The GPS implementation of the sensor fully works. We have menus that allow the user to scroll through different options such as starting a run, reviewing a run, calibrating sensors, choose GPS or GPS and INS, and deleting single or all runs.

We will now focus on getting the INS implementation working. All we have left is coding. If we can't get it done, we will work on it after Design Day since we don't have a sponsor to give the prototype to.

We received the new dspic30f6012a two days ago and managed to get two of them connected and programmed. The new chip boasts 4 times the RAM and 3 times the program memory of the previous dspic30f4013. This new chip provides enough RAM for the Kalman filter and anything else we might want to add.

We are having problems with our SPI flash chip. It does not respond to any commands, despite many attampts and we believe it might be blown. We have ordered 2 new (different) chips that are purely SPI and not the SPI/SQI mix.

We have run into a problem with the current dspic we are using. The Kalman filter requires more than 600 float variables to store data. This means we need 600*4 = 2.4kBytes of RAM JUST for the Kalman filter, but we only have 2kB of RAM on the chip. We have ordered new dspics that have 8kB of RAM which will be more than enough for our application. Since the new processor is in the same family as the previous one, the code should easily be ported over to the new chip. The Kalman filtering code is close to being done, so once we have the new chip, we can test the whole system.

We have begun thinking about the final hardware requirements and ordering the required parts. We have figured out the INS equations and they have been programmed in, but we have yet to test them. We are currently working on figuring out the Kalman filter that will be used. We have received a new LCD screen that runs at 3.3V that is directly compatible with our previous code.

We now have the GPS talking directly to the dspic. The GPS chip has been configured to only send two of the five messages to the controller. This surprisingly reduced the current use by about 30mA! We decided to only recieve two of the messages because the others are not of use to us. This should also save us some computation time, since the INS will need quite a bit.

We have had nothing but problems with the pic32. We attempted to use an ICD3 to program the controller, but no luck. We have decided to scrap the pic32 and start using the dspic30 series chips. One benefit is that some of these chips have 12-bit ADC, which will give us 4 times the resolution for the INS.

Mike Carpenter and Mike Blair have successfully recieved a GPS fix from the GPS chip. They are using a computer to receive data from the GPS through the serial port. We will need the microcontroller to talk to the chip rather than the computer.

We got all our parts in from Spark Fun. The accelerometers and gyroscopes work as expected. GPS will take a bit longer to test.

One problem that we have come across while searching for parts is the packaging. Most of the GPS, accelerometer, and gyroscope IC's that we have found come in a package that we do not have the means to solder to a board. Luckily, we have found some great products at www.sparkfun.com that appear to be exactly what we were looking for. They essentially took the same chips we were looking to use, but have already mounted them to small breakout boards (most smaller than a quarter) that should greatly aid us in the design of our product. One downfall is that the boards are larger than the standalone chip, and thus will force our design to be larger to accommodate the larger parts. This is a sacrifice we're willing to make, and hope to make up for it during our PCB design phase.

Our team has started working on our project. So far we have decided to pursue a Global Positioning System (GPS) solution that will be integrated with an inertial navigation system (INS) to provide the most accurate data to skiers or snowboarders. We believe this design will be an improvement over other designs, such as the purely GPS solution implemented by the Spring 2008 design team.