[time-nuts] Temp/Humidity control systems?

David davidwhess at gmail.com
Thu Oct 27 21:02:49 EDT 2016

Years ago I had to deal with this and the instruments and sensors we
used matched well against dry and wet bulb measurements.  I suspect
consumer level stuff varies considerably in reliability and accuracy.

The capacitive sensors are tricky to use because they require AC
excitation to prevent damage from electromigration.  I see a lot of
integrated sensors are available now and I wonder how well they really

On Fri, 28 Oct 2016 09:39:04 +1100, you wrote:

>You are correct to question commercial humidity sensors.
>It seems to have come about because no-one can make a dollar by selling humidity.
>Manufacturers do not tell the truth, they think ours is as good as theirs so we 
>should claim the same accuracy. People buy these sensors, believe them, and buy more.
>It is not hard to measure humidity/temperature. (they should be measured together).
>Cover the bulb of an ASTM32C thermometer with cotton gauze. Insert it through the wall
> of a short length (18”) of 4” metal tube, insert another ASTM32C thermometer through the side 
>of the tube 4” upstream. Put a computer fan on the outlet of the tube sucking air over 
>the thermometer bulbs at about 4m/s. Wet the thermometer bulb, but NEVER touch it with your fingers.
>In about 3 minutes you can take two temperature readings. There are a number of tables and calculation methods,
>some much worse that others that will convert these values to air temperature, Relative humidity,
>Dew Point temperature etc. You get accuracy of 1% from temperatures measured to 0.1C.
>If anyone is interested I have basic routines for XCEL spreadsheet use to do the hard work.
>This is based on the WMO Reference Psychrometer developed by Russel Wylie of NML Australia.
>> On 28 Oct 2016, at 12:43 AM, Ron Bean <time at rbean.users.panix.com> wrote:
>> And since this is time-nuts: Measuring humidity accurately is tricky. 
>> According to people who have tested them, commercial electronic humidity 
>> sensors, when tested in a lab, have never come anywhere close to the 
>> accuracy claimed in the data sheet. The best you can hope for is 
>> consistent readings, not absolute accuracy.
>> ...

More information about the time-nuts mailing list