Post by thread:[PIC]: 12C509JW weirdness

Hello, I just had a weirdness happen to me. I have a single 12C509JW and was using it today for debugging. After 3 burn-erase cycles I grabbed the chip without observing ESD practice and it started to show dead (my programmer reports 'no chip in socket' for this failure). I scoped the pins and GPIO3 (Vpp) rose with a spike to Vpp and then fell down immediately to 3V (the programming Vdd is 5V), tripping the current limiter in the programmer. I tried this a few times, then I gave up and put it in for a 30 minute erase (normal is 13 minutes). Still the same. I was about to chuck the chip when I did the following last resort attempts: I used a FLUKE DVM on the diode scale to make comparative measurements between Vpp and other pins (esp. GND) in both directions. I could not find anything, but Vpp was not shorted so I set the benchtop supply to current limit and applied gnd to Vss, GP0, GP1 and the supply to Vpp and slowly cranked up the voltage. To my surprize it went all the way to 13.2V without tripping anything. Then I put the chip in the programmer (hoping for a miracle ;-), and it worked (so far). This is one of those weird charge injection ESD failures probably. Has anyone got anything to say about this ? MCLR/Vpp is the only pin on a PIC that has no bulk diode to Vdd. The failure occured when I touched the Vpp wire first (the pic was in a jig with ICSP connector on short wires and I grabbed it by the wires). The place has no ESD measures but it was a wooden desk (not a good insulator), humidity 60% and usually there are no problems. Do you people take special measures with this ? Peter -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Peter, my guess is that you fried (short) the ESD diode on that pin, then when cranking up your 13v power supply on the pin it melted the dead diode to open circuit again. I see this happen all the time in TVs, normal semi fault is S/C, then if it cops more abuse it goes O/C. What current limit did your PSU have when you cranked up the voltage? Was there a current meter? Either way you have to consider this part as suspect for the rest of its career. :o) -Roman Peter L. Peres wrote: {Quote hidden}> > Hello, > > I just had a weirdness happen to me. I have a single 12C509JW and was > using it today for debugging. After 3 burn-erase cycles I grabbed the chip > without observing ESD practice and it started to show dead (my programmer > reports 'no chip in socket' for this failure). > > I scoped the pins and GPIO3 (Vpp) rose with a spike to Vpp and then fell > down immediately to 3V (the programming Vdd is 5V), tripping the current > limiter in the programmer. I tried this a few times, then I gave up and > put it in for a 30 minute erase (normal is 13 minutes). Still the same. I > was about to chuck the chip when I did the following last resort attempts: > > I used a FLUKE DVM on the diode scale to make comparative measurements > between Vpp and other pins (esp. GND) in both directions. I could not find > anything, but Vpp was not shorted so I set the benchtop supply to current > limit and applied gnd to Vss, GP0, GP1 and the supply to Vpp and slowly > cranked up the voltage. To my surprize it went all the way to 13.2V > without tripping anything. > > Then I put the chip in the programmer (hoping for a miracle ;-), and it > worked (so far). > > This is one of those weird charge injection ESD failures probably. Has > anyone got anything to say about this ? MCLR/Vpp is the only pin on a PIC > that has no bulk diode to Vdd. The failure occured when I touched the Vpp > wire first (the pic was in a jig with ICSP connector on short wires and I > grabbed it by the wires). The place has no ESD measures but it was a > wooden desk (not a good insulator), humidity 60% and usually there are no > problems. Do you people take special measures with this ? > > Peter > > -- > http://www.piclist.com hint: The list server can filter out subtopics > (like ads or off topics) for you. See http://www.piclist.com/#topics -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Roman, the supply was limited at 50mA it has an overcurrent light and a LCD meter. Meanwhile I have played with the chip some more and it appears to have a 200K resistor (?!) between GP3 and GND although it remains programmable and works (except GP3 which is apparently hosed - no GPPU, no GPWU, and the Schmitt level is somewhere between Vdd and Vss, not necessarily at the same spot every time I measure it). I am thinking about soldering a 15V tranzorb between GP3 and Vss on the socket ICSP jig I'm using for development to prevent further surprizes. This is the first PIC lost to ESD in ~5 years. Any suggestions ? Peter -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email spam_OUTlistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

Peter L. Peres wrote: > > Roman, the supply was limited at 50mA it has an overcurrent light and a > LCD meter. > > Meanwhile I have played with the chip some more and it appears to have a > 200K resistor (?!) between GP3 and GND although it remains programmable > and works (except GP3 which is apparently hosed - no GPPU, no GPWU, and > the Schmitt level is somewhere between Vdd and Vss, not necessarily at > the same spot every time I measure it). > > I am thinking about soldering a 15V tranzorb between GP3 and Vss on the > socket ICSP jig I'm using for development to prevent further surprizes. > This is the first PIC lost to ESD in ~5 years. Any suggestions ? Definitely sounds like you fried GP3. 50mA would be enough if GP3 was already fused and damaged. I would think that if you can't insulate the connector on the ICSP setup so you can't touch it, then maybe adding a 100k or 56k resistor from each pin to ground should help increase safety quite a bit. I would avoid more active devices like transorbs as these may have enough capacitance to affect programming? My ICSP header has open pins too, and i've never fried a chip, but I don't touch it when it's programming. -Roman -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam@spam@mitvma.mit.edu

Roman, my chip was fried when it was touched by me to be removed from the eraser. I touched GP3 first apparently while the chip was in a jig (stacked socket with icsp SIL connector on short wires), and the rest of the pins were connected to the eraser table. I assume that I have exceeded the ESD protection level by 100 to 1000 times like this (at least capacitance wise). Adding resistors to pins is not an option. This particular circuit did one-pin A/D among other things. My ICSP driver successfully drives Vpp with a 0.047 uF cap to GND or Vcc every time. It's a question of having enough oomph and current limiting at the same time. 5-6usec is easily achieved even across 0.047uF. Check out the Microchip ICSP guide for a circuit that does that. I use something else. Peter -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

>This is the first PIC lost to ESD in ~5 years. Any suggestions ? Use a static strap? I realise this message is coming as a reply a long time after the event, but I have been out delivering some hardware on the that island in the middle of the pond :) While there I was involved in setting up the equipment in a clean room where everything is tied down for static purposes. The clean room suits have conductive threads through them, the gloves are conductive, the equipment was mounted on a metal bench, and I could still trigger the oscilloscope by plugging my static strap into a grounding post on the oscilloscope. Even when everything seems to be tied down to static, it is still possible to cause a glitch. -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads