Introduction

On March 25, 2020, Justin Huang, friend and founder of Solid Factory, reached out to me for my feedback on potential designs to help the University of California, Irvine develop personal protective equipment in response to the COVID-19 pandemic. I asked to be included in the correspondence, and soon found myself taking lead of COVID response research.

I immediately reached out to the community, and was able to get in contact with other engineers hoping to put their efforts towards the pandemic. I realized although there were many efforts, the vast majority of efforts were working in parallel, and not constructively on previous work. I was able to coordinate help from MIT, UCLA, USC and Helpful Engineers to exchange knowledge and resources.

Due to my dedication to the project, UCI allocated resources to me to expedite work on the project. I was given a team of 9 biomed engineering seniors to delegate workload to along with funding and lab access to develop novel personal protective equipment.

The Plan

I soon set out an initial plan on taking the problem:

OBJECTIVE: Develop a reusable respirator solution for the current COVID-19 pandemic.
1. Filter media shall have a viral filtration efficiency similarly to commercially available N95 masks as defined in 42 CFR Part 84.
1.1 Filter media shall be tested to validate it's viral filtration efficiency.
1.1.1 The team shall develop a means to test filter media
1.1.1.1 The team shall work with UCI to expedite the creation and operation of an appropriate testing apparatus
1.1.1.2 The team shall use MIT's existing facilities for testing
1.2 Potential novel filter media shall be identified
2. All reusable parts of the respirator shall be sterilized between use
2.1 The sterilization method selected shall be easily implemented (via an existing process or equipment).
2.2 Reusable parts shall demonstrate adequate chemical, moisture, and/or temperature resistance for it's selected sterilization method.
3. Respirator shall provide an adequately seal against the environment while remaining comfortable when used over the course of a day.
3.1 Respirator shall be minimize pressure and irritation of user's skin
3.1.1 Respirator parts in direct contact with the user shall be made of non-irritating materials
3.2. Respirator seal shall be constructed in compliance with 42 CFR Part 84.

Respirators and Masks

As of July 2020, interest in our masks have become urgent due to the ongoing surge in cases. In order to best comply with FDA regulations, our team of volunteer doctors and engineers will proceed with the following:

Formation of a Nonprofit Corporation to serve as the sponsor of the project, and more easily accept external funding
Implementation of risk management controls and compliance with ISO 13485
Increased automation of the manufacturing process of the masks

Unfortunately, FDA regulations prevent us from discussing specifics of the project’s design to the public. It is important here to reiterate that this is purely a good-will project; the entire team is volunteering their time and efforts, and funding.

PAPR

In addition to the unpowered masks, I am assisting with the design of a PAPR in a consulting role.

Design

Design of a successful PAPR relies on the balance of the following attributes:

Comfort - The PAPR must be comfortable enough to wear for 8+ hours, and thus, must be small and light enough to be widely adopted.
Power - Smaller, lighter PAPRs require more powerful fan and support components, which increase costs, and can paradoxically increase weight.
Noise - As per NIOSH requirements, and out practicality, the PAPR must be located and design in a way that allows for human speech.
Cost - A commercial PAPR solution will easily cost over $600, which is financially unfeasible on a large scale.
Filtration - Due to the large volume of air flowing through the system, filtration must be >99.97% efficient.

It is important for a team to realize all these components work as a balance of these tradeoffs; although measures can be taken to lessen the impact of some component:

Noise can be suppressed with certain controls. First, let’s refer to the Equal-loudness contour, also known as the Fletcher–Munson curves:

It is unsurprising, then, that human beings find high pitched noises particularly annoying.

1) PWM Motor noise/whine can be reduced by decreasing PWM frequency (thus shifting noise to a less perceptible level) or by increasing PWM frequency past the range of human hearing. This frequency is different for each fan, and has to be calibrated for each fan. Additionally, care must be taken to ensure your components are capable of handling higher PWM frequencies - not all combinations microcontrollers or MOSFETS are capable of switching fast enough; and changes may have to be made to made to the microcontoller timers, and/or through the use of MOSFET drivers.

2) Airflow noise can be reduced by absorbing vibration, or by controlling the inlet/outlet of the device. Careful design, including the use of diffusers, can shift the frequency of noise to a different level; however, expansion itself can induce noise!

Cost/Power - A clever way of supplying large amounts of mobile energy for cheap, is to simply use common power banks. Power banks are ubiquitous, and require little additional support infrastructure, training, or cost to use. However, looking at the fan power draw, and total current supplied by the power bank, it’d be quickly evident that the standard USB Battery Charging Specification would not supply enough power. Fortunately, Qualcomm’s Quickcharge technology is now common on most modern power banks, and are capable of supplying between 18W-100W of power at 12V (based on the model), and the handshake required to trigger different voltages from these power bank has been reverse engineered.

A simple voltage divider network is all that’s needed!

Hardware and Software

Using my experience from previous projects, I was able to put together code and hardware for a highly adaptable PAPR using only common parts that would not be subject to supply chain shortages. This design is intended to be highly flexible and modular. To allow a technician or engineer to quickly adapt a PAPR to a new configuration of fan, power source, or filter with only a few simple settings. Additionally, it relies on an auto-tuning PID feedback to maintain constant flow, allowing it to adapt rather quickly.

Work on the hardware and software can be found on my github. I encourage you to to check out the nightly branch, which represents the bleeding edge of the software.

News and Media Coverage

Coverage on our work was published by The LA Times!

Recyclablu: WHO Recognized Research