3D Print - MicroVac Turbo
A small (experimental) 3D printed handheld vacuum cleaner made from scrap parts
Creative Commons - Attribution
This is a small (about 60 mm diameter and 150 mm tall) handheld vacuum cleaner that I basically made from scrap parts. It is more like a proof-of-concept and was designed to fit the parts I had, not to use commonly available components. It is unlikely you will be able to print and build this without first modifying the 3D models, because the motor I used may not be easy to find, and the LiPo battery I used certainly is not easy to find.
This mini vacuum cleaner is built around a turbo style impeller, just because I wanted to see how well it would work (pretty well apparently). After trying out a few impeller designs from other people, I created my own. This was done purely by eyeballing and wet-finger guessing, it probably isn’t the most efficient design possible but it works pretty well for this application.
The vacuum cleaner itself is a very basic design, no fancy cyclones (if there is cyclone action, it would be by accident), just a widening collector to slow down the air and provide a relatively large filter area. I didn’t expect much of it but it works surprisingly well and is useful for cleaning keyboards or electronic devices and hard to reach corners. It isn’t ideal to suck up a lot of fluffy or very fine dust because this quickly clogs the filter, but it is great to quickly get rid of small pieces of junk. When sanding or carving something, I no longer care about the mess I make, I just vacuum it afterwards. It’s also useful to clean the inside of a confined space like a 3D printer enclosure. This is one of the most useful 3D prints I have made so far.
Note: there is no relation between this freely available 3D printable model and any products or brands that may use the MicroVac name. This model is not condoned or endorsed by any company.
IIRC the motor I used came from SparkFun long ago and they no longer sell this model at the time of this writing. The specifications were:
It looks like this motor is a clone of the Mabuchi RE-140RA. You can get these clones from the usual sources like AliExpress. Those who live in Europe, the Reely R140 motor sold by Conrad (article 518398) is a perfect match.
My first motor endured almost 2 years of quite intensive use before its brushes were completely gone. Two things probably contributed to wear:
Learning from the above, the lifespan of the motor can probably be greatly improved by sealing all holes in the motor housing, and only applying minimal amounts of oil or (preferably) grease. Minimising vibrations by balancing the impeller will also help.\
I doubt whether genuine Mabuchis would last much longer. It makes sense to simply buy a stockpile of whatever you can get cheaply. Replacing the motor is easy, especially when using pluggable contacts.
Any motor with roughly similar specs and dimensions should be suitable. Pay attention to the RPM and current ratings. Lower values will not work well. Even if the free-air RPM matches, a motor with lower current draw will be too slow when it experiences the load of having to drive the impeller. After my original motor was beyond repair, I tried a purported Mabuchi (bought via AliExpress, so likely a fake) but it had a higher impedance on its coils and could not draw the same current, which made it too slow. I managed to transplant the rotor of the original motor into this new motor’s housing and this works OK, although the commutator will probably fail next and then I will have to find a better replacement after all.
If you use a different model of motor, you will need to model your own motor mount. It may be a good idea to try a drone motor which will certainly be able to displace a lot of air. However, keep in mind that there will be an upper limit to the speed you can drive the impeller before it explodes into bits. Make sure to wear eye protection if you plan to find this speed limit… Also consider that the impeller will tend to spin faster inside the enclosure than in free air, especially when the inlet is obstructed.
The battery I used is a 220 mAh 551240 LiPo which came from a Chinese ‘spy pen’ gadget, the one I reviewed here. After the protection circuit of this battery failed, I tried to find a replacement and this proved very difficult. This bare battery remained in my stockpile of loose parts until I decided to use it for this project, with a basic protection circuit attached to it. In other words, you will likely need to adapt the 3D model to fit a battery you can easily obtain. A slightly higher capacity than 220 mAh may be recommended although autonomy is pretty decent even with this small value. A bigger battery would also allow more current draw, hence better performance. You will of course need to modify the 3D model to allow mounting a different shape of battery.
The switch is a pretty standard 2P2T slide switch, a classic model that has been around for ages and should still be easy to find. I connected the two arrays of contacts together because the currents it has to deal with may be a little higher than it is intended for. So far, no problems whatsoever.
For the USB socket. I used a simple microUSB break-out board that is easy to find on AliExpress and the like, the dimensions are 14.1x12.7 mm. Of course you could use any other connector, USB-C if you want to be fancy. You could also look for a combination charge/discharge protection board that does everything to keep the battery in good shape and offers an indicator light.
The screws I used are simple self-tapping screws, 3 larger screws that hold the parts of the main body and the motor mount together are 2.9 mm × 9 mm; and 4 smaller screws for the lid are 2.2 mm × 6.5 mm. Anything similar will probably work, you could adjust the models if need be.
You also need some kind of filter. The 3D printed ‘filter’ is insufficient, its only purpose is to hold the actual finer filter in place. The filter should be a circle with 60 mm diameter, and it should be the thinnest possible material that still blocks fine dust. I simply peel and cut one layer from soft multi-ply toilet paper, which works very well.
The recommended layer height for nearly all parts is 0.2 mm. Only the impeller should be printed at 0.1 mm layers with an extrusion width of 0.4 mm. For the other parts I recommend an extrusion width of 0.6 mm, which will be beneficial for strength. Since all parts are basically shells, infill doesn’t really matter and safest is to use 100%.
I used PLA for all parts except the filter ring (see below) and the impeller, which is ABS. The only real reason for this is that I’m able to produce more detailed prints with ABS. Use whatever works best for you.
It is highly recommended to print the FilterRing in an elastic filament like TPU. Otherwise there may be a considerable air leak around the filter edges. Whether you need one or two of these filter rings, depends on how flat and smooth you can make the interface between the CenterPart and Filter parts. If it is perfectly flat, this interface will offer a good enough seal on its own, otherwise an extra gasket is recommended. In my case the surfaces are perfect because I print on a glass bed, but an alternative is to sand and polish these two surfaces.
The impeller is provided in both clockwise and countercw (counter-clockwise) variants. Which one you need, depends on the direction the axle of your motor turns when viewed from the front. The motors I have tried, all turn clockwise when connected with their indicated polarity. You could of course just swap the wires if you printed the wrong version, although I’m not sure whether some motors aren’t optimised to run in a certain direction.\
There are 2 versions of the impeller, the v2 version has blades of alternating height. It is hard to test this quantitatively, but in practice it appears to perform better.
There are two lid designs, cheese has a pattern with round holes, grid is a regular grid and in theory offers slightly better airflow.
Several mouthpiece styles are available. The most useful ones proved to be 3, 4, 5, and 6. These are press-fit, which means you might need to experiment a bit with scale factors to get a good fit. In my case, I have to scale these pieces to 99.5% in the XY plane to get a good fit. I printed the number 4 and 6 pieces in flexible PLA to reduce the risk of scratching things.
A stand is also provided that can hold up to 4 mouthpieces besides the vac itself. When printed well, the two parts will fit and hold together when pushed into each other, otherwise sandpaper and/or glue are your friends.
Assembly is rather straightforward. You should start by pushing the motor into its mount (it’s press-fit) and then mounting the impeller. Tune the position of the impeller on the axle such that it just does not scrape against the intake part, also not when the airflow is blocked and the motor spins at its highest speed. I have created a spacer out of cardboard, fine-tuned with layers of adhesive tape through trial-and-error, such that I can simply push the impeller on the axle with the ideal spacing right away. When this is OK, assemble the other parts.
Cut a fine 60 mm diameter filter out of some thin material, like a single ply of soft toilet paper. You can use the coarse filter as a template. Place this thin filter on top of the coarse filter, preferably at the top side where the grid is slightly recessed. Then place the flexible ring on top of the fine filter and align everything. While keeping the set of filters in this orientation, insert them into the collector with its outlet pointing upwards. If you decided to use two flexible rings, place the second ring onto the coarse filter. Then screw the collector onto the main body of the vacuum cleaner.
The inlet has a tab on the side, the idea was to stick a piece of paper or plastic sheet to it, to act as a valve. This has proven unnecessary though when emptying the collector regularly. Junk will typically move to the sides and not fall back into the inlet unless too much has been collected.
Due to imperfections of the 3D printing process, the impeller may be unbalanced enough that it causes annoying vibrations. Balancing the impeller is fairly easy: mount it on a perfectly straight 2 mm rod (like a drill bit), and let the axle roll on 2 ‘rails’ that are perfectly level and flat. If well balanced, the impeller should not seek out a specific orientation when nudged. If it does, first look for obvious blobs at the side that likes to point downwards, and remove them. If this does not suffice, stick some adhesive tape at the edge that likes to point upwards, and try to dose it such that the impeller no longer has any preferred position. Enjoy your vibration-free MicroVac!
How to wire this thing will depend on the specific parts used, but in general it is rather straightforward. One wire of the motor goes to common ground, the other goes through a switch to the positive terminal of the battery protection circuit. The positive terminal of the battery’s over-voltage protection circuit goes to a charging port. Ideally there should also be a current regulator to limit the current during charging. Usually, a good rule of thumb for the charge current is the same milli-amperes value as the battery’s capacity in mAh (or “1 C”). In my case, the current limiter is extremely simple: a resistor chosen such that the peak current is not much higher than 1 C. This is inefficient, but OK given the small capacity of the battery I’m using and the low current it charges at. For a higher capacity battery, you will want to use some active charging circuit, because a resistor could become dangerously hot.
There are some ready-to-use boards that have a microUSB (or nowadays USB-C) port with charge and protection circuit all-in, but you will need to modify the 3D models to fit those.
If you cannot find a motor that runs at a good speed between 3.8 and 4.2 V, you will have to add some kind of booster circuit, or use 2 batteries with possibly a step-down circuit.
Refinements that are good practice: ensure there is a small ceramic capacitor as close to the motor terminals as possible, to filter out noise (may not be needed depending on the type of motor). Also, a decently sized capacitor across the battery (I used 470 nF), to buffer the initial inrush current when switching on the motor. This should also be placed as near to the motor as possible but must come before the switch. Without this capacitor, the protection circuit of my battery would sometimes engage because the inrush current could be high enough for the protection circuit to consider it a short-circuit.
Using crimp terminals for the motor will make it easier to replace than when soldering it. The RE-140RA and clones have tabs that accept plain flat 2.8 mm terminal connectors. Ideally the connectors should be for 0.3 mm thick tabs but if hard to find, 0.5 mm connectors will also work after some careful squishing.
There isn’t much to say about how to vacuum things, just turn it on and go for it. If you didn’t make a valve to keep dirt inside, the only thing you should avoid is wildly shaking the device or allowing it to fill up to a point where dirt can no longer stay inside.
Depending on your wiring, you may be able to get an extra boost by connecting the USB cable. In my case the gain is minimal due to the rather high impedance of my barebones current limiter. Make sure you cannot send the motor into explosive overdrive when it receives 5 V instead of the usual 4.2 V maximum of a LiPo battery.
To empty the reservoir, unscrew the collector from the rest, obviously while keeping it upright. Then either take out the filters using tweezers, or just immediately topple the whole thing over above a dustbin, catching the filters with your hand. If you’re careful, you can shake the fine filter and reuse it, but the effort to make new filters is pretty small anyway. Afterwards, reassemble according to the above instructions.
The mounting/removing of the filters and the screw thread connection to mount the collector aren’t very user-friendly, but this was only a proof-of-concept after all.
