Shop OBEX P1 Docs P2 Docs Learn Events
OEM BS2 resonators — Parallax Forums

OEM BS2 resonators

$WMc%$WMc% Posts: 1,884
edited 2008-12-18 02:20 in BASIC Stamp
Hello All

Is it possible to·parallel the ceramic resonators or Xtals on the OEMBS2s. I don't want to change the clock speed,I just need·two OEMBS2s "PIC16C57C" to run in sync. with each other.

I looked for some info on this,I couldn't come up with anything?


___Thanks for any info_______$WMc%__________Merry Xmas__

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Truth is out there

Comments

  • PJAllenPJAllen Banned Posts: 5,065
    edited 2008-12-09 01:32
    Resonators in parallel would conflict, not synched and not the exact same freq either.

    I think that the attached would have the two OEM PICs running from the same clock.

    Post Edit -- Here's the link to the PIC datasheet (pdf) from Microchip

    Post Edited (PJ Allen) : 12/9/2008 1:38:13 AM GMT
    spcl_clk.JPG
    707 x 422 - 44K
  • $WMc%$WMc% Posts: 1,884
    edited 2008-12-09 01:42
    Thank You Mr.Allen

    You are the Man!!!!

    Exactly what I was looking for


    ____________$WMc%______Merry Xmas

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    The Truth is out there

    Post Edited ($WMc%) : 12/18/2008 2:23:55 AM GMT
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2008-12-09 20:57
    Bear in mind you have to test the loading characteristics in a design like this. It’s not easy to make changes because you are not able to change the firmware which is setting the drive level. R3 as shown also affects this. What I have seen done successfully is feeding the output of a TTL oscillator into both OSC inputs. When using a TTL OSC feedback and drive are almost irrelevant.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Engineering
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2008-12-09 23:44
    I was all day at work wanting to recommend just that, CS.· Clearly, the external oscillator would be the superior method.·

    Also, I get a ding because the PIC's input buffer is an inverter (see datasheet Figure 4-1.)· So,·PIC #1's·OSCOUT fed to PIC #2's OSCIN would result in the two clocks being 180deg out of phase (another inversion in PIC #2.)


    Post Edit -- dwg attached.

    Post Edited (PJ Allen) : 12/9/2008 11:53:32 PM GMT
    clk_diff.JPG
    380 x 334 - 20K
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2008-12-09 23:58
    PJ,

    In many cases your suggestion would work fine...But on some systems very balanced loading is required and without it, things don't work right. This question came up not too long ago regarding the SX and that is a perfect example of where you have to following the proper load capacitance, feedback resistance and drive. Take care.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Engineering
  • $WMc%$WMc% Posts: 1,884
    edited 2008-12-12 03:10
    Mr.Allen and Mr.Savage

    I would like to think You both for the help and info.

    I know that the "wire / trace" length is important,But I would still like to experiment with this.
    I'll try the Ext. Clock 1st


    ______________$WMc%_____________Merry Xmas

    I know of no other forums that are this helpful and informative(Thanks Parallax)

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    The Truth is out there

    Post Edited ($WMc%) : 12/12/2008 8:51:13 AM GMT
  • Chris SavageChris Savage Parallax Engineering Posts: 14,406
    edited 2008-12-17 23:18
    $WMc% said...(trimmed)
    I know that the "wire / trace" length is important,But I would still like to experiment with this.
    There are guidelines for layout from many manufacturers regarding traces, ground connections, etc.· Trace length, like trace cornering/defelction are somewhat misunderstood I think in general by those who don't have to deal with those issues.· Typically these things are only an issue on extremely high-speed PCB layouts, such as those used in PC Motherboards and in radio equipment.· That said, some things are easily affected by simply having signals near them or not having problem ground plane coverage,·such as OSC inputs with crystals.· Let's take the DS1302 for example.· Not following the manufacturer guidelines carefully when laying out a board using this chip can lead to a slow or (more often) fast clock as the OSC pins are highly susceptible to signal noise from nearby traces.· These signals can ghost extra clock pulses into the mix which over time add up, making the clock seem to run fast.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Chris Savage
    Parallax Engineering
  • $WMc%$WMc% Posts: 1,884
    edited 2008-12-18 02:20
    Mr.Savage

    Thanks for the added info. I started out in RF back in the day. It still amazes Me as to what a ground loop can do. And, induced noise with such little current!!. But it's this kind of experimenting that led to things like terminating resistors,Grounding a shield on one end only, twisted pair ,etc. Sometimes You just feel like playing with things to see what happens?

    Thanks again________________$WMc%___________________Merry Xmas

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    The Truth is out there
Sign In or Register to comment.