Loading ...
Sorry, an error occurred while loading the content.

RE: [cosmacelf] Ceramic Resonator for 18-2 oscillator

Expand Messages
  • bill rowe
    thanks lee. So I would need like 1M and 220 ohms? that would be an RC of 130 ns which seems odd. To: cosmacelf@yahoogroups.com From: leeahart@earthlink.net
    Message 1 of 4 , Aug 5, 2013
    • 0 Attachment
      thanks lee. So I would need like 1M and 220 ohms?  that would be an RC of 130 ns which seems odd.



      To: cosmacelf@yahoogroups.com
      From: leeahart@...
      Date: Mon, 5 Aug 2013 13:03:13 -0500
      Subject: Re: [cosmacelf] Ceramic Resonator for 18-2 oscillator

       
      bill rowe wrote:
      > with avrs i'm used to using 3 pin ceramic resonators like
      > this http://www.digikey.ca/product-detail/en/ZTT-2.00MG/XC1806-ND/440131
      >
      > Can I use one with the 1802, just wiring it between clock and crystal
      > with the 3rd pin to ground?

      The resonator will work, but you also need two resistors. I am now using
      this very same part (only the 1.8 MHz version) on my rev.F Membership
      Cards. The resonator gives you a stable 1.8 MHz oscillator, very close
      to the "standard" 1.79 MHz used with the 1802 and 1861. Here is the
      circuit needed (view with a fixed width font like Courier):

      ..... inverter (inside 1802, or external)
      ... _________|\o______________oscillator output
      .. | ....... |/ ..........|
      .. |____/\/\_______/\/\___|
      .. | ... R1 .... | . R2
      .. | .... _ .... |
      .. |____|| ||____|
      . _|_ . ||_|| . _|_
      . ___ ... X1 .. ___
      .. | C1 ..... C2 |
      . gnd ......... gnd

      X1 is the resonator, and C1 and C2 are its built-in capacitors (about 30pf).

      R1 is a high-value (like 1 to 10 megohms). It is needed to bias the
      inverter's input into its linear region.

      R2 is chosen to act as a low-pass filter with C1 and C2. Resonators have
      a much lower "Q" than crystals. Without R2, it can oscillate at an
      unstable frequency *above* the resonator frequency, set by the RC time
      constant of R2, C1, C2, and the stray capacitance across the resonator
      and R1. If R2 is too large, then it can oscillate at a *lower* frequency
      than the resonator, again due to the RC time constants.

      On the Membership Card, I used a single 1 meg trimpot to provide both R1
      and R2. The inverter is a 4093 schmitt-trigger gate.

      - With the trimpot at one end (R1=1meg, R2=0) it runs as an RC
      oscillator at 2-2.5 MHz.
      - With the trimpot at the other end (R1=0, R2=1meg), it runs as
      an RC oscillator at a very low frequency like 10-20 KHz.
      - With the trimpot set for about R2=100-300 ohms, it "locks in"
      on the resonator's frequency and runs at 1.79 MHz.

      --
      Failure is only the opportunity to begin again more intelligently.
      -- Henry Ford
      --
      Lee A. Hart, http://www.sunrise-ev.com/MembershipCard.htm

    • Lee Hart
      ... The value of R2 includes both the external resistor and the internal output resistance of the inverter. The capacitance also includes the stray capacitance
      Message 2 of 4 , Aug 5, 2013
      • 0 Attachment
        bill rowe wrote:
        > thanks lee. So I would need like 1M and 220 ohms? that would be an RC
        > of 130 ns which seems odd.

        The value of R2 includes both the external resistor and the internal
        output resistance of the inverter. The capacitance also includes the
        stray capacitance to ground, besides that inside the resonator itself.
        So it is rather hard to predict.

        I was using a 4093 as the oscillator inverter. With a selection of
        different brands of 4093's, I got

        R2=0 2.0 to 3.4 MHz at 5v
        1.4 to 2.3 MHz at 4v
        0.5 to 1.2 MHz at 3v

        Notice that 1.8 MHz (my resonator's frequency) is *inside* these limits.
        If the oscillator just happened to be close to 1.8 MHz, it "locks in" on
        this frequency, and wants to stay there. For example, if I powered it up
        at 5v, it might start oscillating at 2.5 MHz. As I lowered the supply
        voltage, the frequency drops until it gets to 2 MHz, when it suddenly
        "jumps" to 1.8 MHz. Once at this frequency, it stays there even if I
        raise the supply voltage back to 5v.

        Next, I studied the range of values for R2 that would make it start at
        1.8 MHz and stay there.

        5v supply R2 = 500 to 2.2k
        4v supply R2 = 300 to 1.5k
        3.6v supply R2 = 0 to 1k

        The actual value for R2 isn't quite that wide; it varies perhaps 2:1 for
        each chip. For example at 5v, one chip might need R2 = 500-1k, another
        needs 1k-2k, etc.

        The internal inverter inside the 1802 probably has a higher output
        resistance than the 4093, because it is optimized to be used as an
        oscillator. So, you'll have to experiment. But I'm quite sure you can
        find a value for R2 that will work with all 1802's at 5v.
        --
        Failure is only the opportunity to begin again more intelligently.
        -- Henry Ford
        --
        Lee A. Hart, http://www.sunrise-ev.com/LeesEVs.htm
      Your message has been successfully submitted and would be delivered to recipients shortly.