Post by thread:[PIC] 16f628 Help

Ok, I'm stumped. I've been struggling with a little 16f628 based project I built the other day. I can't make it work. It seems as though the clock frequency is running at about 500Hz with a 16MHz crystal installed. I didn't trust scoping the OSC lines, so I wrote a little program to toggle a bit over and over...it seems to be running at about 80Hz. It doesn't make sense at all. My bias caps are grounded. Software worked great on the breadboard. I can't understand why I'm running so slow. My config fuses are set to HS oscillator, MCLR on pin, no code protection, watchdog, BOD, or LVP. I read back a chip on another computer and the fuses were set correctly. I've tried two chips, numerous crystals (both 16 and 20MHz) and two sets of bias caps. I've checked my PCB numerous times over the last two days. Power is nice and clean, bypass caps installed. I've pretty much exhausted every avenue I can think of. I've never had a PIC run, but at a tiny fraction of the speed it should be at. At that clock speed the chip is just sitting forever in an opening delay loop. AAAHHHHH! All help muchly appreciated. Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams

Just a little update. I seem to have gotten it working, though I still can't explain the previous behaviour I was seeing. I found an intermittent short between two tracks. HOWEVER, these were two tracks on a data bus that was unconnected earlier when I was trying to figure out what was going wrong. So...I'm not really sure how they could have effected my clock frequency so much. I'm still at a loss to understand why my chip was still running, but running very slowly. It's working now (I think...not all tests have been completed)...that crazy curiousity bug still has me though. Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams

> though the clock frequency is running at about 500Hz with a > 16MHz crystal installed. I didn't trust scoping the OSC lines, > so I wrote a I think you should scope the OSC lines, just to cover all the bases. A x10 setting on the probe will work, or picking the frequency up by radiation. You might also be able make out small oscillator ripple on an active port pin eg http://www.sxlist.com/images/sxlist/an_sc_04.gif

I worded that poorly. I meant that I measured about 500Hz on the OSC lines, but I was worried I was pulling it too much with my probe, so I wrote the bit toggle problem. I was using the x10 setting, and I verified that it worked by scoping another functioning circuit. Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams On 4/12/05, Jinx <spam_OUTjoecolquittTakeThisOuTclear.net.nz> wrote: {Quote hidden}> > though the clock frequency is running at about 500Hz with a > > 16MHz crystal installed. I didn't trust scoping the OSC lines, > > so I wrote a > > I think you should scope the OSC lines, just to cover all the > bases. A x10 setting on the probe will work, or picking the > frequency up by radiation. You might also be able make out > small oscillator ripple on an active port pin eg > > http://www.sxlist.com/images/sxlist/an_sc_04.gif

> I worded that poorly. I meant that I measured about 500Hz on the > OSC lines, but I was worried I was pulling it too much with my probe, > so I wrote the bit toggle problem. I was using the x10 setting, and I > verified that it worked by scoping another functioning circuit. Ah. I'm very surprised that you'd measure a crystal being so far, inter- planetarily far, away from its fundamental. It looks like the OSC circuitry wasn't started or running properly

On 4/12/05, Jinx <.....joecolquittKILLspam@spam@clear.net.nz> wrote: > > I worded that poorly. I meant that I measured about 500Hz on the > > OSC lines, but I was worried I was pulling it too much with my probe, > > so I wrote the bit toggle problem. I was using the x10 setting, and I > > verified that it worked by scoping another functioning circuit. > > Ah. I'm very surprised that you'd measure a crystal being so far, inter- > planetarily far, away from its fundamental. It looks like the OSC circuitry > wasn't started or running properly Yes, well, I was pretty surprised too. I actually suspected the scope at first. I didn't have my good scope handy. I have no idea how this happened...as I said before, an intermittent short on two data bus connections seems to have done this. Weird eh? Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams

This is interresting, I have seen this before on a radio design I did using a 6502 processor. I was new ( still am) and still learning to "code". The 6502 was used to set PLL registers in a radio design. Long story short, I/O shorted and somehow caused the part to starve out and when it came up it latched the I/O to source and sink. The only difference though is that processor was toasted. I did observe 60hz and 400Hz signal across the Osc which was probably my digital scope wigging out on noise. Best Regards, Nick __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

<Josh Koffman notes> > Just a little update. I seem to have gotten it working, though I still > can't explain the previous behaviour I was seeing. I found an > intermittent short between two tracks. HOWEVER, these were two tracks > on a data bus that was unconnected earlier when I was trying to figure > out what was going wrong. So...I'm not really sure how they could have > effected my clock frequency so much. I'm still at a loss to understand > why my chip was still running, but running very slowly. It's working > now (I think...not all tests have been completed)...that crazy > curiousity bug still has me though. Any chance Josh, that the watchdog timer was going off and cycling the whole chip? Id I recall correctly you were looking at an ouput port for a heartbeat, what you may have been seing was each restart triggered by the watchdog. Aza D. Oberman

I had that thought too...so I checked a few times. No WDT. Good thought though! Josh -- A common mistake that people make when trying to design something completely foolproof is to underestimate the ingenuity of complete fools. -Douglas Adams On 4/13/05, Aza D. Oberman <WinKILLspamslycurves.com> wrote: > Any chance Josh, that the watchdog timer was going off and cycling the whole > chip? Id I recall correctly you were looking at an ouput port for a > heartbeat, what you may have been seing was each restart triggered by the > watchdog.