Shop OBEX P1 Docs P2 Docs Learn Events
?? Adc question — Parallax Forums

?? Adc question

HarleyHarley Posts: 997
edited 2009-10-09 03:15 in Propeller 1
AN001, page 17, shows the ∑∆ Adc using two 1 nF (= 0.001 µF ?) caps.

Should these values change if run at a much slower speed (about 76 samples/sec; using 20 bits)?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Harley Shanko

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-10-09 02:18
    Wow! Greek letters in the subject line! How'd you do that?

    To answer your question: I think the caps should stay the same. The frequency of concern is not the sampling rate, but the 80MHz feedback. You'll be be hard-pressed to get a meaningful resolution much beyond 10 bits, though. I don't think 20 bits is even within the realm of fantasy, let alone reason. Sorry. sad.gif

    -Phil
  • HarleyHarley Posts: 997
    edited 2009-10-09 02:34 
    Wow! Greek letters in the subject line! How'd you do that?

To answer your question: I think the caps should stay the same. The frequency of concern is not the sampling rate, but the 80MHz feedback. You'll be be hard-pressed to get a meaningful resolution much beyond 10 bits, though. I don't think 20 bits is even within the realm of fantasy, let alone reason. Sorry.



    The Mac/iMac has had that capability for many years.

    Hmm. I'm trying to sample a slow signal (heart pulse) and was just wondering if those caps should be different. Maybe I need to extract every nth sample to get what I want. Thanks for the guidance. yeah.gif

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Harley Shanko
  • turbosupraturbosupra Posts: 1,088
    edited 2009-10-09 03:06
    http://forums.parallax.com/showthread.php?p=846700

    It has nothing to do with the OS

    start --> run --> charmap

    I used the font simhei

    (I'll save you a click since you were nice enough to build that code generator, even though you came into my other thread accusing me of trying to do something illegal)

    greek1.jpg



    Phil Pilgrim (PhiPi) said...
    Wow! Greek letters in the subject line! How'd you do that?

    To answer your question: I think the caps should stay the same. The frequency of concern is not the sampling rate, but the 80MHz feedback. You'll be be hard-pressed to get a meaningful resolution much beyond 10 bits, though. I don't think 20 bits is even within the realm of fantasy, let alone reason. Sorry. sad.gif

    -Phil
  • KyeKye Posts: 2,200
    edited 2009-10-09 03:09
    The problem is with the fact that the ADC needs a certaint amount of clock ticks for a resolution.

    As in, if you wanted 20 bits of resoltuion you would need to sample it after every 2^20th clock cycles.

    This same idea applies for any bit length. If you do not sample it after a power of 2 time then you will need to multiply up or divide the value down to get the correct result.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Nyamekye,
  • mparkmpark Posts: 1,318
    edited 2009-10-09 03:15
    I hope your tongue is in your cheek about Phil's so-called "accusation." I read it more as friendly advice to you.
Sign In or Register to comment.