Some comments from a friend follow my comments. I agree that crystal oscillators can be FAR more tricky than may have been suggested by some of the comments on this thread. While often enough the "paralllel resonant crystal with 2 x 20 pF caps" is all that's needed, this certainly isn't always the case. Some processor & crystal combinations can be extremely finnicky and may need fine tuning of capacitors and then some. Worst case I have ever met involved a batch of Zilog Z8+s that required an inductor across the oscillator pins to ensure reliable startup in all cases. A most unusual requirement, fortunately. RM ______________________________ Crystal (and resonator) startup is a very complex phenomena and something of a black art. Philips (and other uC manufacturers) have written quite detailed treatises on crystal behaviour and it makes sobering reading for anyone manufacturing in high volumes. Epson also have done a detailed manual on crystals and resonators and their use in reliable oscillators. Murata also have a good manual on resonators and their application. Incidenatlly a major difference between crystals and resonators is their startup time - resonators are typically 10-100 times faster and in low-power uC applications their faster startup can be a real advantage both with regard to power consumption and reliability. We have seen an application where a crystal was substituted for a resonator in such a system and while it worked in some cases it didn't in others. Turned out that the system periodiacally woke up the uC which in most cases then communicated with an I2C bus device exterally before going back to sleep. Problem was that the crystal was very slow to start and ran the uC very erratically while doing so such that the I2C bus transactions were screwed up. In fact in most cases the uC went back to sleep (believing it had done its job) before the crystal had really got going. With a resonator (as originally designed) the oscillator was up to full amplitude and stable 100's of us before the I2C transaction, while with the crystal oscillation had typically only reached about 20% of full amplitude after several ms when it went back to sleep. You can find the Philips app notes on their website or (more easily) from ours at: http://www.acqura.com/technotes.asp?ID=483 http://www.acqura.com/technotes.asp?ID=486 The second is a rework of the first and contains quite a bit of additional theory and coverage of additional issues. Regards, Ken Mardle Applied Digital Research Ltd. Tel : +64 9 415-2514 P.O. Box 6480 Fax : +64 9 415-3514 Wellesley St Mobile : +64 25 879-648 Auckland E-mail : kenmar@adr.co.nz NEW ZEALAND WWW : http://www.acqura.com > -----Original Message----- > From: Russell McMahon [mailto:apptech@paradise.net.nz] > Sent: Friday, 5 December 2003 9:30 a.m. > To: Ken Mardle; Ross McMillan; xoddy; Dr Gavin Higgie > Subject: Fw: [PICLIST] [PIC:] Argh, more crystal questions. > > > Interesting comment about crystal reliability versus mounting on silicon > rubber. > > _____________________________________________ > > ----- Original Message ----- > From: "Bob Axtell" > To: > Sent: Friday, December 05, 2003 3:32 AM > Subject: Re: [PICLIST] [PIC:] Argh, more crystal questions. > > > > I guess I should add my experiences with crystal oscillators. > > > > First- NO question about it- ceramic resonators are very > durable and more > > reliable > > than piezo crystals, particularly in a possible high-shock environment. > > They are also slightly less expensive. But the negatives are > that they are > > susceptible to moisture and their freq stability is much worse. BUT, I > > always use ceramic resonators whenever possible and I've never been > > disappointed yet. > > > > With shock, crystals will fracture then oscillate at a wrong > frequency. If > > you layout the PCB to allow shock-mounting a crystal, such as > floating the > > crystal case in a drop of silicone caulk, with the flying leads > formed to > > prevent shock energy from being transferred, reliability will improve > > 100-fold. It follows that SMD crystals, which have NO leads, will be the > > least reliable: and they are. > > > > The caps are used to force the phase shift of the oscillator, and have > > little effect except at startup. It's best to use the MC-recommended > > values. A lower cap value will usually cause the oscillator startup time > to > > change. Overall, a slight shift in cap values doesn't seem to make any > > difference, especially if both are the same. > > > > In high-precision situations, such as 14.31818Mhz video situations, I > > prefer to use a guaranteed external oscillator package instead of the uP > > oscillator. > > > > --Bob > > > > At 03:42 AM 12/4/2003, you wrote: > > > > -----Original Message----- > > > > From: John J. McDonough [mailto:wb8rcr@ARRL.NET] > > > > Sent: 04 December 2003 03:53 > > > > To: PICLIST@MITVMA.MIT.EDU > > > > Subject: Re: [PIC:] Argh, more crystal questions. > > > > > > > I never did get the ceramic resonator thing. Microprocessor > > > > crystals are > > > > more stable, and no more expensive. The C-301 case jobs are > > > > also a lot > > > > smaller. The HC49 cases are a little larger but cheaper. > > > > But in all cases, > > > > we're talking under a buck (that's about 80% of a Euro these > > > > days), so it > > > > seems worth it to go for a little extra stability unless > > > > you're going to buy > > > > a thousand of the things. > > > > > >They are much more rugged than crystals, so if your product is going > > >somewhere likely to experience shocks and vibration, a > resonator is much > > >more likely to survive. Also, as you say, in large volume > production the > > >cost savings could be significant. > > > > > > > > >Mike > > > > > > > > > > > > > > >======================================================================= > > >This e-mail is intended for the person it is addressed to only. The > > >information contained in it may be confidential and/or protected by > > >law. If you are not the intended recipient of this message, you must > > >not make any use of this information, or copy or show it to any > > >person. Please contact us immediately to tell us that you have > > >received this e-mail, and return the original to us. Any use, > > >forwarding, printing or copying of this message is strictly prohibited. > > >No part of this message can be considered a request for goods or > > >services. > > >======================================================================= > > >Any questions about Bookham's E-Mail service should be directed to > > >postmaster@bookham.com. > > > > > >-- > > >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 > > > > -------------- > > Bob Axtell > > PIC Hardware & Firmware Dev > > Tucson, AZ > > 1-520-219-2363 > > "I lose some on each sale but make it up in volume." > > > > -- > > 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.