Monday, July 29, 2013

DIY positive-pressure workstation for minimizing dust and other particulate

The design and construction of a positive-pressure workstation. 

This is the completed project, less nylon filters over
the intake fans. The dimensions are roughly 16" x 16" x 30"

As those experienced with repairing or modifying tech hardware know, different platforms often mean requiring different tools to do a job effectively and effectively. Having replaced my share of tablet and smartphone displays, I'm no stranger to the frustration that dust can cause when the display and digitizer or front glass are separate pieces. Under normal circumstances (e.g., only having to occasionally take on such a repair), the process of simply keeping the dust-sensitive parts covered until as close to final assembly as possibly, then using some combination of microfiber cloth and compressed air will suffice. Though even then, I've been faced with having to remove a newly replaced digitizer to access and remove dust that somehow slipped by. 

Knowing that my tablet (iPad) repair workload is about to explode, I asked myself, “how can I avoid this problem?” The answer I came up with is to utilize some sort of positive-pressure workstation (PPWS), where filtered air would be pushed into an enclosed work area thereby pushing any dust or other particulate in the air out through the an opening, thereby creating a dust-free, or at least a near-dust free area to work. A bit of research yielded disappointing results (that is, after “calling for a quote” I learned the units I was interested in were running from $1600 - $2300. Granted, these are professionally constructed commercial-quality units for culturing in labs, or hard drive platter data recovery, etc – simply beyond my requirement to eliminate visible dust/particulate in my work area. I figured I could design and build a serviceable unit for a fraction of the price. 

I'd like to share the rough design process with you here, in hopes of it helping someone that doesn't want to spend thousands of dollars for something that (or at least a rough equivalent of) can be put together for well under $100US.

The concept was simple: build a frame for the enclosure; seal off enclosure, less for controlled air flow; be able to see what's inside the enclosure. With this in mind, I made a few sketches looking at various PVC fittings and clear acrylic on Home Depot's website. 

These were some of the original sketches I did (that I just
 pulled out of the recycling bin to take a picture of). Note
that the two on the lower piece of paper assumed I could
find a PVC adapter that allowed for three pieces of pipe
to meet at one corner. The 'design' above worked with
what I had access to, the "tee" and "elbow" fittings.

The shopping list I took to the hardware store looked something like this:
  • ~30 feet of half-inch PVC
  • PVC "tee fittings," and "elbow fittings"
  • PVC cement
  • Machine screws, 1.5 inch, washers, and nuts
  • Duct tape
  • Weather stripping
  • Acrylic sheet (30" x 36", if I recall)
  • "Twinwall" plastic sheeting - I used two 24" x 36"
I already had all the tools necessary for the project. These included a drill with various sized bits, a rotatory tool with cutting wheels, a heat gun, a manual miter saw (probably not necessary), a utility knife, sandpaper, pliers and screwdriver, zip ties, thread locker, and the necessary soldering supplies and electronics (which I will discuss later).

Should you want to build your own PPWS, I say ignore my shopping list for the time being, as after I feel as though I would make significant changes that only came to mind after the build. I will describe the my process, then describe what I would have done differently were I to start from scratch knowing what I do now. 

Working from my sloppy design above, I started to do a (dry) assembly of the pieces necessary. Measuring off and marking the length desired (accounting for the fitting on either end) I used a miter saw for relatively straight cuts, but again, I'm sure there's better ways to cut PVC. 

Note the 'plan' on the sheet of paper to the left. 

I hadn't ever worked with PVC as a structural element before, so when everything came together, more or less as planned, I felt like I was on to something. 

There was a bit of slop in some of spans. I attribute this to the fact that I "measured once, cut once," as opposed to what my father, as I assume many others have shared, "measure twice, cut once." After trimming or replacing the offending lengths, the frame was broken down into smaller parts and PVC cement was applied to make the bond permanent. This was the first time I'd used PVC cement, and you must work quickly to line things up before it sets - I'm not sure if there's a "slow setting" alternative, but for a little more breathing room, I would have opted for such. 

I can't remember what size of bit this was, but needless
to say, it was the right size for the machine screws.

At this point, I was ready to eyeball where the plastic sides and backing would be anchored to the frame. I drilled several holes on the frame for the sides and back. Then applied weather stripping to ensure that I had as close to an airtight seal as possible.

I quickly found out that it was also necessary to pre-drill
plastic sheeting and pop a hole through the weather stripping 
as well in order for everything to go together smoothly.

Using a washer (in my case, it was a finishing washer, as the hardware store only had one package of standard washers, a screw and a bolt, the first pieces of plastic sheeting was secured in place. In order to avoid a crease in the corner, a heat gun was used to to allow for a more graceful bend around to the back of the frame, and additional hardware was used to keep the sheeting secured. The same process was repeated on the other side of the frame. The top of the sheeting was trimmed just above flush to the PVC frame.

Looking down into the frame, seeing the left side and back. Note 
that you can see the two pieces of plastic sheeting overlapping.

Next was the last major structural part of the build - the acrylic sheet. I hadn't ever attempted to use acrylic in anyway that required bends to be put into it, though I knew that it could be done.

I knew that I wanted the clear acrylic to come down 
rouly 8" in front, which would leave an 8" gap under
the bottom of the window for my arms, tools, and hardware
to enter, and as importantly, for air to come out of.

Actually creating the bend was a bit more work than I had anticipated, or at least took more physical force than I would have guessed. I setup the acrylic between two boards as shown in the image below, with the mark just beyond the edge of the bottom board, and the back board recessed slightly - in place of a sawhorse and clamps I used a table and free weighs to hold the setup in place.

After several slow passes with the heatgun on low I was becoming frustrated that the plastic did not want to bend very much at all. I took a break and looked around online to see if I was missing something. It nothing jumped out at me in terms of why this shouldn't be working. After a few more, tries using a similar method, I decided it was time to become more persuasive. With the heatgun on high, making several close passes, then focusing back and forth on one end, I used a third board to apply a firm force across bottom end of the acrylic sheet. Sure enough, if started to give. I repeated the process on the other side and across the length until I had what looked to be an even 90(ish) degree curve, not releasing the lower board until I was confident the sheet had cooled long enough to hold its shape. I was somewhat concerned that I may have discolored the sheet due to the amount (and proximity) of heat, however, it was only the protective plastic that was to blame.

Peeling back the plastic - things are really taking shape.

As on the sides, I applied a layer of weather stripping along the top left, back, and right edges of the frame and put the acrylic in place. It was only at this point that I came to realize how fragile. The bit I used to drill holes for the screws was apparently far too large to use on this piece of acrylic without cracking it. Lucky, I was able to make a score mark on the other side of the most affected area and snap it off without issue. Experimenting with scrap I became confident in using a very small bit to drill a pilot hole, then using a cone grinding stone on a Dremel/rotary tool to open them up to the appropriate size - I'm sure there's a better way of doing this, I just don't know what it is - the method was effective and I was anxious. A layer of (white) duct tape was made for an airtight seal between acrylic and the sides/back.

The grinding stone used to widen the pilot holes in the acrylic.
My assumption is that the friction melted the plastic and the
centrifugal force threw threw it outward as it cooled off.

The fans were placed at equal intervals across the rear/top of the workstation. With the protective plastic still in place, they were traced using sharpie and their four mounting holes marked. Removing the fan, the the corners were marked off to create right triangles as shown in my super-awesome diagram below.
The octagonal shape was cut using a dremal/rotary tool
with a cutting disk, and the dots indicate where screw 
holes were made using the method indicated above.

Next came mounting the fans. To ensure a good seal, I again used weather stripping along the opening around the acrylic in hopes of obtaining a good seal and prevent possible vibrations from the fans resonating through the rest of the workstation. 

A view looking up from inside the workstation.
Fans are mounted and wires zip-tied into place.

As far as wiring the the fans go, I took a couple things into consideration. The first of which being that they're standard 12 volt, 90mm PC fans; next, I wanted to be able to drive them easily from an external power supply. Due to resources readily available in the shop I settled on using an laptop power adapter with an output of 19.5 volts and more than enough current to push ten times the number of fans I'm using. I also had access to compatible DC jacks to fit the power supply. The fans are wired in series drawing ~6.5v/each, they're certainly audible, but not any more noisy than my solder fume extractor or the like.

The DC jack fit into place, prior to securing with epoxy.

A power switch was also added, as seen just on the inside of the window
on the left side of the unit. The cables were temporarily secured with
masking tape while the epoxy bonded it to the PVC.

For filters, I'm currently using a single layer of nylon (cut from tights), which are secured over the fan's intake. Preliminary tests are promising, but I won't be putting the unit to its full potential use for a couple more months, where I expect to see (too) many iPads with broken glass coming into the shop. 

Generally, I'm happy with the outcome, however there are a few tweaks I may make depending upon need. The first of which is to find a more suitable filter material - I don't need anything HEPA grade, but am open to experimenting with anything that will both allow for good airflow and trap dust. The next most obvious thing I could try is upping the drive voltage on the fans, while they're designed to be run through a range, at 6.5, I'm just using half of their rated potential. Third, there's no reason I couldn't tear out the 90mm fans, do some additional cutting and install 120mm, or even 200mm fans. Though these are just notes for the future - it may be that the current setup does more than I need it to do.

If I were to do it again:

The first thing I'd have to admit, is that my design is probably over constructed and unnecessarily complex. While the PVC support frame is necessary, I'm not sure that the cross beams on the top/back were. With the addition of the plastic sheeting, and even the acrylic, additional rigidity is added, and I believe that such a design, while not as robust, would serve its purpose without much bowing or bending. 

Next, I'd lose the weather stripping in favor of a silicone adhesive/sealant - not only would this fill in any tiny gaps the weather stripping failed to, it would also reduce the need for as much hardware securing the the external walls and acrylic to the frame. I could see the use of such an adhesive easily cutting down the number of nuts/bolts/washers I used by a half.

Third, I would change the shape of the workstation. Naturally, where the acrylic bends, the image is distorted - this isn't really an issue for me as the height of my chair at the workbench I use this on is adjustable and I'm able to lower myself into a comfortable position where I'm able to see my work clearly. But I like the design I quickly mocked up below would make this a non-issue (that is, having the acrylic perpendicular to line-of-site, regardless of how tall the user is.

Just for clarity's sake, the red boxes represent the fans,
the area with the blue lines the acrylic, and the black
outline the structural PCV/plastic walls. 

That's about it for this time - the post turned out to be a lot longer than I had initially thought it would be... and I apologize from for my inconsistency between first and third person narrative. I realized I was doing that partway through the post but didn't feel like going back and correcting it.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

Saturday, July 6, 2013


I've successfully constructed a solar filter for my SCT and decided to have a go with it. Shooting the sun, obviously, will require experimenting with new techniques, but I got the concept down, at least.