Group SD0615

From ECE Department Wiki

Description

The objective of this project is to design an interface with a PC that would be useful to a consumer who has an inability to use a regular mouse. The form is a shoulder mounted package that communicates wirelessly with a footpedal base. The base contains a wireless receiver, and houses the left and right click functions. The transmitter package mounted on the consumer would house all the movement functions normally associated with a mouse.

Members

Left to right: Cody Doll, Amber McNeal, Jordan Lucht

Requirements Capture

Goals:

       -To provide a functional interface to a PC that would aid a disabled person.
       -Must be easy and intuitive to use.

Environmental:

       -Wireless frequency must not interfere with other known devices.
       -Long battery life on 2xAA rechargeable batteries.
       -Entire package must be comfortable to wear, and lightweight.
       -Must not interfere with typing.

Mechanical:

       -Floor pedal clicking mechanism with functional “left/right-click” buttons.
       -Interface in the form of a plastic enclosure mounted on an armband, possibly fastening with Velcro.
       -Quick shut-off switch located on foot pedal device.

Electrical:

       -Accelerometer sensor (two dimensional)
       -Wireless interface to receiver base
       -Battery life greater than 20 hours
       -USB interface to PC
       -Easy to use USB mouse drivers

Key Features:

       -Wireless connection to receiver base.
       -Two dimensional movement will be controlled through the movement of the arm using motion sensors (accelerometers).
               -Axis will be mapped directly (up = upwards movement, etc.)
       -Right/left click and kill switch functions mounted in a floor pedal form to be used by the foot.
               -Will be incorporated with wireless receiver base.
               -An example would be a gas pedal and brake interface.

Constraints:

       -Inaccessible client
       -Lack of experience with wireless communications
       -Computer problems in assigned lab; Y: drive inaccessible at times, printer is not networked.

Design Options

Different mounting options have been considered for the wireless portion of the project. Leg-mounted and head-mounted were considered more than other options. The leg-mount was eliminated because of the exhaustion that excessive leg movement causes, and the head-mount was decided to be too awkward to use effectively. The final mounting position decided on was the upper arm (shoulder area). “Clicking” was originally planned to be on the arm mount, but because of the basic reason for this project (handicapped users) it was decided that a simpler arm unit would be better. Adding buttons to an arm unit would mean pressing the buttons with the head or chin in order not to disturb arm movement, but that was decided to be too difficult for the user. The clicking functions will be placed on the base in a footpedal form. We also wanted to add a scroll wheel, and we could not think of a way to effectively integrate this function into a slim design. If the arm mount became too bulky or awkward, nobody would find it useful or comfortable to use. The footpedal base will incorporate the wireless receiver,USB connectivity, and click functions. A scroll wheel may be mounted on the base as well to make viewing documents easier. Also included will be a “kill switch,” which turns off the movement of the mouse. This allows the user to reset the position of the mouse, similar to when one picks up the mouse and moves it. The functionality of the arm unit could best be achieved by using an accelerometer as the basic functional unit. The accelerometer has two axes, mounted orthogonally (x-y style) and has two analog outputs with high sensitivity. We have paired this with a PIC's A/D converter that will output both inputs into a single digital signal. A single channel is far easier to transmit and decode than multiple channels. This signal will be received by the footpedal unit, decoded, and sent to the USB port of the PC along with the click data.

Current Progress

We have all of the main components for the arm mounted package, and all of the components function as planned. We have yet to get them all working together. We have the processing components for the footpedal base as well (wireless receiver, PIC, and USB module). Our wireless transmissions are strong and reliable, and should not be a problem in the final design of the project.

1/20/07 The arm components all function together, and output a fairly strong wireless signal. We are going to continue with ordering enclosures, and we are continuing programming and circuit design for both the arm and footpedal units. Budget, requirements, and timeline documents have been updated.

1/26/07 Work on the PCB layout for the arm unit has begun, along with the battery recharging circuit. We ordered more free samples (mostly for the recharge circuit) and ordered our footpedal enclosure. We will continue with work on the antenna and PCB. Programming will begin shortly for the footpedal unit, and we have yet to select a final PIC for the arm unit.

2/2/07 Arm enclosures were ordered as free samples, saving approx. $80. Charging circuits are to be tested next week for our rechargeable batteries, and enclosure modifications can hopefully get going. PIC programming will continue this weekend and hopefully be done by tuesday. RF antenna work has progressed, and several antennae have been ordered for testing.

2/11/07 More free parts have arrived, giving us freedom to hopefully build the entire arm unit. We experimented with charging circuits, but are not reaching the charging rate we hoped for. Our enclosures arrived, and we are opting for 2xAAA NiMH batteries, which will still give us adequate battery life while freeing us to use a smaller enclosure. Programming for the footpedal unit is still being worked on. Our antennae have arrived as well, allowing us to experiment with those this week.

2/21/07 This update is a little late, but that is due to the fact that we have encountered some difficulty with the networked drives and as a result were not able to get very much done. We have a working battery charging circuit, but we do not know how long it can function at 2.4A before overheating (the CADET cannot supply 2A for very long, the fuse eventually limits current flow to 500mA). Once we get an adapter we will know the true capabilities of our charging circuits. We have a standalone arm unit built onto a breadboard, and it successfully transmits a signal. The signal is not exceptionally clean, but this could be due to many factors (no grounding plane, built on a breadboard, etc.) Fortunately the entire arm unit draws less than 4mA, which on 2xAAA batteries with a combined 1800mAh will give us approximately 500 hours of battery life on one charge.

3/3/07 Due to overheating of the batteries during charging using our circuits, we have decided to scrap the charging circuit and work on other things. The USB module is communicating with our PC, but is not yet sending the data it needs to. Layout of the arm circuitry is nearly finished, and boards will be ordered before spring break. The code for the arm units has been shortened and cleaned up, and we added a third axis from the accelerometer into data acquisition. We are now sending 36 bits at 10kHz, at 100 samples/s.

3/26/07 The arm unit PCB has been designed and turned in for manufacturing. We have added a low battery indicator and power/low battery LED to the arm unit. Code has been modified for the arm unit to send 31 (not 36) bits at 9.6kHz, 50 samples per second. The footpedal code has been written as well. It receives data at 9.6kHz, condenses, formats, and sends it at 19.2kHz, 50 samples per second. Data includes x and y movements, scroll wheel data and L, R, M click functions. We are working out an issue with negative numbers in the footpedal code, and are working with USB drivers as well.

3/30/07 Arm PCB has not been manufactured yet, there have been several problems. The footpedal code is nearly complete, but has been put on a back burner until we have the USB functioning. As soon as the USB drivers work, we can finalize a footpedal PCB and finish programming.

4/09/07 The arm PCB has come in and been soldered (worked on the first try!), as well as mounted in the arm enclosure. The wireless signal seems strong and clean coming from the arm unit, so all seems to be well with that. Everything fits in the enclosure with room to spare. Amber and Cody worked on the PCB for the footpedal, and managed to manufacture one in-house that seems to work. We have yet to add all the necessary components, but we are optimistic. Our buttons came in the mail finally, but there are problems with the length of the buttons vs. the depth of our foot enclosure. The most significant thing is that we switched our footpedal processor from the PIC16F876A to the PIC16C745, which has built in USB functionality. This totally eliminates the need for our USB module. Luckily, the two PICS have the same footprint, and only minor modifications were needed in order to make the footpedal PCB work with either configuration (16C745 or 16F876A/USB Module combo). Jordan found sample code online that mimics a mouse on the new PIC, moving the cursor in a circle (a godsend, actually). The code was compiled using MPLAB and a trial version of HITECH PICC. The USB communication works, and the mouse cursor moves in a circle. Now we just need to modify the code to send OUR data to the buffer ,not sample data (which should not be much trouble). Seems like there is some hope after all...

5/8/07 No updates since a month ago, so here goes... The project is finished, and documents are getting turned in. The final product is wired, since we could never troubleshoot the wireless. The scroll never worked either, so there are two unused buttons. Jordan programmed in sensitivity, so the mouse moves at three different speeds. I could fill this up with a lot of technical writing, but everything is covered in the final report linked below.