Skip to content
Commits on Oct 13, 2015
  1. @tdicola
  2. @tdicola
  3. @tdicola

    Merge comment update from pull #43

    tdicola committed
  4. @tdicola
  5. @tdicola

    Bump library to 1.2.2 version.

    tdicola committed
  6. @tdicola

    Merge branch 'jmdana-master'

    tdicola committed
  7. @tdicola
Commits on Oct 6, 2015
  1. @jmdana

    Divisions turned into multiplications

    jmdana committed
    As correctly stated by Mausy5043, the division is quite slow in most
    microcontrollers, so I have replaced them with multiplications.
  2. @tdicola

    Merge branch 'jmdana-master'

    tdicola committed
  3. @tdicola

    Bump version to 1.2.1

    tdicola committed
Commits on Oct 3, 2015
  1. @jmdana
Commits on Oct 2, 2015
  1. @jmdana

    DHT::computeHeatIndex has been modified so it uses both Rothfusz and

    jmdana committed
    Steadman's equations. As a result, it allows a larger range of
    temperatures and more accurate results for extreme conditions.
Commits on Sep 15, 2015
  1. @tdicola
  2. @tdicola
Commits on Jul 24, 2015
  1. @matthijskooijman

    Use INPUT_PULLUP

    matthijskooijman committed
    Previously, the mode was configured to INPUT and then written HIGH,
    which works on AVR but isn't very portable. Instead, this uses the
    INPUT_PULLUP pin mode (available since Arduino 1.0.1) which is portable.
  2. @matthijskooijman

    Shrink DHT::data to 5 bytes

    matthijskooijman committed
    There's only 5 bytes of data being transferred (2x2 bytes of data, 1
    byte checksum), so no need to keep a useless byte.
  3. @matthijskooijman

    Remove the firstreading variable

    matthijskooijman committed
    By cleverly setting _lastreadtime in begin() to make sure that it looks
    like the last read was alrady 2000 ms ago, even on startup when millis()
    still returns 0, there is no longer a need to keep a separate
    firstreading variable, saving a byte of memory and a bit of code.
  4. @matthijskooijman

    Clean up two-second interval check

    matthijskooijman committed
    There was some handling of millis() overflow that is not needed.
    Standard unsigned subtraction and wraparound already works as expected,
    so this extra check can be removed (it even hurts, since it introduces 2
    seconds after a wraparound where no new data will be read, even if it
    could otherwise happen).
    
    Also, this prevents calling millis() a second time, since its value is
    already known.
  5. @matthijskooijman

    Add force parameter to read methods

    matthijskooijman committed
    This allows forcing a read, even if the previous read was less than 2
    seconds ago. This is useful in cases where the millis() timer is not
    reliable, such as when sleeping. In this case, it is up to the caller to
    ensure that at least 2 seconds elapse between calls with force set to
    true.
Commits on Jul 22, 2015
  1. @tdicola
  2. @tdicola

    Fix #37 by optimizing pulse read to use direct port access on AVR, st…

    tdicola committed
    …ops checksum failures from missed signals.
Commits on Jul 2, 2015
  1. @tdicola

    Merge pull request #36 from Zirientis/master

    tdicola committed
    Cause interrupts to be reenabled if a timeout occurs while waiting for the sensor
  2. Cause interrupts to be reenabled if a timeout occurs while waiting fo…

    Zirientis committed
    …r the sensor.
Commits on Jun 26, 2015
  1. @tdicola

    Merge branch 'MyIgel-master'

    tdicola committed
  2. @tdicola

    Fix merge conflicts, make compute heat index default to Fahrenheit to…

    tdicola committed
    … not break old code that assumes default is Fahrenheit.
  3. @tdicola

    Merge branch 'dynamic_timing'

    tdicola committed
  4. @tdicola
  5. @tdicola
Commits on Mar 31, 2015
  1. @tdicola
Commits on Nov 1, 2014
  1. @MyIgel
Commits on Jul 26, 2014
  1. @tdicola
  2. @tdicola
Commits on Jun 2, 2014
  1. @tdicola
  2. @tdicola
  3. @tdicola
Something went wrong with that request. Please try again.