It is hard to believe that the time for this project is already over, but I am glad that I was able to finish everything on time. I have posted some of the final pictures of the Glove below, and I will either post more pictures or try to get a video of it working. Most of the wires will be hidden under the cover that I wrote about on Post 1 of Week 8, but I didn’t take the pictures with it on. Other than that, hopefully you will like what you see…
In the past few days, I have further modified the power supply that I will be using to power my components, and have continued to prepare the wires for their final connection to the sensors. I have also decided that the circuit board will sit on the glove itself, most likely where the glove’s strap is located. I have attached a second circuit board to the original board with all the components, resembling two stories of boards with enough clearance between the two. The bottom board is to protect the connections of the top one from being damaged by direct contact to the glove. The bottom board will then be attached to the glove by a strip of velcro, making the board removable and easier to maintain.
I will be starting to write the proposal for the final stage tomorrow, and will post it as soon as it is ready.
This weekend I finished a little power supply that will take a 9V DC battery voltage and regulate it down to 5 V. This will power the PIC, the LCD screen, the TTS module, etc. I have included a switch as well, to let me turn the power on and off instead of having to connect and disconnect the battery each time. The power supply will go in the back of the speaker box. I have also been working on the connections between the wires and the sensors. I have soldered pins to the end of the wires, so that these pins can be plugged straight in to the sensors. I soldered the accelerometer to a little mini-board, and this board will be mounted on the top of the hand. I will try to get some pictures of this to help give a better idea of what is going on.
Here are the pictures that I have promised. The first two are of the circuit board with the new additions. You will see more wires for the sensors, wires for the accelerometer, sockets for the wires coming from the LCD screen and TTS (text-to-speech) module, and apotentiometer in order to control the contrast of the LCD display. The potentiometer is the small blue rectangle on the bottom right of the board.
The next picture is of the sleeve that will hold the touch sensor that goes on the thumbnail in order to help recognize the letter ‘E’. The next sleeve picture is one for the sensor on the other side of the thumb for the letter ‘O’. The third picture is for the sleeve that sits on the left side of the middle finger. The touch sensor that goes here will be used to determine if the gesture being signed is ‘U’ or ‘V’.
If you want a little challenge, go to my AMA Glove Reference Material Page, and by looking at the chart, see if you can reason why the sensor locations that I mentioned in the previous paragraph would help to differentiate these gestures from others.
The last two pictures show a little cover that I made for purely cosmetic reasons. I just wanted to have something to keep the sensors from being in plain sight and to maybe protect them from the elements of nature. The first one shows the actual size of the cover, but after it is bent around the glove, it would look like the last picture.
PICTURES
Circuit board with new additions…
Another view…
Touch sensor sleeve for ‘E’…
Touch sensor sleeve for ‘O’…
Touch sensor sleeve between index and middle finger to differentiate ‘U’ from ‘V’…
This week I have continued adding components to the circuit, and will post pictures of the progress as soon as I get the chance. I have been working on how the sensors will be placed on the glove, and sewing the sleeves that will hold some of them in place. I have also been working on the programming code that I will need to make this project work. The programming is very tedious and repetetive, you have to test and try, test and try, etc. What I am doing is obtaining the sensor output on the LCD screen, and then using that value to makeif statements for the final code. In the end, if the conditions for the if statements are met, then it will produce the programmed output, which in my case will be a letter of the alphabet.
This letter C will only be output on the LCD screen if the conditions next to the if statements are met. If the sensor on the pointer finger shows a value between 190 and 230, the middle finger sensor shows greater than 260, and the ring finger sensor shows less than 340, then the LCD will show the letter C. The actual code is more involved, but this simple example should help one to understand the process in determining the correct gesture being signed by the user.
Apart from working on the sensor sleeves on the glove, I have continued to solder more components to the circuit board. The newest additions are the back to back 10 kΩ resistors and the black and white wires. The resistors, which are designated by the brown, black, and orange stripes, are each part of the circuit that will connect to the sensors. They help to form the voltage divider circuits that I metioned in one of the previous posts. The black and white wires will form the actual connection to the sensors.
PICTURES
Glove with a few sensors in place…..
Circuit board with wires to connect sensors and 10k resistors….
I have started to work on fitting the sensors to the glove. I have cut little openings on the glove where each finger connects to the hand, so that the flex sensor going inside of each sleeve can come out of this opening and connect itself to the wires that will interface them to the PIC. This means that the flex sensors will be for the most part out of sight, because they will be under each sleeve and will only be exposed for the actual connection. I am currently working on sewing the sleeves that will hold the touch sensors that will sit on the proximal phalanx of each finger. You can look up what part of the finger this is in the AMA Glove Reference MaterialPage, it will be the second picture.
I posted three more pictures on last week’s post, trying to show my setup and how the LCD displays the value output from the PIC. I tried to make a video, but the quality was terrible because of bad lighting. I will try to get it working on film again today, hopefully it comes out better. In the video you will see how the numerical value representing the voltage seen by one of the PIC input pins, changes as I bend the flex sensor, or as I press the touch sensor, depending which one I use for the video.
Since last week, I have finished soldering the capacitors and oscillators to the board, and have begun to work with the sensors and their corresponding numerical output values. I am still using the breadboard for this testing phase, and once I am happy with the results, (and find the wires that I need), will begin to solder the sensors to the board through those wires. I connected one of the flex sensors to the PIC, and wrote a simple program to have the PIC display a numerical value based on the output of the sensor. That value will be shown on the LCD screen of the box that I built last week.
The way this works is with the concept of a voltage divider. In electronics, a voltage divider is a simple circuit that produces an output voltage that is a fraction of its input voltage. That fraction depends on the values of the resistors in the circuit. The circuit and corresponding output voltage equation are as follows.
In the pictures, you will see that I took a screenshot of my computer while I was testing out the PIC on the breadboard for the first time. The window in the background is the compiler, which is where I write and troubleshoot the code for the PIC. The small window on the right is the PICkit programmer, which lets me download the compiled program to the actual chip. That specific code in the window turns on an LED that I connected to one of the pins on the PIC. This was just to make sure that everything was running properly, which it did.
The final components for the microcontroller circuit finally came in, so I was able to finish soldering them tonight. In the pictures, you will see the small blue capacitors and the silver crystal oscillator on the left side of the PIC socket. The oscillator serves as an external clock for the PIC. The clock is a periodic waveform, which means that it repeats the exact same shape again and again. This clock is used as a synchronizing signal that is used to execute the instructions and commands of the program code. Think of it as the heartbeat to the microcontroller, so without it, the system would cease to function.
The PIC has a built-in oscillator, but the external oscillator lets you achieve much higher frequencies, which results in faster operation. The external oscillator frequency is 10 MHz, which means there is 10 million cycles per second. The PIC also has a 8x PLL, or phase locked loop circuit, which multiplies the external oscillator frequency by 8. This means that my PIC will be running at a total of 80 MHz, or 80 million cycles per second. Fast, right?
I posted another link in the Electrical/Electronics Theory LinksPage as well. It is to a pdf called Microcontroller Programming – The Microchip PIC. It should help to understand anything you might want to know about electronics, circuits, microcontrollers, etc.
This coming week I will finish soldering the connections of the components that I have so far, and will give the PIC a test run to make sure everything is correct.
In the pictures, you will see that I set up a circuit on my breadboard just to make sure that the PIC and sensors work. The green LED on the right will be lit up whenever there is power being delivered to the PIC, so it serves as a reminder that the circuit is on. The next picture shows that when the flex sensor is hardly bent, the resistance is large, shown by the multi-meter as 57.5 kΩ, which is 57,500 ohms. The second picture shows the resistance measured as 9.7 kΩ when the sensor is bent almost 90 degrees.
I also posted two videos on youtube to show the sensors in action. Please forgive me on the poor video quality, I had to use the camera on my phone…
The first video shows the flex sensor controlling the brightness of the red LED. When the sensor is not bent, the resistance is very large, so there is not enough power being delivered to light up the LED. As I bend the sensor, the resistance goes down, turns on the LED, and increases the brightness correspondingly.
The second video shows the touch sensor as the means to controlling the LED.
Since the last post, I have been working on purchasing the parts and assembling the box that will house the speaker and LCD screen, which will let anyone hear and see the sign language translation output from the Glove. I bought a little decorative box from Home Goods that seemed to be the perfect size for what I wanted. I also purchased a 4 ohm, 4” diameter speaker, and the LCD screen. After the text is translated to speech by the Text-To-Speech Module, the TTS Module will output the speech signal through the SP pins, and finally the audible speech will be heard coming from the speaker. The PIC will also send the translation data to the LCD screen, from which you will be able to read the finished translation.
I have not been able to do much more with the soldering because I am waiting for some components that I ordered online to come in so that I can continue setting up the PIC circuit. They should be coming in this week, and as soon as they do, I can finish it and power the PIC to make sure it works fine. I have also been looking for the wires that I am going to use to connect the sensors on the glove to the PIC microcontroller. I am looking for the thinnest wires available so that the wires do not add on too much weight to the user’s hand, and to prevent the sensors from being stressed by a connection to heavy and inflexible wires.
I have continued soldering the components for the PIC circuit to the board. As you will see in the pictures, I have added a capacitor to the right of the IC socket. This capacitor is connected between the power and ground, which means one leg is connected to 5 volts, and the other to 0 volts. The capacitor serves as a ripple filter for the power supply, which means that it will make the power source more stable by helping to prevent voltage fluctuations. I soldered the 6 pin header below the socket; this is where the PIC and the rest of the circuit will connect to either my computer for programming or to a power supply when the project is complete.
I purchased the glove that I will be using to house the sensors. This glove might ease the process because it comes with little sleeves in the fingers, which is perfect for holding the sensors, so I will not have to sew as many sleeves of my own. It also has flex zones and finger gussets for moisture management and flexibility, and air-mesh on the back of hand, which adds breathability and comfort.
I also created another post in the Pages section, called Electrical/Electronics Theory Links. I have been trying to explain myself through this process in a way that anyone with a non-technical background will hopefully be able to understand. If not, I believe these links cover most of what I have mentioned and some other useful information as well.
The sensors finally came in today. It’s about time!
UPDATED EXPENSES
Component
Company
Quantity
Cost
Flex Sensors and FSRs
Images SI, Inc.
20x
$207.10
Memsic 2125 Accelerometer
Radio Shack
1x
$32.99
Mini PC Board
Radio Shack
1x
$1.99
Emic Text-to-Speech Module
Parallax
1x
$72.99
13 Pin SIP Socket
Parallax
2x
$1.98
16 Pin Single Row Header
Parallax
1x
$1.99
Under Armour Yard II Glove
Sports Authority
1x
$29.99
Total
$349.03
PICTURES
The glove…..
From top to bottom, 16 pin header, 13 pin SIP sockets, Text-To-Speech Module…
I have finished soldering the 40 pin socket where the PIC microcontroller will be placed. There is a picture of it below. Not bad, right? The pushbutton and resistor that you see are part of the circuit to set up the PIC. Pressing the button will allow me to reset the microcontroller in case of an error. The way this works is by pulling pin 1 of the PIC, the reset pin as seen in the pin diagram in the Component Specifications Page, to a low state. Being in a low state means that the pin of the microcontroller sees an electrical ground, ideally 0 volts. This is accomplished by pressing the button. With the pushbutton unpressed, the pin is in a high state, which is achieved by a connection to a 5 volt source, through the resistor that you see. The resistor is placed to prevent damage to the PIC by limiting the amount of current flowing into the pin from the voltage source. Therefore, as soon as the button is pressed, the pin will be forced into a low state, and the microcontroller will reset. There are more components for the microcontroller circuit, but I will describe their purpose as I install them so that you may follow their descriptions along with the corresponding pictures.
I have now ordered the Emic Text-To-Speech Module from Parallax, along with some other hardware that I need for building the circuit for the glove. It should be arriving sometime next week, and hopefully the sensors come around the same time as well. I started soldering the 40 pin IC socket to the board. The PIC microcontroller will then just need to be plugged into the socket when it is ready to be tested. The purpose for soldering a socket to the board, and not the microcontroller itself, is so that if the PIC stops working for whatever reason, all I need to do is unplug it and replace it. This is much easier than having to unsolder the 40 pins and redoing it for the replacement chip, so it is a good method for preventing headaches in the long run. There are a few more little components that I need to install to get the PIC running, and I will keep taking pictures of the progress.
I have officially started the project. I started ordering supplies online and made my first trip to Radio Shack to look for what I will need. For the sensors, I have chosen to go with Images SI, Inc.’s Uni-Directional Flexible Bend Sensors and their .5” diameter Force Sensing Resistors. I have ordered ten of each, to account for what I need and a few extra as backup. They should be getting here in a few days. At Radio Shack, I picked up the Memsic 2125 Dual Axis Accelerometer and a mini PC board that I will use to assemble the circuitry needed for the microcontroller and the rest of the Glove. I am still shopping around for the last main component, the Text-To-Speech Module, but I am pretty certain that I will go with Emic’s TTS Module from Parallax. I am just looking to see if there is anything else I might need from their site before I place the order.
For a more detailed description of how the Glove works, or other background material regarding this project, please visit the different links in the Pages section.
My name is Ernesto Vivanco, and welcome to my blog.
A few weeks ago I submitted my project proposal for the 2009 NSCS Service Initiative Grant Competition, which is an opportunity to compete for a $15,000 “service-ship” that would provide the financial support needed to make a difference through one’s own original service initiative. Each of the selected applicants would then receive an initial “service-ship” of $1,000 to go toward financing the beginning phase of the project. At the conclusion, the project that made the biggest impact in the allotted time frame would be awarded $15,000 to take it to the next level.
A few days ago I was notified that I was one of five finalists. Last Thursday, I was in a conference call being interviewed by five members of their committee. Today, I am one of the two finalists that was selected to compete for the $15,000 grant. Cool, right?
I still can’t believe that I made it this far. But obviously, the real work begins now.
For the following nine weeks, I will be posting pictures and explaining the progress made. Hopefully, I can even get video clips when everything starts to come together. You can find the proposal in the Pages section, along with other documents to keep you informed of what the project is about and relevant electronics backround material.
