Simple 3D printer noise reduction with a concrete paver

I print a lot and this also means that many of my prints run overnight. Unfortunately, our bedroom is right next to my office in which my main printers usually are. When I got into 3D printing it was a real pain because the printing noises just kept me from some good night sleep so I usually only ran them during the day. A couple of years ago I started putting cheap pavers below my machines that you can get in many different sizes from the hardware store. You can see them in plenty of my videos and viewers keep asking, what they are and why I use them. What are your favorite cheap and effective printer upgrades? Let us all know in the comments?

So why do I use a heavy piece of concrete under my printers? Well, primarily because they reduce the noise level of the printer significantly. And the important thing here is that they don’t only reduce it in the room in which they are. Even more important, the setup that I use makes the printers way less hearable in the rest of the house.

Printing noise usually has two main sources. The first one are the stepper motors with induce higher frequency vibrations. As their name already suggests, they don’t turn continuously but in discrete steps. Most motors have a step angle of 1.8°, which means 200 steps per revolution. Stepper motor driver chop those full steps down to microsteps the make the motion more smooth, still it will never be 100% smooth after all. This stepping and the magneto-mechanical system that gets excited cause the well-known printing noise of machines that use, for example the good old Allegro A4988 drivers. More recent drivers eliminate that problem very well, but there are still some vibrations caused, especially during fast travel moves. The other, lower frequency vibrations emerge during direction changes, so accelerations, that induce forces on the frame and basically shake it.

Still, most of the sound isn’t actually directly coming from the motors themselves, but from other parts of the printer, like the bed or the frame, that resonate at those frequencies and emit sound. You probably all know these small music boxes. If you hold them in your hand and turn the crank they play a song but not too loud. As soon as you put it on a table the sound gets highly amplified. And this is also what happens if you have your printer sitting on a desk. The printer itself emits noise but the table on which it’s sitting is basically acting like a big loudspeaker. Even worse, the table transmits the vibrations into the floor and into your house, making even the floor in your bedroom a speaker. I put a microphone on the floor of my bedroom and once ran the printer directly sitting on the table, and on a paver with foam below it. I applied similar amplification on the audio so that you can hear the difference.

To avoid that from happening, we need to decouple the vibrating parts from and resonance bodies. This could already start by using those stepper motor dampeners that you put between the motor and the mounting points. What we usually do is to add a couple of rubber feet or felt pads under our printer though there are also other solutions around, like these 3D printed spring feet, which can be a good start. As you’ve seen, I especially like to use these cheap concrete pavers because not only do they give me a level surface, they also enable me to put softer materials like foam under the machine. They distribute the load more evenly for even better decoupling. You shouldn’t put your machine directly a soft materials, because not only can they warp the frame, they also might get ineffective because they get totally squished. And a big piece of foam is a potential fire hazard. Though also with the paver, make sure that no wires constantly rub over the rough surface.

To find out how the different decoupling solutions work, I set up a small test in which I placed my old Prusa Mk2, that still has a bit louder stepper drivers, on a table with varying media of separation and then ran a short test GCode, while measuring the noise level. I’ll also talk about changes in printing quality with these methods at the end of the video. I always ran the test 3 times and calculated the average peak volume. First, I removed the felt pieces for our baseline. If we listen to the test pattern, the movements in the x-direction are pretty quiet, and movements of the bed, so when the y-axis is involved, are significantly louder. The bed acts as a resonator, which is excited by the stepper motor and probably also the linear ball bearings. The maximum sound level is 67.7dB.

Test setup with 1m distance to the printer.

Adding the felt pads already reduces the noise level significantly and we end up at 59.6dB. The felt pads came delivered with my MK2, so how can we further improve on the noise. The next significant reduction is the use of printed spring feet. Here we end up with a maximum noise level of 57.5dB. I used them years ago, but quickly noticed that they might not be optimal because of their springiness, on which I’ll go into details in a bit. These pads consisting out of pressed together rubber pieces are often used under washing machines for a very similar purpose. Putting those under the printer directly or under the concrete paver results in pretty much the same levels leaving us at around 57dB. Next comes probably the visually most pleasing solution, and this is the paver with thick felt feet. Here the resulting noise level was 56.5dB. These blocks made from foam pieces are the parts that I’ve been using for the last two years under my main machine but are sometimes hard to find. Their noise level was pretty much the same as the last one with 56.5dB. I got the best results from this really soft packing foam where we even got below 56dB during the travel moves. And here a comparison between the best, and the worst.

So the soft materials provide a separation layer and prevent the vibrations from traveling from one part to the other. Different types of materials have specific properties and use cases. Rubber is well known for this application and as a rubber mat can be used even for heavy parts like your washing machine. For the lighter 3D printers, this type is probably a bit too hard and prevents good decoupling. Rubber is on the one side springy but also is dampening, and this is a point I really want to talk about because spring and damper are often mixed up. A spring is a component that can store potential energy due to its elasticity and release it again. An ideal spring leaves the energy in the system but can cause catastrophic resonances. A damper, on the other hand, is a device that usually absorbs kinetic energy, so movement, and converts it into heat and radiates it away, so it takes energy out of the system. In reality, there is no pure spring and no pure damper; it’s always a combination of both, and the ratio of spring stiffness and damping ratio is essential for a dynamic system. This is a quite a technical and complex topic and in our case, it’s probably sufficient to use trial-and-error. Still, for our application, this means that we, on the one hand need the spring part for the vibration isolation and on the other hand, if we don’t add an additional damper, the springs can cause the whole printer to wobble around and resonate, potentially causing problems.

For this reason, the spring feet, even though they work, are maybe not the best idea to use under a printer because they only damp very little. People have been tackling that problem by using squash balls between the printer and the table with great success, because they merge springiness with dampening. I tried to illustrate that with a small test by dropping different types of balls and recorded how high they bounced back. A hop ball is the worst in terms of damping and bounces almost all the way back, but that’s also by design. The golf ball lost 24% in height. The squash ball almost didn’t bounce back and lost 84% in height. At last, I also found a juggling ball that basically didn’t bounce back at all. It’s filled with beans, and the friction between the particles absorb most of the potential energy. So might this even be a better solution than a squash ball? Probably no, because, as I said before, we need both the elastic behavior and the dampening component, and the bean bag misses the springy part.

The squash ball seems to be a good compromise between both with some springiness but also lots of damping. Rubber has dampening properties due to the internal movement of the polymer chains. Foam, on the other hand, has also “some” of this property but also dissipates the energy because when it’s squised, the air inside travels through the pores and causes viscous damping. Due to the internal structure, so type of base polymer, open or closed porosity and so on, the amount of dampening can differ. Again I set up a small test where I dropped the golf ball on different materials and recorded the energy loss after the first bounce. The reference again lost only around 20% of energy. The harder and quite dense foam absorbed around 50% of energy. The rubber pad reduced the bouncing height by 64% and the soft foam sponge almost 77%.

So how do all of these results correlate with our noise isolation test? Actually pretty well, because also here the soft, open cell foam performed the best though also the other solutions were not that much worse. What I want to say with that is, that putting anything soft under your printer is usually beneficial because the elastic behavior decouples it from the table. Some damping might additionally be useful to prevent the printer from resonating. Oh, and by the way, if you enjoy this investigation, please consider subscribing and clicking the notification bell to not miss any upcoming videos.

Another interesting point I noticed in the past is, that adding some flexibility under the printer, can even improve your printing quality. For this reason, I chose four different printing bases and printed a 3DBenchy and also my ringing test on each using the same GCode. I placed the printer directly on the table with only the felt pads. I used the spring feet. I placed it on the concrete paver with only felt feet below and finally the paver on some foam. The difference on the Prusa was not huge, but you can see that there are fewer ringing marks on the parts that were on a flexible base. The part that were printed on a rigid surface showed a little more. In the past I have seen this even more exaggerated to my CoreXY Sapphire Pro for example, where the moving print head sits on the top and has quite some leverage for causing vibrations.

Ringing marks on test parts

The explanation that I have for the improvement in quality is, that if you have a flexible base, the forces that are generated during the directional changes don’t cause the frame and other motion components to deform but will cause the printer to move as a whole unit and therefore deform less. The concrete paver here again has advantages because it moves as a unit keeping the base of the printer on one plane. If you have flexible components under each foot, your frame might warp. Maybe you have different thoughts on this topic. If so, then let me know down in the comments!

In summary, putting a $2 slab of concrete under your printers with some soft padding, even only the packing foam that your printer came with, is, in my opinion, one of the most simple and best upgrades that you can make. Not only have we seen that we can reduce the noise in the room the printer is, but we also prevent our house from becoming a big speaker. Besides that, I have found that adding some flexibility to your base can also improve your print quality by reducing ringing marks.