I had this problem with 12F508. The application would run from a 1000uF=20 cap when the power fails and the voltage would gradually drop to less=20 than 2V but would not go to 0V but stay around 1V to 1.5V. At this time if I power up again, the PICmicro would lock up and would=20 never run. I used a MCP1xx (voltage or system supervisor) to keep the PICmicro in=20 reset if the voltage fell below 2.5V. It worked. Cheers Ravi Ariel Rocholl wrote: > Dwayne, > > Can you reproduce the same behavior on a barebone circuit with the PIC + > variable supply and nothing else? I would certainly start by confirming > that, it should also be easier to experiment with burnout flag options to > test with. > > Ariel Rocholl > http://www.rf-explorer.com > > > On Mon, Apr 28, 2014 at 5:53 PM, Dwayne Reid wro= te: > >> At 06:41 AM 4/28/2014, Mike wrote: >>> Dwayne Reid planet.eon.net> writes: >>>> >>>> Varying the supply voltage from 1.0 through 2.5V after the PIC has >>>> locked up has no effect. Dropping the supply voltage down to zero >>>> volts, then restoring to above 2V starts everything back up. >>>> >>>> I'm kind of in a pinch here - this project delivers at the end of >>>> this month. Any suggestions gratefully accepted. >>>> >>>> Many thanks! >>>> >>>> dwayne >>> >>> >>> This has a hallmarks of a latch up state caused by EMI problems rather >> than >>> a simple brown out reset caused by a dip on Vdd. You'll need to check >>> what's happening on all the PICs pins rather than just Vdd. >>> >>> The 2 Amp/5.7v switcher is going to cause a pretty high current draw fr= om >>> the NiCds, does the PIC have it's own ground back to the battery pack o= r >> are >>> you sharing it with the high current paths? >>> >>> Regards >>> >>> Mike >> >> I'm pretty sure that the grounding is good - the entire solder-side >> of the PCB is ground plane. There are 2- 100uF ceramic bulk-storage >> caps sort of centered between the PIC and the 5.7V switcher. Battery >> ground is below all of that area, with that entire area being copper >> ground plane. I can't measure the ESR on those capacitors - it's too >> low for me to measure accurately. >> >> Each switcher also has very good local grounding where all of the >> high currents occupy a very small loop area. >> >> But I am re-creating the problem with both switchers idling - IE: >> there is no significant current flow. That's how I actually >> diagnosed the problem - varying my bench supply above and below >> 2.0Vdc with the switchers idling. >> >> I'm also pretty sure that the PIC is locking up. When the PIC is >> told to shut the system down because of inactivity, I switch the >> internal oscillator to its lowest speed (32 KHz, I think). I first >> tried just putting the PIC to sleep but the quiescent current >> consumption was so low that the 100uF cap across the battery input >> was taking several tens or hundreds of minutes to decay down to the >> point where I could get a reliable power-up. Changing from Sleep >> mode to the lowest internal oscillator speed causes the capacitor to >> decay in about 10 seconds. >> >> However, when this lockup occurs, its taking that tens or hundreds of >> minutes for the supply voltage to decay. That strongly suggests that >> the internal oscillator is not running. >> >> I'm still open to suggestions. I'm reluctant to engage Microchip >> Technical Support right now because they are *SO* un-responsive. >> >> dwayne >> >> -- >> Dwayne Reid >> Trinity Electronics Systems Ltd Edmonton, AB, CANADA >> (780) 489-3199 voice (780) 487-6397 fax >> www.trinity-electronics.com >> Custom Electronics Design and Manufacturing >> >> -- >> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive >> View/change your membership options at >> http://mailman.mit.edu/mailman/listinfo/piclist >> --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .