DIY Powered Air-purifying Respirator

Due to SARS-CoV-2 pandemic in country where I live it is now mandatory to cover nose and mouth in public places. This inspired me to create helmet-like device that uses activated charcoal filters intended for use in ductless (or recirculating) exhaust hood. My PAPR strongly resembles computer case, with its two 120mm fans (one in front, one in the back) and a boxy appearance. This device is highly experimental and not-tested in any laboratory setting. To build it I used only things that I could find around the house, so this design could potentially be greatly improved.

Supplies

Parts and materials:

• 2x powerful 120mm computer fans (preferably two of the same model and in the same state, I ended up using more powerful fan for the exhaust [this configuration creates negative pressure])
• 8x screw for the fan
• 3-cell 11.1V LiPo battery
• 4-pin 2.54mm male pin header for the connection of battery’s balance charging socket
• 2x 3-pin 2.54mm male pin header for the fans (or equivalent)
• simple toggle switch that can handle 1A of current
• few small pieces of prefboard
• some wires
• cardboard (corrugated fiberboard)
• large sheets of thick paper (I used A2 size)
• variety of pressure-sensitive tapes (duct tape, packing tape, insulating tape, scotch tape, double-sided tape, aluminum tape)
• piece of thick clear plastic foil (13.5x22cm)
• glue (something that can hold paper together)
• hook and loop tape (self adhesive or you will need some strong glue to hold it in place)
• elastic band
• activated charcoal filters
• paper towels

Tools:

• phillips screwdriver
• flat-bladed screwdriver
• utility knife
• scissors
• ruler
• set square
• pencil
• compass (drawing tool)
• sewing needle and thread
• multimeter
• soldering station
• solder
• LiPo battery charger

Humidity and temperature control system:

• Intel box cooling
• M5 bolts, nuts and washers
• tap wrench + taps
• copper wires + fan grill
• aluminum tape

Step 1: Preparation of the Fans

Each fan has to be attached to a paper structure of the respirator. To make connection between structure and fan relatively stiff and to keep screws of the fan inside the structure I built a mount that consists of three layers of thick paper. Each of the layers is a square with a 14cm side (mount extends for 1cm from each side of the fan).

There are four circular holes inside each of the layers intended for the screws. Those holes have a diameter of 5mm in the layer closest to the fan (this layer is held by the screws), holes in other two layers have 7mm (those layers only mask the screws). Centers of those holes are located on a vertices of a square with 10.5cm edge length.

There is also a rectangular opening (4x9cm) through which air will flow. I wanted to use small carbon filters and I had only 120mm fans available. This restricts airflow of the fans and wastes power.

Those layers were cut with the use of utility knife (put something bellow the layers, so that you don’t damage your table), then glued together and screwed to the fan.

On the other side of the fan I installed another similar layer of paper, but only 12cm wide and without circular holes. It was held by the tape sealing all the edges. Edges on the other side were sealed by the tape as well.

Mind that one of the fans will be responsible for an air intake, the other one for exhaust. So they will face their mounts from different directions.

NOTE: In the end those fans were not powerful enough, so I replaced exhaust one. Procedure of replacing the fan and more details in Step 8.

Step 2: Building the Paper Structure

You need to cut shapes found in structure.pdf, fold them and then glue them together. Note that folds which are 1.4cm wide on drawing, in pictures are 2cm wide. It is because my paperboard was a little larger than regular A2 size. Then put some tape around the edges, both inside and outside, to let glue harden and to make sure that no gas can come though the edges.

Later glue fan mounts to the front and the back, and cover their edges with tape.

I lined top of the helmet with cardboard. I used rectangular 60x27cm piece. Note that flutes are parallel to the longer edge. Cardboard was folded so that it could fit inside PAPR and on the sides approximated shape of the human head. Cardboard was then attached to the rest of the helmet by one piece of tape in the front (so that it can be later raised a little). Fronts of the folds were trimmed so that they did not obstruct field of vision anymore.

Two more rectangular pieces of cardboard (36x12cm) were used as beams in the lower section of the PAPR. After being folded into roughly rhombic cross-sections they were attached to the structure by the airtight layers of tape.

Another cardboard piece (12x6cm) will hold battery with self adhesive hook tape, so that loop tape was glued (and taped with another tape) to it. This piece was placed on top-back edge of the structure and attached with tape. Small part of it is located between structure and 60x27cm cardboard piece, so when they both are in place, the larger one should be fastened on the back side as well.

Step 3: Wiring and More Cardboard

At this point I soldered wires, connectors and switch together. This is so simple system that I won’t go into details. Connections were later modified to allow for connection of the fan on the outside (switch was placed outside as well then).

Two pieces of cardboard (13x3cm) were taped to the exhaust fan, as well as small piece of paper towel, so that hair won’t be drawn into the fan. When fan was replaced (Step 8) configuration of those pieces was changed a little.

Another piece of cardboard (20x11cm) filled the gap in bottom-front part of the PAPR. On both sides it was attached with densely packed tape.

NOTE: To this cardboard piece I later connected humidity and temperature control system (see Step 9). If want to install it as well, you probably should do it before attaching the 20x11cm piece.

With all those cardboard pieces in place, I covered bottom edge with yet more tape (while leaving some exposed paperboard around air inlet/outlet).

Step 4: Collar

I considered the name skirt, as this is the part of a hovercraft that blows air outside the craft. I used around half (52x30cm, when lying flat on the ground, 2 layers of plastic in total, lying one atop of the other) of a waste disposal bag. When extended, it forms a side of a stubby cylinder. Also, it was a scented rubbish bag, so I can tell people that I’m wearing a perfumed scarf.

Part of the bag that I used had a string inside it, so I taped corresponding openings. Then I started attaching bag to the opening in the helmet. First inside, back to front, then another layer of a seal on the outside.

Circumference of the bag was greater than that of the opening, so in front there is a fold in the bag, held in shape by a tape, so now circumference on the bottom of the collar is greater than on the top.

Collar is supposed to be tucked into the shirt. To improve sealing around the neck I constructed kind of a elastic belt, made from a thin elastic band and sewn to it longer piece of hooks tape and shorter of a loops tape. Be careful not to strangle yourself.

Later, I increased amount of tape in the corners of the opening in the base of the helmet, so that area of the opening is reduced.

Step 5: Window

I used thick, clear plastic for an eyepiece, which was a part of cotton bed sheet packing. Window is a rectangle with 22x13.5cm dimensions. Before installing it, I applied a long piece of a clear tape to the edge of the window opening, on the inside. Then I attached window to it at kept it on the outside. After that, I lined rest of the opening’s inside with clear tape and fully placed window in its position. Later, another layer of tape covered the window’s edges on the outside.

NOTE: When opening is empty, it allows for an easy access to the interior of a PAPR. You may want to install window later. You also should reinforce paperboard structure around the windows with corrugated fiberboard. I used double-sided tape to attach reinforcements made out of 11x6cm pieces left over form holder’s frames that will be presented in the next Step.

Step 6: Activated Charcoal Filter Holders

From corrugated fiberboard I’ve cut 6 rectangular frames (13x8cm) with a rectangular openings inside (11x6cm) into which filters will later be placed. Each of the holders (one for intake, one for exhaust) were constructed out of 3 cardboard frames glued together. Then each of the holders was covered with a tape, from all sides.

Later, I glued holder to air intake, and started to place tape around it, so that seal is better and holder is supported when glue stays in liquid form. Same thing was made for the exhaust.

When holders were in place, I covered all exposed paper on the outside of the helmet with tape, to increase resistance to tear and wear, allow for a easier disinfection, and of course to make whole construction more airtight.

Charcoal filters were cut from a larger filter intended for a ductless exhaust hood. They should fit tightly inside holders and during cutting you should be extremely neat, clean and hygienic. When inside the holders, filters are secured in place by cheap, clear tape, completely covering a filter-holder border.

Note that you may want to install some other filter behind intake activated charcoal one, to prevent inhaling small chunks of it. I did not do it and I didn’t had any problems (yet).

Step 7: Paper Towel Prefilter

I’ve also made additional holder for a prefilter made out of a paper towel, because placing it into small housing of charcoal filter restricted airflow too much. It is made out of two parts, each of them consisting of 3 pieces of cardboard glued together. Cardboard pieces are in the shape of a circular segment, its chord length is 22cm (the same width as PAPR) and arc length is 25cm (width of the paper towels that I use).

Calculating radius of the circle, so that compass can be used to draw it is quite an interesting problem as formula for radius of the circle that I produced is:

R = half_of_arc_length / arcsin( half_of_chord_length / R ) 

and it has to be solved by iteration (LibreOfffce Calc has problems with this equation for some reason). I produced equation only after calculating radius with slightly different method in the spreadsheet. Both of those methods can be found inside circular_segment.zip

In the end, radius is 14.45cm (roughly) and height of the circular segment (sagitta, distance from the center of the arc to the center of its base) is 5.075cm. I first drawn one copy of segment’s shape on paperboard, then cut it out and used it as a template for making more copies of it from thick cardboard.

Pieces of cardboard were then glued together and attached to the the front part of the PAPR. Make sure that tape covers whole holders. When this was done I placed long pieces of tape where paper towel will be placed, as to make surface more smooth. I also placed smaller pieces of cardboard covered in tape between main circular pieces, so they won’t move too much.

Paper towel prefilters should have dimensions of 25x13.5cm and be attached to the holder by clear tape covering all sides of the prefilter. Make sure that you got this right, as towels easily slip to the inside of the holder.

Step 8: Exhaust Fan Replacement

Fan that I mounted in the back turned out to be much to weak. I replaced it with something that draws more power and is much nosier. New fan had a grill that I retained, so 12x12cm piece of paperboard was no longer required on the intake side of this new fan.

To remove previous fan, I stripped tape seal around it and then pried and pulled it and its screw out of the paper mount (which remained in its place).

New fan had LEDs, wires of which were desoldered from fan’s PCB in order to save power and prevent fan from producing any weird blue auras around my head. This fan also had weird connections. One of them (2 pins) was for 12V power, and it fitted just as well to the PAPR connector made from standard 2.54mm male pin headers. Tachometer had a separate 3-pin connector, was not used, and along with LED wires it was wrapped around sides of the fan and covered with tape.

Then I prepared new layer of mount, which was now only 12x12cm, had rounded edges and small holes (5mm diameter) for the screws. After that, screws were placed in their positions and fastened, and this new layer was glued to the mount. Then, I replaced the tape seal around the fan. When fan was installed and connected, I noticed that due to the small area of opening in the mount it blows quite a lot of air back inside PAPR. To remedy this I covered rim of the grill with two layers of tape, creating some radial flow inside this axial fan. This also reduces chances of hair entering the fan.

The two 13x3cm cardboard pieces were taped to the new fan in such a way that they formed an arch above opening in the grill, allowing air to pass while reducing chance of hair coming in. I discarded exhaust paper towel as it caused too much air resistance.

Step 9: Humidity and Temperature Control System

I used cooler (heatsink+fan) that was shipped with Intel Pentium Dual-Core processor (1.8GHz, 65W, LGA 775). It is a fully aluminum construction, with a DELTA fan rated for 12VDC 0.60A . Diameter of the fan is close to 77mm.

Ten years ago I drilled a hole in a base of heatsink and made a M3 thread inside it. I most likely used 2.5mm drill bit, and then I removed some material with the use of greased Screw-tap set 3 pc. This hole will now to be used to attach a condenser made out of 120mm fan grill and 5mm diameter copper wire. If I were doing the hole now (and had M5 tap and 4.20mm drill bit) I would made it a little larger.

One of the grills “screw holes” was angled by 90 degrees, screw will come only through this “screw hole” and it will connect condenser with the heatsink. 5mm diameter copper wire was angled and twisted so that it would made one loop around grill’s length and would touch the heatsink's base in two places at least, while holding quite firmly the grill. Additional 0.5mm diameter copper wires secured larger wire in its place.

Cooler was to be mounted to the underside of the bottom-front cardboard piece (20x11cm), but not by plastic “clips” which were removed, but by M5 screws (around 20mm total length), nuts and washers. Five circular holes in the cardboard were made to accommodate installation of the cooler. Central one has a radius of 2cm and will expose base of heatsink to the inside of the helmet. 4 other holes are on vertices of a square with a side length of 7cm. Their diameter is 5mm and M5 screws will come through them and fasten cooler to the PAPR. Centers of those 4 holes are 4.95 cm from center of the large hole, which should be located in the middle of the cardboard, but a little towards the collar.

When I marked this main center with compass’s needle point, I made two circles one with 2cm radius, another with 4.95cm radius (on the photos you can see two sets of circles, one of them is drawn mistakenly). I placed heatsink on those circles an marked where exactly mounting screws will be with pencil. Then I started making those 4 holes, at first with a needle point, then with small wood screw, and then I dragged 5mm diameter copper wire through them. Largest hole was cut with utility knife.

Then I could try fitting condenser to the large hole. Also, before mounting heatsink I wanted to cover edges of its base with aluminum tape. Before doing that I cleared less accessible side of heatsink base with aluminum tape similar in its trapezoid shape to the pieces of tape that I will be sticking to the heatsink in a short while.

I installed some pieces of aluminum tape between heatsink base, its fins and mounting holes before fastening heatsink with screws. It was done to airtight this mounting (you may try using some epoxy to make it even more reliable). Also, before fastening heatsink to cardboard, I mounted condenser with M3 screw and washer, but it may not be necessary to do it in this order.

Cooler was attached to the cardboard with M5 screws (bolts), nuts and washers (2 per screw). Aluminum tape was pushed against the cardboard, as to make it stick. Less rigid tape was used to reinforce aluminum tape and cover M5 bolts heads. I used densely-packed aluminum foil to fill 2cm radius hole on the inside of the helmet. Also, two additional copper wires (2mm diameter) were connected to the pairs of bolts on the inside of the the helmet, and held in place by M5 nuts (no washers here).

Grill was additionally secured in its place by placing some amount of clear tape between its two side “screw holes” and walls of the helmet.

I also made a tiny slit next to the cooler through which 3-wired ribbon cable can pass. This cable is connected to the prefboard on which 4-pin header for the outside fan connector and switch is located. Switch on the outside allows to temporarily disable fans and reduce noise during verbal communications. Cable slit, was tightly covered with tape, as well as all electronics on the outside. Also remember to cover electronics inside with insulation tape (so that connector stay connected), and tape them to the walls, so that they don’t wiggle freely everywhere.

To the cooler small piece of cardboard was attached, secured with two long strips of narrow tape wrapping many times around heatsink, one in the bottom of the fins, another on the top. Cardboard was facing collar and was there to prevent accidental damage to this plastic foil, as this Intel cooler has its fan quite exposed. You should be careful (although being hit by the exhaust fan’s blades is more painful and damages skin more) .

Step 10: Side Notes

PAPR provides air supply that allows 93kg (BMI 23.7) male to walk fairly comfortably few kilometers. I do not recommend running in it or performing other highly intensive activities, as air supply is not that abundant.

Most problematic issue is condensation of water on the window which happens when using helmet in cold environment. Humidity and temperature control system reduces problem slightly and allows for walking longer distances without vision being completely blocked. I suspect that better heat exchanger on the inside (or even building dehumidifier with a heat pump and bucket for a condensate) or changing window material would help with this issue immensely.

Field of vision is large enough that 27” screen can be viewed fully with both eyes when those eyes are approximately 50cm from the center of the screen (I usually view this screen from 70cm distance).

PAPR in a configuration that I presented has negative-pressure inside, making it more likely to the contaminants to enter through small imperfections in its sealing layers of tape. But it also should prevent contaminants inside from escaping, so in a way it a safe configuration for epidemics.

Best technique for breathing that I found is inhaling with mouth and exhaling with nose. To improve airflow at some point I had installed kind of a straw that was supposed to bring air intake opening closer to the mouth and use additional pressure from lungs to draw air more efficiently, but I later removed it to make place for the condenser.

Two 120mm fans on the inside combined consume 0.33A during normal operation, outside 77mm fan consumes 0.08A .