Issue link: http://iconnect007.uberflip.com/i/1108006

Contents of this Issue


Page 35 of 83

36 FLEX007 MAGAZINE I APRIL 2019 necting the rigidized islands without changing materials (Figure 1). Armed with a solid solution for the flex board, the next challenge involved assembly and sensors. One thing almost all wearables have in common is the need for some form of sensor interface to the real world. The Graft- Worx device measures hemoglobin, SpO 2 , heart rate, temperature, and volumetric flow with a powerful fusion of 11 sensor data streams. Unfortunately, because of this interface, sen- sors typically have some form of vulnerability beyond standard ICs. Our MEMS accelerome- ter can be damaged by excessive shock from a pick-and-place machine. The microphone has a port open to atmosphere, which exposes the delicate internal diaphragm. It can be damaged by particles, fluid, and excessive percussive pressure changes. The optical sensors have transparent windows that can be damaged by aggressive mechanical handling. Solutions in a high-volume assembly pro- cess can be found in the application notes from the component manufacturers and have been developed from experience with other products, such as smartphones. These can include adjustments to a standard assembly process, such as tip material choice, pick- and-place parameters, use of a no-clean flux, and reflow profile. Because each type of sen- sor can have different sensitivities, all must be reviewed to determine if they can be accom- modated in a single pass with regular ICs, or if they are going to require an additional custom pass. Without a willingness to review what may be novel requirements for assembly and adjust and requalify a standard process, sensor components will be damaged, and yields will suffer. Once fully assembled, we had a challenge with our battery. Our wearable is a single-use device. To achieve an acceptable lifetime, we engineered a very efficient power management scheme with hardware and firmware that can achieve a quiescent current of 50 nA as well as respond to outside stimuli by waking up and performing tasks. The boards had to be tested and programmed with this power management firmware before solder attach of the primary cell, or the battery would start to power the circuit. In sufficiently high volume and with a mature code base, we could have had our firmware image pre-loaded into our microcon- troller, but for this stage, the lead time and cost were not supportable. Neither did we have a production worthy pogo or bed-of-nails fixture to test and program the board. Lenthor's flex- ibility enabled the GraftWorx engineers time to test and program the assembled boards in their facility with a prototype fixture and system just before the final battery attach. The resulting electronics assembly shown met all of the requirements wonderfully (Figure 2). The tethers were flexible enough that they Figure 1: Cross-section of layers in the customized PCB process for the GraftWorx wearable. The blank cores provided sufficient rigidity to meet reliability requirements while the thinner tether regions were extremely light and flexible, enabling maximum comfort for the patient.

Articles in this issue

Archives of this issue

view archives of FLEX007 - Flex007-Apr2019