Shop OBEX P1 Docs P2 Docs Learn Events
5mhz Crystal Substitute in SMC? — Parallax Forums

5mhz Crystal Substitute in SMC?

JBWolfJBWolf Posts: 405
edited 2013-12-18 14:46 in General Discussion
Just wanted to make sure I'm not missing something before ordering....
I need a smaller SMC solution.
Is this a suitable substitute for the 5mhz crystal?

http://www.digikey.com/product-detail/en/NX8045GB-5MHZ-STD-CSF-3/644-1202-1-ND/3894081

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2013-11-02 15:49
    The load capacitance spec of that crystal is low at 8 pF. I believe the Prop provides more like 18-20 pF, which means the crystal will probalby oscillate, but slightly off the specified frequency.

    -Phil
  • Duane DegnDuane Degn Posts: 10,588
    edited 2013-11-02 15:56
    It seems like all the really small crystals I've seen in Propeller boards have been 10MHz. The Propeller Backpack uses a 10MHz crystal. Perhaps you can get the part number for Backpack's crystal.
  • hover1hover1 Posts: 1,929
    edited 2013-11-02 15:59
    The load capacitance spec of that crystal is low at 8 pF. I believe the Prop provides more like 18-20 pF, which means the crystal will probalby oscillate, but slightly off the specified frequency.

    -Phil

    Phil, I thought it was 6 pf per page 25 of the data sheet. Would 8 pf be close enough?
  • jmgjmg Posts: 15,173
    edited 2013-11-02 16:43
    JBWolf wrote: »
    Just wanted to make sure I'm not missing something before ordering....
    I need a smaller SMC solution.
    Is this a suitable substitute for the 5mhz crystal?

    http://www.digikey.com/product-detail/en/NX8045GB-5MHZ-STD-CSF-3/644-1202-1-ND/3894081

    Another option is a ceramic resonator - slightly lower precision than a Crystal.
    http://www.digikey.com/product-detail/en/CSTCR5M00G53Z-R0/490-4720-1-ND/1219380

    and if Size matters more than price, the Si504 Prog Osc comes as small as 2x2.5mm SMD, and you can define any Freq you like.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2013-11-02 16:54
    hover1 wrote:
    Phil, I thought it was 6 pf per page 25 of the data sheet. Would 8 pf be close enough?
    You're looking at the wrong spec. The 6 pF is for I/O pins. The correct spec is on page 14, where it shows 36 pF for both pins (18 pF apiece) for a 4-16 MHz crystal. So, no, 8 pF is not close enough.

    As to ceramic resonators: okay if you're not doing video. The BASIC Stamps all use ceramic resonators, BTW.

    Personally, if size is a problem (and I know jmg will disagree), I would use (and have used without issue) a 10 MHz crystal at PLL8x. It is outside the 4-8 MHz stated spec for PLL locking, though.

    -Phil
  • SRLMSRLM Posts: 5,045
    edited 2013-11-02 17:26
    You're looking at the wrong spec. The 6 pF is for I/O pins. The correct spec is on page 14, where it shows 36 pF for both pins (18 pF apiece) for a 4-16 MHz crystal. So, no, 8 pF is not close enough.

    Just to clarify a bit more, you should be looking for a crystal with 18-20pF of capacitance.
    This apparent discrepancy was explained somewhere, but I wouldn't know where to begin looking for it. I think it was either Mike Green or Tracy Allen who explained it. IIRC, the 36pF figure represents a value that is twice that of the crystal spec. Whether that's because it's the sum of two pins, while the crystal spec is per pin, or something else, I don't remember. But, whatever the reason, 18-20pF is the correct crystal spec to look for in a 5MHz crystal.

    -Phil
  • jmgjmg Posts: 15,173
    edited 2013-11-02 21:32
    This apparent discrepancy was explained somewhere, but I wouldn't know where to begin looking for it. I think it was either Mike Green or Tracy Allen who explained it. IIRC, the 36pF figure represents a value that is twice that of the crystal spec. Whether that's because it's the sum of two pins, while the crystal spec is per pin, or something else, I don't remember. But, whatever the reason, 18-20pF is the correct crystal spec to look for in a 5MHz crystal.

    The crystal 'sees' the two Pin C values in series., so that makes 36pF/Pin -> 18pF ideal Crystal operating point.
    A crystal with a different operating point will still oscillate, but not precisely at the factory trim mean value.

    The other XTAL options will likely also oscillate, and you could try a series R on XOUT to improve the phase margin if using the higher F settings at 5MHz.

    How many ppm do you actually need ?

    eg I've found some notes here, on a CMOS inverter oscillator
    (Note the pins are not quite the same in ppm/pF)

    HC49U/4 Xtal, 11.0592MHz Pull-ability
    One Pin is ~ -2.3328ppm/pF pull and the other is ~ -7.052ppm/pF, on HC49U Xtal.

    A 24MHz Murata crystal gave -0.332ppm/pF, & -3.132ppm/pF.


  • JBWolfJBWolf Posts: 405
    edited 2013-12-17 14:38
    - Update -
    Looks like more products are now available.
    Only difference I see is the ESR (equivalent series resistance).... is there a critical value needed for use with the prop?

    Digikey list of SMC & Through Hole 5mhz 20pf Crystals
  • Tracy AllenTracy Allen Posts: 6,664
    edited 2013-12-17 16:11
    In the range of ESR values they quote on the Digikey selection table, it won't make a difference. How small an SMT crystal do you need? I have one 5MHz AT cut 20pF from Jauch, 7mm x 5mm. JXG75P2 series.

    There is a diagram of a Pierce oscillator showing the series arrangement of the capacitors here, How-does-the-load-capacitance-of-a-crystal-effect-its-clock?
  • JBWolfJBWolf Posts: 405
    edited 2013-12-17 16:48
    7 x 5 is very nice... my circuit board is 12mm wide, so size is everything. I am using a CNC to cut .385mm wide channels between traces. so almost no component is too small for my use (except some of those microscopic resistors).
    I couldnt find anything at digikey relating to product id JXG75P2... can you find it there by chance?
  • jazzedjazzed Posts: 11,803
    edited 2013-12-17 19:16
    I gave up finding a small enough 5MHz crystal, and decided to use 10MHz instead (5mm x 3.2mm).

    http://www.mouser.com/ProductDetail/ABRACON/ABM3C-10000MHZ-D4Y-T/?qs=sGAEpiMZZMsPDM5321osT8c7QnCB39Ur
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2013-12-17 20:07
    I have used the small cylindrical crystals for quite some time but for the last couple of years I have been using the small 5x3mm SMD packs. Now the reason I use these smaller crystals is not necessarily because they are smaller but because I disagree with SMD lead forming a HC49/4 pack which ends up with ridiculously large pads spaced so far apart. On a typical double sided pcb we don't have all those extra layers for full ground planes and shielding so I choose to use a package that does not obstruct routing tracks from the I/O next to the crystal pins. The result is that I can place the crystal very close to the Prop with high noise immunity and still have my I/O go past the crystal. In contrast the HC49s span the whole width of a QFP and if I had to use the package I'd rather it were through-hole as at least the pads are minimal and closer together.

    Of course 10MHz is "outside" the range but that isn't normally a problem unless you are running at the extremes of the Prop operating conditions anyway. Besides Hydra used 10MHz and I have dozens of designs in the field using 10MHz.
    5x3xtal.png
    706 x 548 - 24K
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2013-12-17 20:17
    The Propeller Backpack uses a 10 MHz crystal (due to size contraints), with no reported issues.

    -Phil
  • tonyp12tonyp12 Posts: 1,951
    edited 2013-12-17 20:23
    For a beginner demo board I think a 5mhz crystal is a must, nearly everything in obex have that as default value.
    And having crashes due to forgetting to change the clkfreq when you trying a few demo codes is not fun.

    I'm checking mouser/digikey once a month to see if that 7x4mm 5mhz crystal will show up one day, if it does I will post it.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2013-12-17 20:31
    tonyp12 wrote:
    For a beginner demo board I think a 5mhz crystal is a must, nearly everything in obex have that as default value. ...
    That's why I wrote clkset.spin:
    {
    
    ┌──────────────────────────────────────────────────────────┐
    │                          CLKSET                          │
    │(c) Copyright 2009 Philip C. Pilgrim (propeller@phipi.com)│
    │            See end of file for terms of use.             │
    └──────────────────────────────────────────────────────────┘
    This object determines whether a 5MHz or 10MHz crystal is
    being used and sets the PLL accordingly for 80MHz operation.
    The main program should use the following settings:
    
    CON
    
      _clkmode      = xtal1 + pll8x
      _xinfreq      = 10_000_000
    
    OBJ
    
      clk   : "clkset"
      
    Then clk.set should be called first in the toplevel program.
    If the crystal is determined to be 5MHz, the clkmode will be
    changed to xtal1 + pll16x. clkfreq will still return 80_000_000.
    
    Version History
    ───────────────
    
    2009.05.23: WARNING: Trial, tentative, very alpha release!
    
    }
    
    PUB  set
    
    '' Automatically set the pll mode for a 5MHz or 10MHz crystal.
    
      cognew(@setpll, 0)
    
    DAT
    
                  org       0
    setpll        movi      ctra,#%0_00001_011      'Set ctra for pll on no pin at x1.
                  movi      frqa,#%0100_0000_0      'Set frqa for clk / 4 (= 20MHz w/ 10MHz crystal, i.e. too high).
                  add       x,cnt                   'Give PLL time to lock (if it can) and stabilize.
                  waitcnt   x,#0
                  movi      vcfg,#$40               'Configure video for discrete pins, but no pins enabled.
                  mov       vscl,#$10               'Set for 16 clocks per frame.
                  waitvid   0,0                     'Wait once to clear time.
                  neg       x,cnt                   'Read -cnt.
                  waitvid   0,0                     'Wait 16 pll clocks = 64 processor clks.
                  add       x,cnt                   'Compute elapsed time.
                  cmp       x,#$40 wz               'Is it really 64 clocks?               
            if_z  mov       x,#$6f                  '  Yes: Crystal is 5MHz. Set clock mode to XTAL1 and PLL16X.
            if_z  clkset    x
                  cogid     x                       'Stop the cog: we're done.
                  cogstop   x
                  
    x             long      $1_0000                 '65536 clks for pll to stabilize.
    
    {{
    ┌──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┐
    │                                                   TERMS OF USE: MIT License                                                  │                                                            
    ├──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┤
    │Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation    │ 
    │files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy,    │
    │modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software│
    │is furnished to do so, subject to the following conditions:                                                                   │
    │                                                                                                                              │
    │The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.│
    │                                                                                                                              │
    │THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE          │
    │WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR         │
    │COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,   │
    │ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.                         │
    └──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────┘
    }}             
    
    It allows you to use either a 5 MHz or 10 MHz crystal and handles the details of figuring which crystal is being used and doing the appropriate clock settings. No other programmer or user intervention is required.

    -Phil
  • jazzedjazzed Posts: 11,803
    edited 2013-12-17 20:46
    Nice idea Phil.
    That's why I wrote clkset.spin:
  • Mark_TMark_T Posts: 1,981
    edited 2013-12-18 04:34
    jmg wrote: »
    Another option is a ceramic resonator - slightly lower precision than a Crystal.

    Hundreds of times lower accuracy than a quartz crystal - I'd never choose a ceramic resonator myself, there are
    many things that want accurate frequencies and low jitter.

    Ceramic resonator - +/-5000 ppm typical spec (0.5%, 7 minutes a day, sensitive to temperature)
    Quartz crystal - +/-30ppm typical spec. (0.003%, 3 secs/day)
    TCXO - +/- 2ppm typical spec. (0.0002%)
  • JBWolfJBWolf Posts: 405
    edited 2013-12-18 06:40
    I'll have to take a look at 10mhz crystals... didnt know they were an option, thanks phil! :)

    Peter can you give a link or product number for the 5x3mm SMD packs? I would like to look at those, that size is just perfect.
  • Tracy AllenTracy Allen Posts: 6,664
    edited 2013-12-18 09:07
    Oh, sorry, my 5x7mm 5MHz/20pF crystals were special order full reel from Jauch. A couple of other companies I talked to could not do that combination, but they could start at 6 or 6.25MHz in the 5x7 package. 6.25MHz is nice (as Bill Henning showed) because the Prop can then operate at 100MHz.

    The official range for the crystal is 4 to 8 MHz, which gives PLL operation from 64 to 128MHz. 10MHz is a bit outside that range, PLL at 160MHz and prop operation at 80MHz on PLL8. It seems to be working okay for many designs now, but I wonder about extended temperature ranges. Has anyone been using them at extreme temperatures?
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2013-12-18 14:46
    JBWolf wrote: »
    I'll have to take a look at 10mhz crystals... didnt know they were an option, thanks phil! :)

    Peter can you give a link or product number for the 5x3mm SMD packs? I would like to look at those, that size is just perfect.

    I've used a couple of different ones, here are some links to Digikey.
    10MHz 12pF 30ppm 48c @100
    10MHz 8pF 50ppm 48c @100
Sign In or Register to comment.