Skip navigation

This week i have been experimenting and tesing different ways to attach the flex sensors and accelerometer to the polyester inner glove.

I have tried sewing the flex sensors directly to the glove in a kind of loop fasion but found this not only looked unprofesional but didn’t secure the sensor as well as i would like.

I decided i would need some kind of sheeth for each sensore that would itself be attached to the glove and which the sensor could slip inside.  I decided that the sheeth would need to be made of something that did not have elastic properties so chose to use a cotton handkerchief.

I cut the handkerchief into a 40mm wide strip, folded this over and sewed down the open side creating a tube 20mm wide.


This was the same width at the fingers of the gloves.  The flex sensors are much narrower than 20mm so i found that the fit was very loose so decided that i would need to sew a channel into the cotton sheeth to ensure a better fit, using the sewing maching i made a channel of 10mm to fit the 6.5mm flex sensors.  This seemed to work pretty well, now i needed to figure out a way to attach these to the fingers of the glove.

When turned inside out the glove has two hems on either side of the finger, as pictured

These were perfect for my needs, i could now sew the sheeths directly to the gloves as shown below

The flex sensors have been slipped inside the sheeths on the backs of the fingers before a small hole has been piered at the base of the sensors and thread used to sew this to the glove to reduce any pull on the soldered connections.  Thread has also been used below the tape which covers and protects the connections again to try and reduce the strain.

Now with the flex sensors sorted i moved my attention to the accelerometer.  I had already decided i was going to use a piece of vera-board sewn to the back of the glove so went about getting this sorted, i figured it would be a smarter idea to test my sewing skills before i glued the sensor to the board so below is a picture of my 3rd attempt


My gloves arrived last week so i have been experimenting with how to attach my sensors.  I have discovered that the flex sensors are going to be more complicated than i first thought, they need to be secured at one end while being free to move at the other but still need to bend with the finger.  For test perposes i have been using tape to secure one end but i will require something more durable and hard-wearing for the final glove, for the other end have used cotton thred through the bottom of the flex sensor and to the glove itself, this seems to work pretty well but i’m going to have to improve my sewing skills if this is the method i use.


The accelerometer is to be attached by first using a hot glue gun to glue it to a small piece of veraboard, then using the small holes in the board i will use needle and tread to sew the board and sensor to the back of the glove

Over the last week i have purchased a licence for Max and intalled and have it running.  I have also downloaded and installed the Maxuino software

Max is, as described by Wikipidia:

Max is a visual programing language for music and multimedia developed and maintained by San Francisco-based software company Cycling ’74. During its 20-year history, it has been widely used by composers, performers, software designers, researchers, and artists for creating innovative recordings, performances, and installations. for further details visit

a screen shot of a basic Max window

Max Screen

Maxuino is, as descrided by their website:

Maxuino is a collaborative open source project for quickly and easily getting the Cycling ’74  Software “Max 5″ talking to the i/o Arduino board.  This allows Max to read analog and digital pins, write to digital and PWM pins, control servos, listen to i2c sensors and much much more.  It is in use all over the world by artists, musicians, inventors and gizmologists.

Basically a piece of software that can recieve information from the Arduino and display this directly through Max, for more information on maxuino visit the website

a screen shot of the Maxuino Software

Maxuino Screen

the following websites help with installation

I have tested the software is working and have also ensured that i can recieve data from all of the sensors.

This week i have been considering what gloves to use and how to construct a durable and hardwearing interface.  The following research helped me form my choices whilst considering quality and cost.

here are some more expensive examples

and off-the-shelf or project examples

I want to achieve the look of a proffesional example but at the cost of a home project.  I want to attach the sensors much like the second set of gloves but want to have the workings concealled.

This can be achieved two ways while remaining cost effective, either using two gloves; one thin inner glove with the workings and a second larger and more hardwearing outer glove to hide and protect the workings or finding a large hardwearing glove that would have sufficeint room to insert the sensors

I decided that my sewing skills arent up to much so im going to use two gloves, one inside the other.  Ive been looking at small thin polyester gloves and have ordered four pairs (allowing for tests and errors) of white inspection gloves below from at a cost of £11.70

and for the outer glove i have decided to use what is described as a ‘running’ glove.  A elasticated, lightweight 95% polyester glove as show below.  Which has been purchased from DW Sports for at a cost of £5.99.

My new component arrived this morning and i have been testing it and establishing its peramiters and requirements.

Initially i have just hooked the accelerometer straight up to the arduino simply to test all is working fine but this will eventually be done with capasitors to limit the band width and reduce any noise. I used the following code retrieved from this web site

Accelerometer test

Using the data sheet i have calculated that using 0.47uf capasitors on each output will give an approximate band width of 10Hz

I’ve sketched out a diagram of the circuit i will use for the final piece which i also intend to draw up on Multisim to give a clear idea of whats going on which should be posted next week.

Ordered a 12 month student liecence for Max 6.0 today as i feel that being able to work with Max MSP and Pro tools at home will improve the quality and complexity of the final piece.  This cost a total of $60 which is about £37.

After a troublesome day i have discovered that the Digital Accelerometer i have is proving to be more compicated than first expected. After further research into the area i have discovered more support material available for anaolog accelerometers and in particular the Triple Axis Accelerometer Breakout – ADXL335

and the help page i have found most useful

This is the new component i have ordered from at a cost of £22.50 with the accompanying datasheet below.


Until this arrives i will concentrate on the software i will be using, Maxuino and either ProTools or Ableton Live (whichever had the better capabilites i require.

I need to test each of the sensors to ensure they work and their thresholds ie. voltage and Arduino serial data.

Here is the Arduino code i found to help me test the flex sensor


and using the following Voltage Divider where R1 = Flex Sensor (25 – 45Kohm), and R2 = 22kohm resistor

and using the following formula, calculated that the voltage should vary from approx 2.34v and 1.64v when Vin = 5v

and using the serial monitor in the arduino sorfware noted that the data varied from 260 when the sensor was straight to 130 when the sensor was bent.

A rough test showing that the sensor worked.  The same was repeated with all 6 sensor to ensure no defects.

After considering which sensors would be most appropriate for the types of hand gestures i i want to use i have decided on Flex Sensors and Accelerometers.  These have been purchased from at a total cost of £61.20 for

2 Triple Axis Acceletometer Breakout – MMA8452Q


• 1.95 V to 3.6 V supply voltage

• 1.6 V to 3.6 V interface voltage

• ±2g/±4g/±8g dynamically selectable full-scale

• Output Data Rates (ODR) from 1.56 Hz to 800 Hz

• 99μg/Hz noise

• 12-bit and 8-bit digital output

• I2C digital output interface (operates to 2.25 MHz with 4.7 kΩ pullup)

• Two programmable interrupt pins for six interrupt sources

• Three embedded channels of motion detection

– Freefall or Motion Detection: 1 channel

– Pulse Detection: 1 channel

– Jolt Detection: 1 channel

• Orientation (Portrait/Landscape) detection with set hysteresis

• Automatic ODR change for Auto-WAKE and return to SLEEP

• High Pass Filter Data available real-time

• Self-Test

• RoHS compliant

• Current Consumption: 6 μA – 165 μA

6 Flex Sensor 2.2″


– Angle Displacement Measurement

– Bends and Flexes physically with motion device

– Possible Uses

– Robotics

– Gaming (Virtual Motion)

– Medical Devices

– Computer Peripherals

– Musical Instruments

– Physical Therapy

– Simple Construction

– Low Profile

– Flat Resistance: 25K Ohms

-Resistance Tolerance: ±30%

-Bend Resistance Range: 45K to 125K Ohms (depending on bend radius)

-Power Rating : 0.50 Watts continuous. 1 Watt Peak

-Temperature Range: -35°C to +80°C

Data Sheets



I have been investigating Sensor types that would work well with my intended project.

For my previous project i had used Piezoelectric sensors on the tips of each finger of a glove.  The principle of a piezoelectric sensor is that under pressure it produces a voltage.  With my previous project i found that these sensors were unreliable and had a short life-span due to their fragile design.  I had the task of researching the sensors that were available and thier suitablilty for my final project.

Sensors that are commonly used in musical applications and their characteristics are as follows (Information gathered from the book New Digital Musical Instruments)

Force-Sensitive Resistors.   Typically made up of a conductive polymer film sensor whose conductance is proportional to the applied force; electrical resistance will decrease with the increase of pressure applied to the device.

Strain Gauges. Resistive elastic sensors whose resistance is a function of applied strain due to mechanical stress.  Their resistance decreases with compression and increase with tension.    Strain Gauges are more accurate than Force Sensitive Resistors and noramlly used for sensing weight,

Bend/Flex Sensors Consisting of a strip of plastic with conductive ink, bending the sensor causes its resistance to increase.  Bend sensors are very useful because they can be easily attached to body parts or textiles

Infrared SensorsBased on the measurement of properties of light signals

Accelerometers Measures linear acceleration in one or more axes and can also sense inclination.  Also useful for measuring shock.

The first task was to research the products already available in the current marketplace and where research was being undertaken in this area.  This was a field i was already very familiar with as i had undertaken research for a physical computing project earlier in the year.  Below are several links to products or research i found that was relevent to my proposed project

Commercial Examples – an american site with a collection of commercial Data gloves using bend sensors and pitch-and-roll sensors ranging between $600 for a single glove and $1400 for a pair. – shows a review of an open-source data glove using accelerometers in each finger to register independent movement at a cost of $499 for one glove. – a company specialising in advanced data gloves using up to 22 sensors to capture even the slightest hand of finger movement.  No prices given. – shows a prototype of a data glove using flex sensors but as stated on the website it remains a work in progress but visually differs from the previous examples as this has been designed with a unique and interesting appearance and is described as Hybrid Apparatus for Social Interface. – a data gloved designed for 3D games and virtual environments at only $59 using flex and sensors and accelerometer – shows a concept of a glove designed for the user to idly tap out beats with their fingers.  As only a concept to technical detail or cost available. – Piano Gloves which use sensors on each finger that when stuck on a flat surface produce piano notes at a cost of £50.

Below are some examples of projects people have undertaken

Below are some links to video clips showing different attempts by people to great music with different types of data glove.

Below is a list of books and articles that all feature data gloves or cover relevent technologies

New Digital Musical Instruments: Control and Interaction Beyond the Keyboard. Miranda & Wanderley2006.

Physical Computing: Sensing and controlling the physical world with computers. O’Sullivan and Igoe. 2004.

Transducer Interfacing Handbook. Sheingold. 1981.

Getting Started with Adruino. Banzi. 2009.