Thank you Dom. The 100 nF decoupling cap on this board was almost half-an-inch from the pins. Soldered another one across the pins on the back of the board, and it now starts every time. (It's a one-off so the extra cap is a perfectly fine solution.) Thanks again! And thank you, Joe. I didn't try your suggestion, but that's just the kind of thing I was wondering about. Seems like good practice, and I may just implement it anyway.=20 Foster =20 > -----Original Message----- > From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] On Behalf > Of Picbits Sales > Sent: Wednesday, February 23, 2011 2:06 PM > To: Microcontroller discussion list - Public. > Subject: Re: [PIC] Internal Oscillator doesn't start reliably >=20 > I might be totally off base here but do you have adequate decoupling > capacitors on the power pins as close to the PIC as possible ? >=20 > Most issues I've had with startup instability have been due to inadequate > or > (once) faulty decoupling capacitors. >=20 > Dom > ----- Original Message ----- > From: "adastra" > To: "'Microcontroller discussion list - Public.'" > Sent: Wednesday, February 23, 2011 7:19 PM > Subject: [PIC] Internal Oscillator doesn't start reliably >=20 >=20 > >I have just created a simple application using the 8Mhz internal > oscillator > > of a PIC18F2520. The problem is that about half the time the oscillato= r > > doesn't start on power up. I am required to use an existing analog 12V > > power supply ahead of a 78L05 regulator. I have discovered that the > > oscillator will start every time if I connect the 12V to the board afte= r > > the > > power supply has already stabilized, so I inferred that it was the slow > > rise > > rate of the analog supply that was causing the problem. However, it > looks > > like I am getting 0 to 5V at the PIC Vdd pin in about 20 ms, which seem= s > > to > > be well within the minimum rise rate spec of 0.05 V/ms. I have tried > > various combinations of Configuration bits, i.e., Power-Up-Timer enable= d > / > > disabled, Brown-Out-Detect enabled / disabled, etc. but that doesn't > seem > > to > > make any difference. The hardware is extremely simple, so I must be > > missing > > something in the firmware. Any suggestions for where I might look woul= d > > be > > much appreciated. Thanks, Foster > > > > -- > > http://www.piclist.com PIC/SX FAQ & list archive > > View/change your membership options at > > http://mailman.mit.edu/mailman/listinfo/piclist > > >=20 > -- > http://www.piclist.com PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .