GPS path in Itasca State Park, Minnesota

The image on the left is a path recorded from my bicycle ride  in Itasca State Park.  The path starts from the headwaters of the Mississippi River and follows the wilderness drive & bike route to the visitor center.  I also road from the visitors center to the headwaters but the GPS reception was blocked by the ice pack for my lunch.

The GPS logger takes data from a Linx GPS module and saves it to a SD/MMC card.  The data is formated in a KML flie which can be opened with Google Earth or uploaded to Google Maps.  The GPS module outputs NMEA sentences at a 1Hz rate.  The GPS logger uses the GGA sentence for longitude, latitude, and altitude.

The GPS logger is powered off of a rechargeable 3.6V Lithium ion polymer battery.  The battery is charged using a USB power supply and a lithium ion polymer charger integrated circuit.

The source code for the AVR microcontroller can be found on github under the name GPS-Wolf.  The PCB board layout is also uploaded to github.

PCB designed in EAGLE

I had thought the TV in the background was mute but the audio sounds like I forgot.

Since the video below was shot, I’ve made some modifications to the code to improve it’s autonomous navigation.

From Disc golf target

The first thing I noticed was that a single ring of chain worked for most puts but a few that should have been caught got through.  The simple fix to this was to add a second ring of chains.  I added six more chains halfway between the center and the outer ring.  Very few discs actually make it through the chains.

The second problem was that discs can roll on their side out of the basket.  It didn’t seem like it should happen too often because the disc had to be on its side rolling at an angle.  The solution to this would be to either add more dowels so the disc cant fit through or to add a second plywood ring onto the dowels.  I found that after I added more chain the problem happened less often.

The third thing I noticed was a few of my dowels were not actually hardwood.   They were some lighter weight wood that looked the same.  When the disc hit these dowels they cracked where the screws went in.  I added some glue and a few zip ties to strengthen them.  The best solution would be to make sure all the dowels are some type of hardwood.

Lastly, the extra part of the u bolts that sticks up can really eat away at your discs.   Cutting the excess bolt off and smoothing it with a file helped a lot.

Tools required

• drill & drill bits
• string & pencil
• jig saw

Supplies

• 3/4″ Ply wood
• 1 in pipe
• 12 U bolts
• chain
• Large (3″) key ring (can substitute with wire)
• nuts and bolts
  • 12 nuts, bolts, and washers for attaching the Galv Floor Flanges
  • 24 nuts to fit U bolts (2 extra for each)
• dowels 3/4″ or 1″
• 2 – 1″ Galv Floor Flange
• 1- 1 1/4″ Galv FLR Flange
• 1- 1 1/4″ Galv nipple

I tried to make my disc golf target fairly close to the ones at my local course which uses the Innova DISCatcher^® PRO Disc Golf Target. Most of the dimensions for that target can be found by following the link above.  The only dimension I couldn’t find was the depth of the basket which I measured to be 9 inches

Step 1

Cut 3, 26 inch diameter circles from a sheet of plywood.  Take one of the circles and cut out a 22 inch circle within in.  The resulting ring will be used as the rim of the basket portion of the target so be careful not to cut through it.  The 22 inch circle will be the top portion where the chain is attached.  The easiest way I’ve found to cut the circles is to use a nail, pencil, string and jigsaw.  Tie the pencil and nail together 13 inches appart pound the nail in the center and use it to draw a 26 inch diameter circle.  Cut out the circle with the jigsaw.

Step 2

On the 22 inch circle draw a line from the center every 30 degrees (12 lines)  to determine where the U bolts should go.  drill holes for your u bolts.  Each U bolt will require four nuts.  A nut will go on each side of the plywood.  washers are optional.

Note: The U bolts are are sort of optional although I highly recommend them.  You could stick a bolt through the last link of the chain and bolt it to the plywood.

Step 3

Cut your chain in approximately two foot lengths (for me this was 20 links).  Slip one end of the chain on each u bolt and attach them too the plywood circle.  Place the other end of each chain on a large key ring.  The key ring can be substituted with wire but wire will not hold the circular shape.

step 4

Cut 12, 8 inch long dowels and attach them to one of the 26 inch circles (again every 30 degrees).  Attach the ring to the the dowels as well.  now you should have a basket

step 5

Cut the pipe to 55 inches and thread each end.  Most hardware stores will be able to do this for you when you purchase the pipe.

step 6

Attach the 1″ Galv FLR Flanges to the top and base circles.  Drill a 1″ diameter hole through the center of the basket, screw the 1 1/4″ nipple into the 1 1/4″ Galv FLR Flange and attach it to the basket.

step 7

slide the basket onto the pipe and screw on the top and bottom.  Position the basket where it looks right and drill a hole through the nipple and then through the pipe.

step 8 (optional)
An inner ring of chins can be added to the target for better catching.  You will need 6 extra u bolts and some more chain.  the length of chain for the inner ring should be shorter approximately  16 inches (18 links of my chain).  The Innova targets at my local course have 3 rings of chains. for a total of 24 chains.  The single ring seems to work but occasionally lets the disc through.

please see my update on this disc golf target

Octave is a free open source alternative to Matlab.  Nearly all the functions found in Matlab are the same in Octave.  So far the only function that I’ve not been able to use in Octave is xlsread.  Octave by itself is basically just a shell.  Where you can enter in commands and outputs are displayed.  By installing a front-end program such as QT Octave you get the whole environment that you would in Matlab.

At first I wasn’t confident that the two programs would be compatable so I always checked to make sure my program would work on Matlab.  After having all my projects work flawlessly between Octave and Matlab I’m confident that my code will execute on both programs.

Link to code: Sensor

For this project I learned how pulse width modulation worked and how to operate the SRF05 ultrasonic range sensor.  I decided that since the pic18F1220 had a built in PWM module, I would learn how to use that.

In order to generate a pulse width modulation the pic microcontroller uses a timer and compares the timer value to two other registers.  One of the registers (PR2 in the 18F1220) is the value at which the timer resets.  This value is related to the period of the pwm waveform.  The other register (CCPR1L in the 18F1220) is the duty cycle.   When the timer is less than this value, the waveform is high.  Otherwise the waveform is low.  When CCPR1L is equal to PR2 the duty cycle is 100%.

The SRF05 creates a pulse who’s length is proportional to the distance an object is from the sensor.  In order to begin the distance measurement the SRF05 needs a 10us trigger pulse.  The SRF05 then measures the distance and outputs a pulse that is between 100us and 25ms.  Measuring the pulse can be done by waiting for the output line of the SRF05 to go high, then setting up a timer with an appropriate prescaler to run until the line goes low.  Once the pulse is over the timer value is read and can be mathematically manipulated  to get the distance. The SRF05 recommends waiting 50ms before getting another measurement so the previous ultrasonic pulse doesn’t interfere.  Documentation on the SRF05 can be found here

For this project I used a simple loop where I got the distance from the SRF05 then moved the timer value into CCPR1L.  I then waited for 50ms before getting the distance again.  The led was hooked up to the output of the pwm wave.  The pulse widht modulation controlled the brightness of the led.

This program logically was not difficult but inexperience made the process slow. First I didn’t think of ordering an oscillator when I ordered the sample pic. After some reading in the data sheet I figured out there was an RC mode where I could use a resistor and capacitor. This seemed simple enough until I realized that I didn’t have the capacitor that I needed. I read online how someone used just a resistor and the stray capacitance from the breadboard. I decided to give this a try. This is the most unstable method you could use. It worked alright but sometimes the pic would be slow and other times it would fly through the program.

This was the project that I learned all the general information about programming a microcontroller. I learned how to configure ports for input and output and how to move and manipulate data in the microcontroller. The most valuable thing I learned was how to look up information in the data sheet. At first It was difficult to follow and understand how things were organized in the data sheet. After a while I got good at searching for key words.

Link to code:  7-segment

I had a new bicycle in my plans for mid summer but after I started having spoke problems with my Gary Fisher Nirvana agan I decided to look into a new bicycle. While the spokes were being repaired I looked at road bikes at Scheels, the Spoke bike shop, and the Flying Penguin. Scheels had a good selection but they were all over priced. the spoke also had a good selection but they didn’t have the right size frame in the store for many of the entry level road bikes. I went to the flying penguin where I bought my skis two years ago and found that they carried the Giant bicycle line like the spoke and they had the right frame size. I was still debating if I really should get the bicycle so I went home to think about it.

I went out for a ride on the repaired wheel and had another spoke break within 5 miles. I decided that It was only going to get worse and I didn’t want to spend too much more when I was planning on a new bicycle this summer so I decided to go back to Mankato and get the Giant OCR2 from the flying penguin. I got a deal on it because I already had the clipless pedals and they also threw in a water bottle and holder.

So far I have put 17 miles on the new bike and the ride is really nice. Shifting with the break levers is a little bit different but it works really nice.