[Weather] termperature offset on Dalsemi station

[Tim Bitson]

Ernie,

The original DS1820 temp sensor has a design flaw. The design was  
corrected and the fixed parts became DS18S20. Some of the early  
weather stations had the defective parts in them. At the time, Dallas  
Semi was offering free replacements, but I doubt that's true any  
more. You can order a DS18S20 from Hobby-Boards.com or Digikey and  
replace the sensor in your wind instrument (it is soldered in place).

Just out of curiosity, do you have the shorting jumper installed on  
your wind instrument? Usually the lack of a jumper causes the temp to  
read zero, but it could cause the problem you're experiencing.

Using the DS2438 (the 1-Wire chip in the humidity sensor) is  
perfectly acceptable. It has a slightly higher accuracy (2 degs C  
verses 1 deg C for the DS18S20)., but if your getting good reading,  
then there is no reason you shouldn't use it.

Tim
http://www.weathertoys.net



On Mar 31, 2007, at 10:31 AM, Ernie Oporto wrote:

>
> I forgot to mention that the internal sensor temp is actually 7-10  
> degrees LOWER than the real temp.
> But I can guess that because the thing is in the shade it keeps the  
> cool air inside from the night before and never heats up.  I made a  
> pagoda a few years ago for the humidity sensor out of a lawn lamp,  
> so I'll just do the same for this.
>
> Does anyone recommend AGAINST using the Trh sensor for temp?  I  
> seem to recall that had issues too.
>
>
>
> On 3/30/07, Rich Mulvey < weather at mulveyfamily.com> wrote:
> Ernie Oporto wrote:
> >
> > You can see at the link below that my Dalsemi station seems to send
> > data that is 7-10 degrees off from other stations around me.   
> Most of
> > the time this is because my station is in the shade due to my condo
> > complex's rules about where I can mount things.  Of course, at  
> night,
> > we are all in the shade and the temperature difference is still
> > there.  This is one of the original dalsemi 1-wire stations.
> >
> > http://www.findu.com/cgi-bin/wxnear.cgi?call=CW0861
> >
> > Surprisingly, the Trh in my humidity sensor (one of the original
> > humidity sensors that were sent out in samples) seems to show the
> > right temperature.  I'm sure I can fudge OWW to read the temp  
> from the
> > humidity sensor for sending to findu.com <http://findu.com>, but I'd
> > like to see if I can get to the root of the problem and correct the
> > temp.  Is there any known issue I should be looking for that might
> > cause this?  Any recommendations?
> >
>
>
>    Is this the temperature sensor mounted in the DalSemi weather head?
>
>    If so - you're *never* going to be able to get a reliable,  
> consistent
> reading from it.  The problem is that the sensor is mounted in a  
> sealed
> enclosure, and thus is not measuring the air temperature - it will be
> offset based on the amount of sunlight, cloudiness, how much heat is
> being absorbed by the weather head, etc.  So you'll generally see the
> temps running high in the daytime, and low at night.  You can't just
> apply an offset to the temps, because there are so many variables that
> will affect the measurement.
>
>    The solution is to mount a completely separate temp sensor in a
> properly ventilated enclosure - google for "weather pagoda" or
> "stevenson screen" for many ideas in that regard.
>
> - Rich
>
>
> _______________________________________________
> Weather mailing list
> Weather at buoy.com
> http://www2.buoy.com/cgi-bin/mailman/listinfo/weather
>
>
>
> -- 
> Ernie
> http://www.shokk.com/blog/
> _______________________________________________
> Weather mailing list
> Weather at buoy.com
> http://www2.buoy.com/cgi-bin/mailman/listinfo/weather

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www2.buoy.com/pipermail/weather/attachments/20070331/969be33d/attachment.htm