs. Clearly this would draw more than the 25mA spec limit, but > would it survive? Is there any known internal protection? > > Thanks, > Matt >

Ariel Rocholl wrote: > In fact the datasheet does indicate what will happen Did you actually read that block of text before copying it? it says may not will > when (if) you exceed > "Absolute Maximum Ratings" in this case 25ma / IO pin either sourced or > sunk: > > *NOTICE: Stresses above those listed under “Absolute Maximum Ratings” may > cause permanent damage to the device. This is a stress rating only and > functional operation of the device at those or any other conditions above > those indicated in the operation listings of this specification is not > implied. Exposure to maximum rating conditions for extended periods may > affect device reliability.* >

Wouter van Ooijen wrote: > Matt Rhys-Roberts wrote: > >> I can't find anything in the datasheets about whether a PIC (say 18F >> series) can withstand a short circuit on any of its pins when configured >> as outputs. Clearly this would draw more than the 25mA spec limit, but >> would it survive? Is there any known internal protection? >> > > it is not guaranteed to survive. People have reported chips to survive > such an experiment, but no guarantees. > No gaurantees but if short duration shorts frequently resulted in failure i'd think we would know about it by now. During prototyping short circuits, misconnections etc are very common.

On Mon, Jan 25, 2010 at 11:42 AM, Matt Rhys-Roberts <.....mattKILLspam@spam@nu-ins.com> wrote: > I can't find anything in the datasheets about whether a PIC (say 18F > series) can withstand a short circuit on any of its pins when configured > as outputs. They do not specify correct operation outside of listed maximums. In other words, you can't find this text because they do not gaurantee it. > Clearly this would draw more than the 25mA spec limit, but > would it survive? It depends completely on the circumstances. CMOS naturally limits its own current under some circumstances (as the chip heats up, the junction resistance increases, which decreases the current, etc, etc). Further the chips are generally overengineered to some small degree so that process variations won't eliminate a whole batch of chips that end up at the wrong end of the variance. Some hobbyists use this to current limit LEDs - leave out the resistor, and just let the CMOS take the heat. > Is there any known internal protection? There is no intentional internal protection. Such protection is not needed most of the time, and costs more in terms of semiconductor size. Further, the protection needed depends on the situation - a simple short in a low current situation isn't hard to protect against, but a static shock takes more to protect it. How far do you go before that protection is enough, and should you make all your customers pay for it, when only a small portion of customers would actually need it? You should understand, though, that even though the chip will not always be destroyed by going outside the limits, it may be damaged in ways that are not entirely obvious. The point of static workstations is not because chips kept dying on the assembly line - they kept dying under 'normal' stresses out in the field. Turns out that static directly to the chip didn't kill it, but weakened it enough that it failed later. The chips are cheap enough that it's rarely worth your time to keep using a chip that's been abused - you'll waste far more time debugging a stupid problem with a bad chip when it finally fails, and you won't recognize the problem until you give up and swap with a new chip. -Adam

On Mon, 25 Jan 2010 16:42:53 +0000, "Matt Rhys-Roberts" <mattKILLspamnu-ins.com> said: > I can't find anything in the datasheets about whether a PIC (say 18F > series) can withstand a short circuit on any of its pins when configured > as outputs. Clearly this would draw more than the 25mA spec limit, but > would it survive? Is there any known internal protection? Hi Matt, In my experience, PICs can survive a longterm single pin short circuit pretty well at lower supply voltages. If you get over 4.5 volts I wouldn't count on it surviving. You will lose that pin, and maybe more of the chip as well. But an intermittent short during development is not likely to do it any damage. In a production device, never allow a short circuit by design! Cheerful regards, Bob -- http://www.fastmail.fm - Choose from over 50 domains or use your own

On Mon, 2010-01-25 at 16:42 +0000, Matt Rhys-Roberts wrote: > I can't find anything in the datasheets about whether a PIC (say 18F > series) can withstand a short circuit on any of its pins when configured > as outputs. Clearly this would draw more than the 25mA spec limit, but > would it survive? Is there any known internal protection? It's not speced to survive. In practice it often does. Since it's not speced to survive, assume it won't. TTYL

Where the data sheet says "may fail" in the absolute maximums section, it MAY fail catastrophically, in which case you may be lucky. It also might sort of work most of the time, or not. As is the nature of such things, it will probably fail at the worst possible time or during a demo test. If you exceed maximum ratings, there's no telling what will happen in the real world. Jon On Mon, Jan 25, 2010 at 3:54 PM, Ariel Rocholl <EraseMEforosspam_OUTTakeThisOuTarocholl.com> wrote: {Quote hidden}> 2010/1/25 peter green <plugwashspam_OUTp10link.net> > > > Ariel Rocholl wrote: > > > In fact the datasheet does indicate what will happen > > Did you actually read that block of text before copying it? it says may > > not will > > > > You're right, it is a language thing. It says "may", I am used to safety > critical requirementes where shall/will/may makes no difference. > > > > -- > Ariel Rocholl > Madrid, Spain > -

> No gaurantees but if short duration shorts frequently resulted in > failure i'd think we would know about it by now. During prototyping > short circuits, misconnections etc are very common. But thorough testing (including full temperature, power supply, and clock ranges) is very rare during prototyping, and problems are easily blamed on other issues. So no conclusions can be drawn except that the chip often appears to survive. -- Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products docent Hogeschool van Utrecht: http://www.voti.nl/hvu

Matt Rhys-Roberts wrote: > I can't find anything in the datasheets about whether a PIC (say 18F > series) can withstand a short circuit on any of its pins when configured > as outputs. Clearly this would draw more than the 25mA spec limit... Not necessarily this would draw enough current to kill by the current density. It may kill the output by power dissipation by even 25mA when it gets short-circuited under 5V or even under much lesser voltage when other outputs are heavily loaded.