Weather Underground data

6 posts / 0 new
Last post
Weather Underground data

If I decide a different Weather Underground location provides more accurate information for my location, can I change that and the local weather data be backfit for the new location data, or will I just see a "jump" in weather. The one I selected is consistently reading several degrees lower than the surrounding stations.

Paul's picture
How the weather data works.

You will see a jump in the data. Change the weather station in the form. Every 5 minutes the hive Pi will pull the WX data from Weather Underground, save it locally and send it to the public database on the web server. Unfortunately, we currently do not keep track of which data came from which weather station. That would be nice. There is no way to back fill the old data with data from the new site. That might also be nice.

How about getting a DHT22 and recording the ambient temp/RH? One area we hope to research is whether a local WU PWS (could be 5 or 10 miles away) accurately reflects conditions in the yard or whether we need local sensors. I see a 4 to 7 degrees difference (higher) in the temps in my Athens, GA yard in the suburbs compared to the Athens Airport (less than 5 miles away).

So I think sensors in the yard will be necessary due to the micro climate. We are moving toward a complete weather station in the yard, although barometric pressure shouldn't have that big of a gradient.

DHT22 vs TEMPerHum?

If I decided to do the local readings, is the DHT22 preferred over the TEMPerHUM?

Paul's picture

I prefer the DHT22. It is smaller and cheaper.

The TEMPerHUM uses Type A USB connectors. They do not latch and I question the reliability of the Type A connectors. If you don't set up udev rules, sometimes when you reboot, the sensors will swap device names meaning inside temp/RH will become outside and vice versa. But the TEMPerHUM will work on any platform that has USB ports and doesn't require soldering or crimping.

The DHT22 is about 1/4 the price and 1/2 the size (if you get the sensor without the wires). It requires soldering and crimping. We use the RJ connectors that latch and are reliable. If you buy the DHT22 from Adafruit, it is mounted on a PCB with .1" headers and includes a pullup and bypass cap. In our testing, you can get by with out them.

Walter's picture
No longer at Adafruit

I've been working on a parts list. I don't see the DHT22 w/ PCB at Adafruit, only the pre-wired DHT22 or the bare DHT22 with pullup resistor. I can see that the pre-wired version is slightly larger. Not sure if that will work inside a hive?

Perhaps this is the same as the DHT22 w/ PCB-

It looks the same but not sure if it includes a pullup resistor/bypass cap?

Paul's picture
DHT22 on Ebay.

I don't like to post ebay links as they go stale so quickly but here is one with the board. I was wrong in thinking it was Adafruit. Carl or Adrian are the ones ordering these and they sent them to me for testing. This one is mounted on a board with R and C for $4.13 but doesn't have free shipping:

This one is just the bare sensor and has free shipping for $3.09:

We (Adrian, Carl and I) don't recommend the "wired version". Yes it is too big to go in the hive and at $15 each, getting close the the TEMPerHUM in cost.

At some point we plan to buy them in bulk and have them included in the dev kit. I've seen them under a dollar each in quantity (50).

(Note to self and Nate: This comment thread needs to be moved to the temp sensor Forum.)

Log in to post comments