Weather Station Hardware > The WxTech Dream Machine

The Dream Machine - II

(1/2) > >>

The Dream Machine - II will use individual sensors communicating by XBee radios.  The Arduino Uno microcontroller connects to an XBee radio and to the host PC via USB and/or an Internet connection using an Ethernet shield.  Several programmable XBee radios form an outside network to allow flexibility in locating the weather sensors.
The XBee radios cost about $21 each and has a range of 30 meters (1 mile for the $40 pro version).  The radios can have a dual function as repeater for distant radios.
I have the parts for this but will concentrate on the VN1 version first.

Sketch of ideas while reading Arduino and XBee books.  I forgot to include a display console.

I realize that this thread hasn't had any comments in sometime and I'm a new member as of today so just trying to get caught up.

To give you a little background on myself, I have a small business that I started some time ago but just never really did much with it. It was there just in case I got an opportunity to do something that I was interested in. The small business is a side project, but my normal dialy job is as a Senior Electronics Technician for the NWS repair depot in KC. My shop is responsible for all the Upper-Air balloon tracking systems, non-ASOS surface and hydrological equipment including the COOP network.

I have a passion for trying to redesign alot of this old equipment and am currently in the process of trying to take the current NWS min/max temperature system and convert it to a wireless device utilizing a custom config I designed that uses an arduino ATmega2560 processor board (Chip only), a 24 bit ADC w/ 8-Channel mux (for the thermistor reading), and utilizing 900MHz Xbees for the RF link. I'm looking at the option of an ethernet or Wifi connection once I get the info inside the house.

There are a few other projects that I'm working on including a NOAA Weather Radio Receiver / SAME Decoder w/ integrated GPS and data logger.
Very first iteration of just the basics can be seen here.

Larger scale version of the radio itself in its initial pprototype form is here.

I did design a consumer breakout for the SI4707 IC receiver SAME decoder here.

I don't have pictures yet, but I'm also working on an SDI-12 Arduino interface board as well.

My reason for saying all this is that I would be very interested in collaborating on a dream machine if it was still something that enough people were interested in. I don't have all the know how, but enough to give it a go. My vision, at least within my NWS projects is that the NWS will never be in the sensor making business, leave that to the companies with the funds to do the proper R&D and produce those. Building a logger to extract the data from those sensors shouldn't be that hard to do though. Most sensors are either I2C, SPI, Serial, or Analog with a few exceptions. The Arduino will do all of those, so why not harness the open source information available to that platform to build to suit.


Thank you for your interest in 'The Wxtech Dream Machine'.  I'm Wxtech.  AKA Al.
My main interest was in building a replacement for the Nimbus max/min temperature system.  I'm a Co-Op in Downtown Lexington, Ga.  LXTG1.
I don't have a lot of confidence in the MMTS and use back up systems to verify the data.  A modern sensor to the display link should be wireless.
Only a few members have shown interest in helping to design a dream machine.  Lately there has been Davis VP & Vue members' ideas/activity using the Arduino, Raspberry Pi, Picaxe, etc.  Our members will jump in and build after a prototype is produced and kit parts with instructions are available.
I have ideas but lost my focus on this project.  We need someone who can rally several of our member builders into a common direction.   

I understand your issues with the NIMBUS. Most of the time the issue doesn't necessarily rest with the sensor itself, but within the way the NIMBUS itself works. The thermistor is actually a fairly precise unit (spec'd to within +/- 0.3 deg C). The NIMBUS however uses a resistance to frequency conversion through a PLL. There are a series of calibration resistors within the device that will typically become out of tolerance over time and therefore cause the drift in readings. They are also highly susceptible to RF interference because of the frequency conversion.

I can't make any promises, but this is the system that I'm redesigning for 900MHz. We'll see how it goes.


[0] Message Index

[#] Next page

Go to full version