Dogs who are likely to need this collar are active in all conditions.  So, in order to keep all of the electronics safe, we really need a case.  I just designed this one.  You can find a render of it in the gallery.  It includes supports for the board inside allowing it to be secured using #4 sheet metal screws that are 1/4" long.  The case uses an 3/32" o-ring that will need to be cut to size, and the top screws down using #4 sheet metal screws, compressing the o-ring and making a water resistant seal (I can't say waterproof because I have not tested it yet.)  There are 4 tabs with 1/4" holes in them that allow you to use paracord to attach the case to the dogs collar.  The solidworks files are attached for editing, and the STL files for the top and bottom of the case can be found there as well.  You can use the STL files to 3D print the case yourself or send them to a service to have one made for you if you do not have a 3D printer at home. 

The original design for the GPS dog collar includes the use of Li-ion batteries.  The batteries are rechargeable and have significant power capacity to weight ratio which is ideal for this mobile application.  The only issue is that there is no way to know how much charge the battery has.  I am working on integrating a "fuel gauge" IC that will allow monitoring of battery status so that the battery won't just go out and you lose your furry friend.  It will likely be included in board V009.  Board V008 will be posted soon and includes the 555 timers instead of the 556 timers.

Board Update

The LM556 dual timer is going to be discontinued, so I working on an updated circuit using the 555s and am going to integrate it into the PCB.  There will be an updated set of boards that I will be posting shortly.  

Testing and comparing GPS modules is a tedious process because claims often vary greatly from actual performance. Thankfully, our friends at Sparkfun have done the work for the ones that they carry -Sparkfun GPS Tutorial. After reviewing the information they found, I decided to use the Copernicus II because it consistently had the most accurate results.

NOTE: This GPS technology has the correct amount of accuracy for farms or people with yards over 1 acre. In order to be accurate enough for most city yards, RTK GPS technology would be necessary. The cost for such technology has dropped significantly in recent times, but it is still expensive relative to the WAAS GPS of the Copernicus II and at the time was not necessary for my needs. WAAS GPS technology was developed for landing aircraft in the USA without having a visual of the runway and is fairly accurate.

RTK Accuracy: Plus or Minus 2-3 centimeters XYZ

WAAS GPS Accuracy: Plus or Minus 1-2 meters in XY and Plus or Minus 2-3 meters in Z

When I first began this project, the only microcontroller family that I was familiar with were Arduino boards. But for a large range geofence, I decided that it was also important for it to be a connected device. With all the excitement about pet tracking, it seemed like a logical choice. Also, with all of the APIs already created for particle, the ease of improvement and innovation on the basic concept would be seemless. I am not a programmer by trade. What I know, I have learned from others or the internet, so I am excited to see the innovative additions that will be created on top of this open source project.