Shop OBEX P1 Docs P2 Docs Learn Events
Dead crystal. How did I kill it? — Parallax Forums

Dead crystal. How did I kill it?

MicrocontrollerGuyMicrocontrollerGuy Posts: 12
edited 2010-08-20 19:12 in Propeller 1
Of all the things on the Propeller education kit board, the crystal seems like one of the least likely to fail. Yet after a couple of hours of checking wires, testing voltages and even swapping in another Propeller just in case I had fried the one on the board, my problem finally ended up being a dead crystal. I put a new crystal in and everything was back to normal. The dead crystal had worked fine for several hours before it died.

Was this just really bad luck with this part, or is there something I should be doing to prevent these crystal failures?

Comments

  • Mike GreenMike Green Posts: 23,101
    edited 2010-08-17 21:23
    It's pretty hard to damage a crystal electrically in a CMOS oscillator circuit like the one used in the Propeller. You can damage a crystal with a mechanical blow or you can heavily overdrive it (impossible with the Propeller). You may have had a defective one or it may have been dropped at some point.
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2010-08-18 01:14
    I have known for xtals to just die, without severe provocation. I guess that since it is a mechanical device the a certain amount of failures would be expected.


    "When I was young", the xtals were much bigger and quite rare and expensive. They would have been more robust and treated with great reverence. Now they are a few pence/cents and come out of bags of 1000's.


    I cannot help myself if there is a scrap board, with a rock on it, it has to be harvested "just in case".
  • MicrocontrollerGuyMicrocontrollerGuy Posts: 12
    edited 2010-08-19 00:28
    I just killed another one. This time it was a 16 MHz crystal that I was using for an AVR microcontroller. I think I know what I did, too. I had been trimming and bending the leads to get the crystal to sit closer to the board, and I think in the process of bending the leads I've damaged the crystals. From now on I'm going to let them stick up from the board rather than risk breaking them.
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2010-08-19 00:42
    The packages that xtals come in are quite robust, but if you are bending the wires (which are usually steel based) it would be better to put some fine nosed pliers against the base of the xtal and bend against that. This is true of most components though.

    When the xtal is already soldered onto a PCB, flush to its based then a minimum heating time should be used as the very short leads will heat up the innards very quickly. Surface mounted xtals can be a right ******* to get off, especially on PCBs that have tracks that are thicker than they are wide.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2010-08-19 05:56
    "quite rare", "quite robust"
    I'm... almost speechless.

    WTF? Why is MicrocontrollerGuy posting to the Propeller Forum about a non-Propeller matter? It's about a crystal and that's for an AVR project. WTF?
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2010-08-19 06:03
    It's the "new age" of being polite, to newbies.

    The start of this thread was Prop based.
  • MicrocontrollerGuyMicrocontrollerGuy Posts: 12
    edited 2010-08-19 12:18
    I was trying to offer information that might be helpful to others, specifically about the risks of damaging a crystal while bending the leads to make it fit better in a breadboard. Apparently I shouldn't have mentioned this, because the second crystal (i.e., the one that allowed me to determine what I was doing wrong) happened to be connected to a non-Propeller microcontroller.

    Rest assured that I will only post about topics and components that are directly connected to a Propeller microcontroller in the future. Have a nice day.
  • schillschill Posts: 741
    edited 2010-08-19 12:24
    I was trying to offer information that might be helpful to others, specifically about the risks of damaging a crystal while bending the leads to make it fit better in a breadboard. Apparently I shouldn't have mentioned this, because the second crystal (i.e., the one that allowed me to determine what I was doing wrong) happened to be connected to a non-Propeller microcontroller.

    Rest assured that I will only post about topics and components that are directly connected to a Propeller microcontroller in the future. Have a nice day.

    I thought the additional information provided by the "AVR" example was good and fit in with the thread. As long as you are talking about the crystal it had direct bearing on your original post.

    Since you were not posting about AVRs, I expect the complaint above was meant to be (at some level) humorous - although I think the language (even in acronyms) is too strong. There has been a lot of discussion recently about comparing different companies products (which you did not do) and I believe this was just a reaction to those discussions.
  • John R.John R. Posts: 1,376
    edited 2010-08-19 12:42
    Ya gotta realize, PJ is our unofficial "keep the forums groups pure" guy, and takes great exception to "off topic" posts. Once you get to know him better, he kind of grows on you...
  • Heater.Heater. Posts: 21,230
    edited 2010-08-19 12:53
    Next time you kill a crystal, just say the "The Prop did it". Then everything will be OK.
  • RavenkallenRavenkallen Posts: 1,057
    edited 2010-08-19 13:13
    @PJ allen

    1: Because pretty much nobody checks the sandbox
    2: He is a newbie, cut him some slack. Didn't you see Ken's post about wanting us to be more friendly to newcomers
    3: His initial question did involve the use of/ interfacing to a propeller

    We have got to be friendly to those who don't know the rules/ traditions of these forums. Their experience here could directly contribute to the success of Parallax. WE have got to make a good impression and develop a thicker skin
  • jmgjmg Posts: 15,185
    edited 2010-08-19 14:46
    PJ Allen wrote: »
    WTF? Why is MicrocontrollerGuy posting to the Propeller Forum about a non-Propeller matter? It's about a crystal and that's for an AVR project. WTF?

    Oops, someone failed to read everything.

    The OP followup was very good, and something to be encouraged. Threads should leave useful information behind, for others to learn from.

    The 'AVR' info is not heresy, but valid as it gives a second reference point, that a failure is NOT related to the Prop.

    It underlines that physical stress matters, and If I understood the time-lines, that the failure was not post-stress-instant either.

    The only suggestion I'd add for the OP, is to include the package of the crystals he was using - to give his experience full context.
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2010-08-20 06:27
    A mechanical stress on any component will likely result failure.
    Watch out for heat, too, "newbies", because it can be a killer (or, in keeping with demonstrated banality, a real killer.)
  • John R.John R. Posts: 1,376
    edited 2010-08-20 06:32
    See, PJ really is a nice guy...

    Everybody say "Awwww....." :-)
  • Greg GlennGreg Glenn Posts: 17
    edited 2010-08-20 07:55
    I killed a crystal on a propeller project. It happened when I had to resolder a QFN chip no less than 3 times with a rework hot-air iron. The crystal was right next to the chip. That big metal housing really soaks up the heat.

    I was puzzled at first, because I could load my code onto the prop OK, but it would not run. When I ran a small test program that did not have the _xinfreq and _clkmode definitions, it ran fine because it was using the internal clock on the propeller. That told me that the crystal had gone bad.
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2010-08-20 13:30
    You can get heat resistant pastes/gums to go over the nearby components and set to a rubber like form. After the heat work has finished it just peels off.
  • yarisboyyarisboy Posts: 245
    edited 2010-08-20 19:12
    I really like the pin socket arrangement Gadget Gangsta uses on the Propeller platform. When laying out a board one must bump the hole size up to the diameter of the pin socket after you cut it out of the plastic carrier it came in. If my app is going to see a lot of shock and vibration I'd still solder it in with the insulator the xtal comes with. I've potted some boards with epoxy just for good measure.
Sign In or Register to comment.