DIY Laser Cutter Air Filter

Today, we create an air filter for the K40
laser cutter. The air exhaust of the K40 laser cutter blows
the fumes out of the window. Working with the window open is annoying – especially
in winter time when its cold outside and the laser cutter is running a bit longer. So we ordered a fan and a couple of filters
to create an indoor air filter. This is a bit of an experimental build and
if you have any ideas how to improve it please leave a comment. We ordered these pocket filters with the intention
to use them directly. However after laying them out on the table
this seemed too massive for a small laser like the K40.

We take them apart and will cut them into
pieces to get the necessary filter material. We aim for a 30 by 30 cm filter size and create
a pattern to hold the filter in FreeCAD. On the CNC we bore 4 holes in a piece of 4mm
plywood. We then use these holes to hold the piece
to the table while the CNC is cutting a grid.

If you don’t have a CNC you can either just
drill holes with a big forstner drill or cut a router template on the laser cutter that
you can then follow with a flush trim bit. The CNC leaves a rough surface that requires
a lot of sanding. After sanding we finish the filter holders
to make them easier to clean. The parts get a few coats of paint followed
by a final clear coat. We cut the filters from the pocket filters
but you most likely would be better off with buying the filter material without the frames. We will put links to the pocket filters and
the raw material in the video description. We use the following filters: A coarse filter that is classified as G3 which
is a coarse material that filters the big dust particles.

For all the filters we first mark the 30 by
30 area and then cut the material to size. A finer filter of the class F7 comes next
and will filter the finer dust. And finally we have a very fine filter F9
that removes most of the smaller dust particles. We made the four holes to screw the filters
together. We poke small holes in each filter, sandwich
it between two plywood pieces and secure it with M4 screws. The last filter we create is a bit special
as it should become an active carbon filter. We cut a steel mesh to the size of the two
panels of this filter. The top side gets secured by gluing a thin
strip of wood to the mesh and the filter. For three sides OSB strips are glued together
at the corners and secured with a few nails. While the glue is drying we bore holes in
four pieces that will act as spacers. We drill the holes on the wood lathe but you
could use a drill press or just drill them by hand.

The holes are a bit larger than the threaded
rod we will use in the next step. The OSB sides are glued to the mesh on one
side. In both meshes we cut small holes and put
the threaded rods through the holes. After putting the distance holder pieces over
the threaded rod the side is glued to the OSB as well. After the glue has dried the sides are secured
with a few screws from the outside. We also bought this exhaust fan. It comes without a plug so we will have to
wire a plug to the fan. All of the products we used will be linked
in the description. On the CNC we cut a circular opening in a
scrap piece of kitchen countertop for the fan to fit.

To connect the box the the hose we 3d print
a simple connector piece. The model will be available on our Thingiverse
website. The filter box gets a hole on which the 3d
printed connector can be installed. In this case using a router and an improvised
circle cutting jig is faster than using the CNC. The 3d printed connector is glued to the OSB
with a healthy amount of silicone. We cut all the parts for the box from cheap
OSB sheets. We also cut some OSB pieces to size that should
hold the filters in place. Their corners are rounded over on the router
table. Putting all panels side by side allow us to
glue and nail the strips that hold our filters in place across all thre panels at the same
time.

This way they are perfectly aligned. We put the first strip in place, then use
the filter itself to space the second strip and continue like this for all the filters. We can now also cut the panels to their final
length. Unfortunately we messed up the direction of
the box and cut the one end too short as we will see later. We cut through the strips but not the side
pieces on the table saw and also remove some more material of the freshly glued on strips
so that we can assemble a box. Before the final assembly we route a shallow
groove in the top of all four sides. We then assemble the box with glue and a lot
of clamps. You can now see the missing part where we
cut the box too short. At the corners we connect the grooves with
a chisel. The rubber band seal is glued in the groove
so that we get an air tight seal. When the fit of one of the filters it too
snug sandpaper or a plane will help to get a perfect fit. When putting on a lid we quickly realized
that there is still a bit of a gap between the filters and the lid.

Every filter therefore gets a thin MDF strip
on its top. To hold it in place we glue two additional
strips of OSB to its sides. And for an air tight seal we use a rubber
door sealing band that is self adhesive and can be glued on the MDF. Each filter has now its small head that makes
sure all air goes through the filter.

Off camera we doubled the filter fabric in
each filter stage so that the fit is a bit tighter than with only one filter. Now is a good time to fill 5 kg of active
carbon in the last filter. The top is closed with a couple of suitcase
hardware pieces. To screw them in plage we press the lid down
with clamps. Then we first install the upper part of the
mechanism.The lower part gets secured with a dab of hot glue before it's screwed in place. The air exhaust hose from the laser is connected
to the filter and the power of both fans get connected to the laser so that both are activated
when the laser is running. Now the big question is: does the filter actually
work? To answer that question we use an fine dust
measure device from ebay, connect it to the USB port of the computer and measure the PM10
and PM2.5 particle density. These are standards to measure particle matters
in the air.

We first measure the air quality without the
laser. When cutting cardboard and not having any
of the filters in the box the measurement go quickly up to over 500. After placing the filters back in and cutting
the same amount of cardboard the measurements stay fairly stable. So the filter seems to work. There is however quite some room for improvement
and it would be great to get your thoughts in the comments: Could we make the filter smaller? Could we sandwich all the three filters directly
together? Should we use more or less activated carbon? Should we use a stronger fan to get rid of
the fan at the laser itself? What are other improvements you would like
to see? We will most likely create a new version 2.0
at some time in the future. In the meantime please subscribe and follow
us on Instagram for a look at future projects and more improvements on the K40..

As found on YouTube

Related Posts