Using Flexible PC Boards in Consumer Electronics

In traveling around to visit different consumer electronics developers I’ve seen a new issue that affects sensor providers working with the newest cellphone and similar designs. It’s this: Can the sensor operate from the main PCB instead of from the typical flex PCB (FPCB) that brings the sensor closer to the glass? Potentially, a tough call. Look at the typical implementation for a sensor solution, on a FPCB where the infrared (IR) LED and the sensor, whether an ambient light sensor or proximity sensor, are placed on the flexible strip, or FPCB. The sensor is located at a gap of 0mm to 0.2mm from the bottom of the cover glass. See Figure 1. The FPCB and its connector to the main PCB cost on the order of 20 to 30 cents.

Figure 1

With the emitter and detector close to the glass, there is no dead zone, crosstalk is minimal, and the field of view is wide.

With the emitter and detector close to the glass, there is no dead zone, crosstalk is minimal, and the field of view is wide.

After seeing the kinds of lengthy negotiations between suppliers and customers that go on over saving a penny or two, you can imagine how tempting it can be to want to remove the FPCB connection. If the IR LED and sensor are placed on the main PCB, the gap (or distance from the bottom of the cover glass) could be as little as 0.5mm to typically 5.0mm. In a few cases, it's even greater than 5mm. The greater the distance, the bigger the challenge to the operation of the sensor. See Figure 2.

Figure 2

When the emitter and detector are set farther back, a blind zone shows up, crosstalk increases, and the field of view narrows.

When the emitter and detector are set farther back, a blind zone shows up, crosstalk increases, and the field of view narrows.

One of the figures of merit of an ambient light sensor or a proximity sensor is its field of view, shown in both figures. Notice how the distance behind the cover glass in the case where the sensor is on the main PCB reduces the field of view. The greater that distance, the more that field of view is reduced. In the case of ambient light sensing, the light has to enter the aperture above the sensor through that very limited field of view. (The system alternative is to add a diffuser, but this adds cost and is commonly not included.) In the case of proximity sensing, objects will only be sensed if they are within that same limited field of view.

This may or may not be desirable depending on the application. For a cellphone, the object to detect is usually a human ear or cheek to signify that the user is bringing the phone up to their head. In this specific situation, a smaller field of view may suffice. However, if the proximity sensor is used to sense gestures or interactions with a hand or finger, then the field of view must be much wider. There is no international standard for field of view, so each smartphone maker has its own spec. We understand that an acceptable field of view tends to be ±30 degrees from center, or a total of 60 degrees. The limited field of view is one tradeoff from moving the sensor from the FPCB to the PCB. What field of view would you like on your phone? Would you like the screen to dim or brighten only when the light level changes behind you (narrower field of view)? Or maybe when the light level changes anywhere in the room (wider field of view)?

Here is another issue: The distance created by increasing the gap also causes an increase in internal reflection (refer again to the figures, above). So, system designers can add a barrier between the IR LED and the sensor to minimize crosstalk. Sometimes barriers match the distance between the cover glass and the PCB. While this is mechanically sound, it risks creating a blind zone, a detrimental condition where an object can be close to the sensor, but no longer detected.

The IR LED has an area over which it spreads its energy and the sensor has a three-dimensional area over which it can receive signals. These areas are represented by the angles projecting upward in the figures. The blind zone can be fixed by lowering this same barrier to the height that allows the angles to intersect within the cover glass. There may be an added cost though, and knowing how tight component price and cost are to the consumer electronics providers — well, it can be tricky, right?

8 comments on “Using Flexible PC Boards in Consumer Electronics

  1. Michael Dunn
    March 7, 2013

    I suppose if you're a large enough user, you could get taller components made! (within reason)

    Of course, in olden times, components had things called leads, and we could place them as far above the PCB as we liked.

  2. eafpres
    March 8, 2013

    Hi Tamara–interesting problem.  I suppose the final decision includes many variables, including can you use that FPC to support any other components, signals, power, etc. that might spread the cost over more of the system, and possibly reduce complexity somewhere else?

    I wonder if you have seen anyone using a daughter rigid PCB in a stack to raise the parts?  This could be relatively cheap to do–design a small rectangular PCB with the pads for the sensor on both sides, large vias to pass power.  Assembly would be a challenge, however, as either the daughter or the sensor would have to go on in a 2nd operation.  

  3. TheMeasurementBlues
    March 8, 2013

    @Tamara, can the aperture be larger? I'm guessing not. If that case, you need to raise the LED and senros or lower the aperture. Basic physics at work here. But then, when did management ever believe that the laws cost couldn't overcome the laws of physics?

  4. mhanuel
    March 13, 2013

    I think it;s possible to include the IR sensor on the lens sensor, that is, the touch screen sensor. I haven't seen a dedicated FPCB for the IR sensor yet.

    Very interesting article.

  5. TamaraISchmitz
    March 14, 2013

    Currently, the IR sensors are too large and would cause black dots on the screen.  This would be unwanted.  With technology scaling, someday this might be possible.

  6. TamaraISchmitz
    March 14, 2013

    Manufacturers guard their aperture sizes like my brother guarded his halloween candy.  The aperture (and the candy stash) can get smaller, but not bigger.

    As for management understanding, I sing the durge along with you!

  7. TamaraISchmitz
    March 14, 2013

    Yes, I've seen that!  It was a pretty nifty design.  The issue was solid connection through the spring connections from the daughter card to the main board.

  8. TamaraISchmitz
    March 14, 2013

    This is a brilliant suggestion!  I am surprised that manufacturers haven't asked for this.  My suspicion is that the tooling (handlers) for assembly wouldn't be cheap to adapt to the nonstandard heights.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.