The betaKnit research took a first step towards a full integration of electronics within textiles, whereas current wearable projects mainly use electronic components mounted on the surface of textiles. With the prospect of being able to create knitted electronic parts as complex as a whole speaker, the since long envisioned ideal of having soft tactile electronics comes closely within reach. If these prototypes will be professionally produced and made available to the commercial market by the textile industry this would imply a big technological advance in the field of wearable technology.
FabLabs use computer aided production processes to create objects. During this research project, ways to apply these processes into the realm of textiles were explored. Instead of using a CAD file to print a 3D print, a code for the knit-work to be “printed” would be used. The concept of Beta Textiles was introduced to describe this approach. Beta Textiles refers to the fact that fabric does not consist of purely thread, but on a meta level also contains this code of a digital knitting pattern which can be fed into production machines.
As a proof of concept for this idea of Beta Textiles, an electronic textile containing a knitted audio speaker was created. The samples of this textile were produced using a basic punchcard knitting machine. The pattern and method of production can be shared with the community as open source patterns. From there it could for example be used by individual crafters, as was demonstrated by the creation of a ‘Sound Hoodie’ based on the knitted audio speaker, or be incorporated and multiplied as a sonic module for larger fabrics by the industry.
For distribution and sharing of patterns, the research proposed to set up an online database. Such a database would feature a simple user interface to upload patterns and translates them into various types of codes. One code that can be read by professional knitting machines, thus allowing industry to reproduce it, another code for home-use knitting machines (or even hand knitting), to enable a community of DIY makers to copy and modify the pattern.
To see how such an approach would translate from an individual sample to small scale production in a FabLab with machines for textile production, V2_ set up a rudimentary Smart Textile FabLab, the eTextile Sweatshop.
The research’s main aim was to develop textile solutions to reproduce audio. A project by Mika Satomi and Hannah Perner-Wilson („The Crying Dress“) involving embroidered forms of speakers was a starting point for the betaKnit research. betaKnit however strictly focused on the method of knitting.
The concept for knitting audio speakers is based on the construction of regular speakers but is modifying the coil elements to form a plane surface. To create a fabric speaker, a coil is produced using insulated copper wire. Then a membrane and a magnet are added. If sound waves are sent through the coil, the speaker will create sound:
The audio waves sent through the coil basically consist of small electric charges that will create a magnetic field. The magnet counteracts this magnetic field and thus moves the membrane in between. The membrane then moves the air and thereby creates sound waves.
The first experiments served as proof for the functionality of creating flat forms of coils and were made by taping copper foil and wire directly to carton, whereat the carton serves as the membrane. The next specimen were already knitted with copper wire. In the later samples a round form was used that corresponds to the actual size of the used magnets. This type of coil was then also used in the production of the protosample sound hoodie.
Parallel to experiments with coil forms, the team experimented on materials for the mem- brane. Early samples were tried with carton membranes whereas later tests were made with textile forms of membrane.
Findings & Improvements
To illustrate possible applications for the speaker module, a sample prototype was pro- duced. The protosample sound hoodie created for this research is fully functi- onal. It is audible and a simple 3,5mm audio jack cable can be attached to plug it into any music playing device.
The knitted coils are able to produce sound at a decent level of audibility, but have to be held close to the ears. This creates quite discrete forms of speakers that produce sound rather locally than filling whole rooms.
After consulting an expert on electromagnetism the team was made aware that the magnetic fields mostly cancel each other out on an intertwining knitted copper wire mesh. This was improved in the follow-up samples by knitting only every second row with wire, allowing for small stable magnetic fields to form.
Another difficulty was posed by the fact that in the shape of a knitted wire the magnetic fields change direction in every other loop. This could be counteracted by putting very small corresponding opposite magnets under each knitted loop. Another solution to this problem is to change the form of the coil by turning the stitches by 180 degrees, thus creating a more stable magnetic field going in only one direction.
In the latest sample, as seen in the bottom image, two of these modifications were realized: knitting the wire with a row of nonconductive inbetween and turning the single stitches so they form whole loops. These changes resulted in significant improvements of the speakers loudness.
For improving the overall quality of the speaker both in loudness and clarity, there are a few more approaches to be tried out in future research:
1. Create more differentiated coil forms - similar to above mentioned method of turning the conductive stitches 180 degrees - to create a more stable magnetic field. This could include making the mesh wider or larger, or using more sophisticated conductive threads.
2. Another approach involves researching other types and shapes of magnets. Potentially the use of tiny magnets that correspond exactly to single stitch size and are mounted alternatingly on a base plate will significantly improve sound quality. Ring shaped magnets could be sewn together into a magnetic flexible raster to reach the same effect. It is also thinkable to create a magnetized thread or wire that is knittable for creating the counter magnet.
3. A change of membrane material would change its behavior and therefore also lead to improved sound quality. The membrane needs to be absolutely airtight and stiff under the coil whereas the anchorage of the membrane should be flexibly connected to the base material of the speaker. This will reduce the loss of lower frequencies. A better membrane will lead to improved sound in terms of frequency quality and spectrum.
About Smart Textile Services: