anthony.a.douglasOk, first of all an important mention goes out to Edderic Ugaddan (twitter @eddericu ) who is currently making and selling a cool device called the Laminair. It's his research that showed me that this kind of device was valuable, because he has shown that you can get good protection factors from directing HEPA filtered air at your face. If you want to "just buy" something, there's that. I haven't tried one but I know he's a good guy.
There are "personal air cleaners", which supposedly use ionization to get particles to stick together and fall from the air. They don't work for jack.
NOTE: I think the H11 filter is probably a better idea, the H13 one gives kind of low flow even after adding the plastic rings to the fan. The H11 is usable even without adding the rings.
There are other "personal HEPA" filters, however other than the Laminaire, I have not seen any for sale or diy-able that actually direct clean air toward your face, which seems like an inherently more efficient approach for stronger protection than trying to clean the whole room with a small device.
I think this is probably because the size and weight is too great to get a good result, for comfortable neck mounted wear, without the conical annular jet.
But now... there is the conical annular jet. It causes air to flow at the edges of the Clean Air Zone at a higher speed/volume than the center. The stuff in the center does not need to move fast, and making it move fast takes a lot of clean air.
By reducing the total CFM (cubic feet per minute, of airflow, a CFM is 28.3 liters per minute) required to give a given level of stability and reliability to the Clean Air Zone, it helps. Lower CFM means the whole thing becomes cheaper easier, quieter and smaller. Less of everything, and less is more ;-). A large filter is also the way to go, it means a less powerful and quieter fan, smaller cheaper battery, no need for high powered electronics.
I estimate it chops the size of the unit, for the same degree of protection, but about a factor of 3 or so, which is enough to make a real practical difference, and if it had to be bigger you'd be stuck for a fan because this is the biggest easy to get suitable fan.
I estimate flow on max to be about 25 CFM, however you won't need to use it on max much. I crank it up when I'm walking around in the grocery store because you can't hear it anyway with the background noise and when you are moving around max flow is perhaps warranted.
I would be the first to say I do not think these are "the solution", however I think it's a legit tool in the box, like if you are on the bus or in a hospital or eating or something. I figure you should get fairly good protection most of the time, and you can also use it in addition to a mask if you want to hedge against leaks, although a papr might be a better idea at that point, this is a little easier and less intimidating.
Regarding exactly how much protection you get in various situations like near field and far field, it's pretty hard to know. One day I'd like to tape a PM2.5 sensor to my nose and another to the intake of the device and walk around during my day to see what particulate level reduction I get. I've also got it on my list of things I'd like to do to drill a hole in the base of the filter, insert a nicotine-free e-cigarette, and it should spew fog. I could set up a camera and walk around and see how reliably my face stays in the clean air zone. Obviously this will confuse some people greatly because foggy air represents clean air. I try to avoid those people :P.
This or better yet a refined device could also work for allergies. Clearly if it's going to be employed it should be improved to shrink it etc.
Not everyone can wear masks for various reasons, it can cause problems with your eyes, for instance.
It is possible to further shrink this device, I think by at least 50%, however that requires, primarily, designing and making a new fan of superior design. There is probably a smaller even more suitable filtre out there.
This one is the best filter I could get, I tried 4 different ones. This one, this on and this one, and the final 120 mm GDStime centrifugal fan, which is available on many retail sites.
UPDATE: actually, what happened was I ordered the filter for this vacuum, but it said it was H11. That was the V1.0. I gave that one to my Dad for xmas. I bought a 2 pack of filters for the same vacuum cleaner, but the listing says they do 99.97% filtration, which would be H13, not 95%, which is h11. Now, amazon listings hardly mean jack (and that's why I tried to use P100 filters, so you know what you are getting). However I noticed the pleating density is higher and the impediment to flow is also higher, with these 2 new filters. The media also looks different. So it's entirely possible that these are H13, and the other was H11. However obviously we have a reliability issue here - how do I know these filters aren't just trash? Well you could have that with any HEPA filter appliance, too, just look at what happened with the Levoit ones (not that that was a problem, it's different for a room type filter appliance, only CADR matters). Well. You can do a test with a PM2.5 sensor. I don't have one right now and it's not really a priority. I'll just carry on for now, and some day someone could test it. But then there is a sourcing problem. Well, aside from sticking with PM100 filers that are niosh approved, I don't really have any answers to that. You could combine the plastic ring modification with 6-8 P100 filters, perhaps. Trying any other product isn't going to solve this problem, every manufacturer makes stuff up these days. You could also possibly use a well tested filter like the Ikea Starkvind or something, those are probably a reasonable bet. If it really mattered, probably buying a sensor for $80 is a good idea. Has to be a reliable sensor though!
There are people like Nukit which are working on PAPR devices that use P100 filters. The fan module from that could be used perhaps with more than one P100 filter, that could possibly work, but the entrance and exit flow paths tend to be rather constricted. Basically we need a source of air, and the nozzle.
I have also investigated small portable HEPA filters, and there are some USB ones that are small enough you could probably just plonk a nozzle on one. However I try not to get into that stuff because then you buy this stuff and break it and it often doesn't work. It will trigger the clogged filter thing or something else, they may make too much noise or not really be HEPA or whatever then you are fuked. You might as well roll your own, basically.
I think a better approach for this whole subject of frankly consumer devices in general would be to deliver highly reliable components, not these unreliable and low quality complete machines. It would give much better results and with much less investment and lower per unit cost, for everyone involved. Then let other people mix them into good designs, with low capital input, test and verify everything. Right now the world of consumer manufacturing etc. is a real shit show, we have real problems and they really aren't getting solved.
You can dismantle the fan and glue in place small rings of plastic in the right areas to reduce air leakage past the blades, thus getting more airflow at max power and also more airflow with less noise because you can turn the fan down.
I included suitable sized rings and there is a pic to show what I mean, they leave about 0.5 mm of space between the case and blades. If they are not glued in place fully flat etc. the blades could hit the rings, and you will know because it makes noise, then you would have to take the rings off or sand them down which is a lot of work, so I would just remove them and print shorter ones or try again to glue them. The upper one, near the inlet hole, is thinner than the one below the blades. The gaps are different sizes. The pic does not show the lower one in place, it is missing in the pic.
If you use the sleeve bearing type fans this might not work as the rotor is usually capable of moving more, so it can hit the rings. You'd have to inspect one of the fans.
The rings are a hassle but they do work. To remove the tiny circlip I used two tiny Torx head screwdrivers. That works pretty well but the circlip gets bent, then I bend it back.
There is a spring between the two bearings, around the shaft. Don't lose it. It's also kind of hard to get the circlip back on. I eventually finagled it, basically you push the circlip down with two screwdrivers, this expands it and also pushes the bearing down at the same time. It might not be a bad idea to print a small object that makes this easy, it would simply be a cylinder with a hole slighly bigger than the shaft, in the middle and would only take a couple minutes to print.
You don't need to do the ring thing, it makes it better but it still works ok if you don't, you can also do it later, after gluing the filter on. You can even replace the fan without re-gluing the filter, you open the fan by putting a screwdriver into the little clips holding the upper case component on. I had to dremel away some stuff to be able to get a screwdriver in there.
It's pretty annoying they don't just do a better job on the fan design to begin with, probably the blade angles and densities etc. are nowhere near optimal, either.
Be careful not to break the little clips that hold the two pieces of the case together, I broke 2 of them but 2 is still enough if they are on opposite sides. You could also tape it together.
I investigated several different filters: multiple in parallel (4) of those pink and also the grey P100 from 3M, the hard plastic grey P100 from 3m, the eclipse P100 filters, and an H13 vacuum cleaner filter for the DCV5011H vacuum, which are smaller than this one that I finally ended up with: an H11 one made for the DC515 vacuum cleaner.
I was trying to level up to H13 or P100, but the resistance to airflow was way too high for all the other filters, possibly just because they were too small a filter media area.
Again, if you designed a better fan they could work, but not with any of the fans I could get a hold of off the shelf.
If you had a quieter fan you could also put it closer to your chin, or you could anyway, that's probably a good idea. It's just a hackjob right now, it works, that's all.
It's useful to note that because efficiency usually goes up when the airflow rate through a filter goes down, we are probably getting better efficiency than the H11 spec indicates, because the flow rate is of course much lower here than in a vacuum cleaner.
I didn't try to design a printable fan because I already did that a couple times on other projects and the problem is you can't get a good motor driver, so until that is solved there is little point, really, and in reality the number of people that can make their own fan is pretty small. This is something you can throw together and use *now*, like a CR box.
You just glue the filter on with Shoe Goo, or any other glue really. I like Shoe Goo because the tubes don't go bad and it's strong and seals well etc. when it's time to replace it, just pry it off and glue another one on.
Just put the nozzle in the outlet of the fan, it stays in fine, you may have to chamfer the edges very slightly or something to get it in. Most printers, unless they have linear advance, will bulge the corners out so they have to be trimmed down.
The nozzle geometry and design is reasonably well designed, it seems to get a fairly good distribution of air velocity in the right places, but could use a little tweaking. I deliberately reduced flow on the face side so it can be higher elsewhere, giving more resistance to disruption of the Clean Air Zone during forward walking in e.g. stores.
I recommend https://jlc3dp.com/ , it's about $5 USD including shipping for one of the nozzles, I checked.
Total cost can be less than $60 including shipping, plus you get some spare voltage regulators or whatever.
The case uses M3 screws, you can get cheap ones for a couple bucks on ali express, I like the self tapping kind, you can use the standard ones though.
You can also print the case for the speed controller from the stls provided, or you can use something off amazon etc. for controlling the speed of a 4 pin fan, there are many such products but I have not tried them, some are overpriced.
You can just reduce the voltage you feed the fan with but I find that makes a bit more noise.
I have provided the original CAD files should you wish to fork or modify the device, which is welcome!
My friend pointed out that air in motion can dry out mucus membranes and increase susceptibility to infection although it's not clear if this is a strong effect (I don't think it applies much to us here except for the eyes). I investigated a humidifier to add to the system, they make tiny usb powered ultrasonic ones, but the one I got was DOA so that's on hold. More to do for another time.
The code for the pico is almost trivial but is included.
Not all power banks can power the fan, you need a good one but it can be an old one, just capable of about 1.5 amps, but actually capable of that, many say 2.1 or 2.4 but are not actually good for that.
The wiring diagram is in with the other files as an image file for easy readability, no sense getting fancy for something so basic. It's more important to make sure it's easy and quick to read.
I used these voltage regulators but any step up regulator is fine, you can also use a PD board to get 12 volts but then you need a step down regulator anyway to power the pico. Either way, you can only get one voltage from the power bank. https://www.amazon.ca/gp/product/B0C54TDQXC/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
They are adjustable and good for lots of things, I just wish they were step-up and also step-down, that would be even more general purpose and not necessarily cost more.
The fans are the GDStime 120 mm centrifugal fans, also available on amazon, I like the ball bearing ones. Again, you can use the 2 pin ones and just PWM the power or feed them a varying voltage to control their speed but the 4 pin is kind of handy, and in my experience quieter and a wider adjustable speed range.
I just used some random ribbon stuff I had for the strap, and it goes through the holes on the fan on one side and gets taped to other stuff on the other side. The little carabiner to attach it around your neck is optional.
I used servo wire, 3 conductor, to go from the fan to the speed control box but you could also put the control box on the device and just run the usb cable, or use any other wires.
You can make it closer to your face, I find that as shown in the pictures is good, I prefer to have the noise of the fan a little farther away and the Clean Air Zone a little bigger, and air velocity a little lower, but adjust to taste.
Print the nozzle with the large end of the cone down, no supports needed, I use a brim but that may be optional.
To load the program on to the pico, see the raspberry pi foundation tutorials.
Those green connectors soldered to the pico and boost converter are called screw terminal connectors, 2.54 mm pitch. They are very handy and I use them by default in my projects, they give a mechanically durable connection and can re-configured even more easily than soldering. Also you can put multiple wires together easily and add or change things later. The three connection points per block on the boost converter terminals are all soldered together so I can easily add more devices or multiple ground connections or whatever, without messing up the old connections (it's nice to have a separate port for each wire).