Site hosted by Angelfire.com: Build your free website today!

 Well this project is finally complete. I've learned more then I could of hoped from all of this. I started this project over a year ago. First it started out with just brainstorming and research. After I got the proper test equipment, I would say I spent about 5 months of actual work on the glove. Improvement redesign. I started with an entirely analog system. That was a mistake. Then I went digital but the circuits were still to space consuming. Finally when I got my act together I went with a ATmega microcontroller. Cheaper overall, less space consuming, and easier to work with.

 

 Excluding the equipment costs this project will cost approx. 100 to 150 dollars. With equipment, unless your just doing it for the challenge, I wouldn't recommend this project for most people. The programmer for the microcontroller was 100 dollars with a 250 dollar oscope and a variety of tools. One thing I will do if you really want to do this project is burn the chips for you. Send me the money or chip and I'll burn one for you and send it out as fast as possible. That will keep your cost down. You will definitely need an oscope for troubleshooting and repair. Otherwise, it will be cheaper to buy a 5DT glove for around 495 and it will probably be a lot more "commercial".

 This project is to show that it is possible to produce homebrew VR equipment. With the test equipment I've purchased initially I can now produce a variety of other VR projects. So enough talking lets start with what you'll need.

  You will need:

1. Approx everything in this material list.

2. Desktop PC with serial port.

3. Oscope.

4. Soldering Iron.

5. Dremel with attachments.

6. ATI programmer-STK500.

7. Hack saw and vise.

8. Hell of a lot of time!!!!

  I recommend building the unit on a breadboard before putting it together on your PCB boards. It may save you a lot of time in troubleshooting and will also get you familiar with the unit. Other wise the first step is to build your custom PCB boards. This takes some time.

 

Here are some pictures of my PCB board layouts on paper. The thing to do is take a piece of paper and fold it in half. The top half will be the top of your PCB and the bottom with be the bottom. Take your enclosure and place it on top of the top section and trace it. Now flip over the enclosure and make note of the holes you'll have to drill and make note of them on your traced box. Basically you want your traced picture to look exactly like the inside of your enclosure where the PCB will be mounted. This will give you a template to draw your PCB layout.

The easiest thing for you to do will to try to reproduce the layout in my pictures. Or improve them in any way you see fit. The receiver side will be the same process except you will have a back side. Now I didn't put the crystal on the back side like the picture shows, but everything else is the same. With the receiver you want to take your drawing and fold it in half so you can hold it up to a light to see through the drawing. This way you can see where your traces on the back are relative to the front. Make your layout accordingly.

Here are some pictures of my rough PCB's with the components installed. Use this as your guide along with the schematics to place your components. I recommend using a 40 pin socket inside of soldering the IC directly to the PCB like I did. Also make sure all your holes are drilled in your enclosure before you stuff your boards. This will prevent any damage from happening to your boards.

One thing you may wish to do that I didn't was to place test points on the PCB's. A few that I would recommend are: the microprocessor out (pin 15), the receiver out, and the MAX232 line driver out. These points will be useful in your troubleshooting and signal conditioning if needed.

You may notice I may have a few extra components in my PCB pictures then in the material list or in the schematic. This is why I had packs of extra components in the material list. You may have to add caps and or resistors to help the circuit with signal conditioning. This will depend highly on the quality and layout of your PCB. The receiver picked up a lot more noise when I placed it in the enclosure compared to the breadboard. So I had to place addition caps between Vcc and Gnd and a few other key points, like the line driver. Experiment with the circuit a little and see what best fits your needs.

Once your circuit is operational hook it up to the serial port and a terminal program to see if data is being received on your port. It will be received in HEX$ format. You should be able to see a format of 80 0 X X X X X 0 0 0. The X's being the finger flexure data. The first two bytes are for syncing and the last three are for x,y,z coordinates that will be added in a later project.

Here is a test program I wrote in VB6 for the glove. It displays formatted data in decimal format and also displays a rough hand model the mimics your hand movements. It is not extremely fast so some lag is noticeable. Also here is the program that is burned into the ATmega163 chip.

It you notice that the hand model is glitching a lot, then you need to adjust the receiver on your PCB. One thing might be to drill a hole in your enclosure to adjust the receiver from the outside. You need to adjust the inductor on the receiver for the best signal. The best way to do this is with your Oscope.

Well I know this is kind of a rough, quick run down of the project but it should get your started. If you have any question please email me. I have ENORMOUS amounts of information on this project. I have spent countless hours on it, so don't hesitate.

Counter