Russell said:

>Note that when using Sigma-Delta A2D converter that the time constants of
>the drive resistors and integrating capacitor can be important (even though
>some people will tell you that they are not). Plot your actual input voltage
>versus the SD converter result. If you get sudden steps, typically but not
>always at the 50% and/or 25% points or interesting non linear patterns try
>changing capacitor (or resistor) value(s) and also the sampling rate and see

    A2s.asm uses Port B to display the count. On my breadboard,
    the count runs up to 255 for only a few tenths of a volt
    input, then repeats as the input continues to increase.
    I tried capacitor values of 0.1, 0.33 and 0.47uF. Changing
    the cap doesn't make nearly as much difference as the input/
    output resistors.

>Note also that the SD process relies on a hardware switching point that is
>well defined and very consistent. This will not always be the case for all
>processors input pins. It seems from Scott's reports that the PIC performs
>well in this respect. My experience with the Z8 showed that it's
>auto-Schmitt-triggered inputs greatly limited achievable converter accuracy.
>Use of the Z8's hardware comparator as the switching level detection input
>produces good results.

   Good tip. One of the parts I was contemplating using this scheme with
   has Schmitt inputs. I need to set two alarm points, at approx. 0.75 volts
   and 3.75 volts. Accuracy is not as important as repeatability from unit
   to unit. This is an attempt to get some cost out of an existing product.

   Thanks for the input. I'm continuing to experiment with it.

   Reg

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu