The very first experiments often involve the use of double sided tape. There quite different grades of double sided tape however. Cheap stuff used to bond carpets comes to mind at first. It looks very sticky to plastic surfaces like mylar. Does it offer sufficient adhesive strength? Most certainly not. The fact that tape adhesive never hardens makes it very unreliable under static stress. Even small tension will cause it to creep over time. Creeping action is first observed at corners of the panel as small wrinkles. Those wrinkles might look innocent at the first glance. However they buzz when low or even mid-bass frequencies are reproduced. This can add harsh character to the sound depending on level. Therefore almost all tapes are ruled out quickly. Some people report success by using certain high strength 3M VHB(very high bond) tapes. They appear quite expensive though. Glue and plastic spacers might be a more reliable alternative after all. A lot of different glues have been tested but only few were found to actually work well. The very thin gap between spacer and mylar prohibits evaporation of even very volatile solvents like acetone. Some glues which were found to work are discussed below.
1) Slow curing, low to medium viscosity epoxy . The application is best performed by using a syringe or small bottle with a thin nozzle. A thin line of mixed epoxy is applied over the perimeter. It is not necessary to disperse epoxy after this. Frame with tensioned film is then placed on the stator and left as is. The weight of the frame squeezes the glue to a uniform, thin layer withing 10-20 minutes. Therefore it is important to use slow curing epoxy. At the first glance the bond of epoxy does not seem very strong as it's quite easy to peel mylar off. However the shear strength of the bond is very high and film would tear long before epoxy gives up. Furthermore, the adhesive force increases over longer time.
2) Acetone based polyurethane contact adhesives. The adhesive should be diluted with acetone to a viscosity almost like water. The adhesive is clear, very fast drying and should be applied to both surfaces via brush. Timing is critical. Both surfaces should seem dry but still adhere to each other. Too fast, and wet glue is embedded between mylar and spacer and will take ages to dry completely. Too late, and some spots fail to bond. Therefore it becomes somewhat difficult to obtain consistent results. Problems quickly increase with panel size, as time required to apply adhesive is increased and uniform drying is more difficult to achieve.
1) Slow curing, low to medium viscosity epoxy . The application is best performed by using a syringe or small bottle with a thin nozzle. A thin line of mixed epoxy is applied over the perimeter. It is not necessary to disperse epoxy after this. Frame with tensioned film is then placed on the stator and left as is. The weight of the frame squeezes the glue to a uniform, thin layer withing 10-20 minutes. Therefore it is important to use slow curing epoxy. At the first glance the bond of epoxy does not seem very strong as it's quite easy to peel mylar off. However the shear strength of the bond is very high and film would tear long before epoxy gives up. Furthermore, the adhesive force increases over longer time.
2) Acetone based polyurethane contact adhesives. The adhesive should be diluted with acetone to a viscosity almost like water. The adhesive is clear, very fast drying and should be applied to both surfaces via brush. Timing is critical. Both surfaces should seem dry but still adhere to each other. Too fast, and wet glue is embedded between mylar and spacer and will take ages to dry completely. Too late, and some spots fail to bond. Therefore it becomes somewhat difficult to obtain consistent results. Problems quickly increase with panel size, as time required to apply adhesive is increased and uniform drying is more difficult to achieve.