Announcement

Collapse
No announcement yet.

Anemometer calibration...

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Anemometer calibration...

    I've built my own anemometer and am trying to calibrate it. The mcsTemperature User's Guide states:
    8.3.5 Wind Speed Calibration
    The conversion from counts to wind speed on a DS9097/DS9490 interface is

    1.2265 * Counts / Sample Interval * WindCalibration

    For a standard AAG wind instrument that operates normally the WindCalibration is 1. Adjustments can be made in the upward direction by increasing the calibration in integer increments.
    I am finding through experimentation and direct measurement that the formula does not seem to reflect what happens in mcsTemp. I have also discovered that non-integer values in the calibration field in the browser interface are accepted and seem to work.

    Direct measurements of my test airflow place 83 counts at approx 9 mph (794 ft/min).

    Using a sample interval of 10 sec and a calib value of 1, 83 counts yields a display of 8 mph. The same interval with a calib value of 2, 83 counts yields display of 16 mph. By trial-and-error, I came up with a calib value of 1.208 to provide a velocity display of 9 mph.

    Michael, would you be able to verify the currently implemented calcs and let me know if they differ from the manual?
    Best regards,
    -Mark-

    If you're not out on the edge, you're taking up too much room!
    Interested in 3D maps? Check out my company site: Solid Terrain Modeling

    #2
    Ok, I have it now. While I have the sensor update rate of set at 10 seconds set, by enabling 'general logging' in mcsTemperature I discovered that sensor readings happen more like every 13-14 seconds. That's why my calcs were not working.

    So, the calcs in the manual are correct and my anemometer appears to be calibrated. Now to find a wind tunnel and verify my calibration!

    Thanks again for a great plugin Michael!
    Best regards,
    -Mark-

    If you're not out on the edge, you're taking up too much room!
    Interested in 3D maps? Check out my company site: Solid Terrain Modeling

    Comment


      #3
      When we were collecting aerodynamic data for a space vehicle we used the wind tunnel at NASA. It was rather expensive. A vacuum cleaner may be good wind source, but then you still need to find a golden unit to verify against.

      The sampling time is actually time between when one cycle finishes and the next starts and is variable due to the multitask environment of Windows. If you have more sensors and slower ones then the interval will skew even more. I believe xapmcs1wire does a better job of holding the rate, but it even has variation expecially when the lower rates sensor intervals align with the higher rate one so something needs to get pushed to the right.

      As you noticed even though the sample interval has jitter and skew the interval is timestamped so the speed calculations should not be measurably affected by the variation.

      Comment

      Working...
      X