There are two factors at play that cause this. The first factor is that encapsulants are applied thicker than their dielectric counterparts. This means that more UV light is absorbed or reflected before it can penetrate all the way to the bottom and fully cure. The second deals with the incident angle of UV light. The light intensity increases as the incidence angle approaches 90 degrees which is the situation encountered in most dielectric applications. When encapsulant is placed over a component, there is a 3 dimensional structure with areas that receive lower incident angle light and regions coated thicker than a normal dielectric. Both of these factors imply that encapsulant curing takes longer than dielectric curing.

The short answer is no. There are switches that have been in the field for more than a decade with no incidence of failure because of uncured component encapsulant. The purpose of an encapsulant is to mechanically secure the LED to the substrate so that the conductive epoxy joint has less chance of being stressed and broken, and this mechanical anchoring is accomplished by the encapsulant that surrounds the LED.

Some manufacturers put a dot of clear adhesive in the center of an LED to anchor the LED more securely to the substrate and to prevent encapsulant flow under the component.

It is possible to make a UV encapsulant that is rubbery but rubber tends to have the wrong material properties to perform as an encapsulant. A good encapsulant should work to hold the component on the surface and distribute flexural or thermally induced stresses over the entire surface of the substrate and component. A rubber in this situation tends to elongate and this would transfer all of that stress to the epoxy joint of the LED at the silver ink surface leading to premature failure.

Conversely, if an encapsulant is too rigid it will cause the entire encapsulant to break off and bring the component with it.

For these reasons, a good encapsulant lies somewhere between rigid/glassy and flexible/rubbery. One should be able to see a slight indentation in an encapsulant after being pressed with a thumbnail.

This is by far the biggest source of failure on membrane switches. The most frustrating thing about these “latent” failures is that it is virtually impossible to determine what actually caused the failure.

We have seen instances where we have investigated a single LED failure on a finished switch with a customer, and we could not cause failures on any of the other LEDs by mechanically stressing them. In one notable instance, we took the finished switch, stripped the graphic layer off carefully, and then pulled the circuit over the sharp corner of a conference room table and could not get the working LEDs next to the failed one to stop working. Even bending the circuit to about a ¾” radius in the area of the other LEDs did not produce a failure.

Nicomatic, LP (Warminster, PA) reports that probably 90% of LED failures on completed circuits are due to incorrect placement of LED components.

In the case of a latent failure, the effect of a poor surface mount joint or incorrect component placement is that the LED may work perfectly well during final testing. When it is shipped out to the end customer, the joint is stressed, causing a microfracture in the conductive adhesive or at one of the interfaces with the component or substrate, and the LED stops working. This stress can be caused by thermal expansion during shipping or by normal flexing during handling and assembly of the membrane switch into the final component.

The causes of this type of failure are numerous, but can be summed up as:

1. Improperly mixed conductive epoxy adhesive
2. Too little conductive epoxy adhesive used
3. Epoxy adhesive not cured completely, or
4. Component not placed correctly (offset from conductive adhesive and ink trace, not applied with enough pressure to seat it into adhesive, or canted to one side so one end seats well, but the other end barely makes contact with the conductive adhesive)

Of these four, the last two are the largest contributors to LED latent failures. If the epoxy adhesive joint is not cured completely, it will not have optimized mechanical strength. If the joint is stressed, a small crack can appear that will push the silver particles apart far enough so that a high resistance junction is created.

If the component is not placed correctly, mechanical stress from handling the circuit can cause a small fracture between the component and the epoxy.

In either case, the LED will show as a failure, but in most cases slight pressure can be applied to the component and it will start working. However, after a while, the small crack will slightly open once more and the LED will stop working.

The best way to minimize the chance for a latent failure to occur is to consider the following recommendations when surface mounting components on membrane switches:

1. Use a two part adhesive to ensure that the adhesive will continue to cure completely at room temperature after the heat curing cycle.
2. Do not use epoxy adhesives that have high levels of solvent content.
3. Use a center dot of non-conductive adhesive along with the silver conductive adhesive to mount the component for added mechanical strength.
4. Use a UV encapsulant to secure the component to the substrate.
5. Cure the conductive adhesive completely using the supplier’s recommended guidelines.
6. Follow recommended guidelines for component placement to ensure optimal contact is made between the component, conductive adhesive, substrate and ink trace.