P10002 Dynamic Keyboard

Alexander MoultonMarie HammerXingwang GaoAndrew Robertson

Team LeadMechanical EngineerElectrical EngineerElectrical Engineer

Project GoalsThe purpose of this project is to capture key-

strike dynamics for integration into a full keyboard

Enhance text based communication by providing an analog signal in parallel with binary keystroke data

Accurate differentiation of typing forces appliedEncapsulate typing forces with keystroke data

and communicate with a PCCharacterize human typing forces for future

projects

Revised Project GoalsOriginal project goal:

capture emotion while typingComplications:

Keyboards are binary devices Users are not trained to pay attention to how

they typeRevisions:

No association between emotion and typing patterns

Conscious user input expected

Customer Needs and SpecificationsNeed Specification Metric Ideal Value

Enhance text based communication

Establish a voltage output linearly proportional to force applied while typing

Linear relationship between force and voltage output

Couple analog data with keystroke character

Character data transmitted with analog data (Boolean)

True

System is able to measure a large range of input force

Range of force (N) 0 to 10N

Differentiate user input accurately

Users are able to establish up to 8 distinct outputs while typing

Number of output partitions 8

Use sensors with static output Variation in output voltage under static load (dV/dt)

dV/dt = 0

Output is independent of simultaneous keystrokes

Variation in output voltage with and without concurrent loads on multiple keys (V)

Compatible with modern PC/Laptop

USB protocol used for communication USB communication (Boolean) True

Output is readable by PC software

Applications able to monitor USB port can be programmed to interpret and display the data received

Pass/fail ofCommunication (Boolean)

True

2

20

d V

dF

0V

Design Concept - ElectricalAnalog data acquisition is independent of the original

keyboard designFour stages:

Thin film pressure sensitive device acts as a variable resistor in a voltage divider

Conditioning circuitryAnalog to digital conversionCommunication

Micro Controller

Conditioning Circuitry

Design Concept - MechanicalKeys: scissor switch, buckling, dome springMaterials: ABS plastic, silicone, foamMethods of Manufacturing: re-fabrication of current keyboard, rapid prototyping with ABS plastic, injection molding, machining raw material

Test Plan Sensors have a static output (i.e. no capacitive loads)

Load a sensor with a static weight and measure any variation in the output over time Establish a voltage output linearly proportional to force applied while typing

Calibrate device output (Voltage vs. Force) using weights ranging from 100g (~1N) to 2kg (~20N) A linear best-fit line should be possible

Force transmitted through the key to the sensor matches the force applied at the top of the key within ±10%.

Calibrate the device output with and without the key and spring Output of key strikes must be independent of simultaneous key strikes

A test key is loaded with a static force while a second key is fully depressed The variation in output voltage with and without the second key being pressed is

measured Characterize human typing force

Objective is to establish a baseline of normal typing force for future reference Result are compared with results from previous studies in typing force (1N to

2N) to ensure device accuracy determine the resolution of human typing force

Objective is to determine the minimum amount of force a user can consistently increment Tap key with successively increasing force average difference between keystrokes is measured

Test Data

Force (N) ΔV (mV)

1 0

5 0

10 31

Variation in output voltage for 1, 5, and 10N test forces with a second key fully depressed

Modified: y = 0.002701 - 0.046Unmodified: y = 0.002477 + 0.148%errorm = (0.002701 – 0.002477)/0.002477 * 100% = 9.04%

ΔV << Vmax/(# of output partitions)31mV << 3.7V/8 = 462mV

Test Data (cont.)Average typing force among users

User Trial 1 Trial 2 Trial 3 Trial 4 Trial 5Alex (N) 1.73 1.79 1.68 1.68 1.74Marie (N) 1.70 1.74 1.76 1.80 1.69Xingwang (N) 1.81 1.69 1.84 1.75 1.72Andrew (N) 1.69 1.78 1.75 1.72 1.80

Number of distinct partitions without feedbackUser Trial 1 Trial 2 Trial 3 Trial 4 Trial 5

Alex 6 6 7 5 7Marie 7 5 7 6 6Xingwang 6 7 6 6 6Andrew 6 6 7 7 6

Number of distinct partitions with feedbackUser Trial 1 Trial 2 Trial 3 Trial 4 Trial 5

Alex 8 7 8 7 6Marie 6 8 7 5 7Xingwang 8 7 6 8 7Andrew 8 8 7 8 8

Meeting SpecificationsEstablish a voltage output linearly proportional to force

applied while typing - PASSCouple analog data with keystroke character - PASSSystem is able to measure a large range of input force -

PASS (0 to 13N)Users are able to establish up to 8 distinct outputs while

typing – Not met, only 6 levels were achievedUse sensors with static output - PASSOutput is independent of simultaneous keystrokes – PASSUSB protocol used for communication - PASSApplications able to monitor USB port can be programmed

to interpret and display the data received - PASS

Future Project RecommendationsPrinting Force Sensitive Resistors in a matrix

underneath the keys for future keyboardsDesigning modified keyboard to hold more

circuitry as an alternative to modifying the keyboard.