Underbed piezo with PCB 1.2

  • Hi,

    How does the drawing look like? Just put 3 piezo disks in a row and connect to the PCB? Or something other is to Do?

  • Do you mean how do you wire them together or how do you fit them to your printer?

    Electrically there are three connections for piezo disks, they are all connected in parallel so it doesn't matter which one you use.

    Mechanically, there are many ways to do it, which one is best depends on how your printer works. Have a look at the reprap wiki page here for some ideas.


  • What for type bed do you have?
    Underbed works best for stationary beds like delta type printer or a printer like mine with moving gantry.

  • I have a CoreXY - Hypercube Evo

  • Hey guys,

    please take a look at here: https://youtu.be/6UOGXfSk7g0

    I just created a underbed mount for my HEVO and testet it in 1 corner. The test was very good.

    Tomorrow I will install a 4 corners and run a test.

  • @hevilp We can't see that link. It is for your login only.

  • @punamenon


  • @hevilp : did you test while moving the bed?
    This sort of setup is very sensitive and requires a lot of tweaking (as you will find out when you put more piezos).
    I found that adding more piezos made the setup more sensitive but the trigger range went up (from 10ยต to about 120ยต) which is not what you want.
    I had originally 4 piezos on my setup ( one in each corner of the bed). I changed this so that only the front corners have a disc and 2 extra piezos attached on the printer chassis to cancel out mechanical noise of the Z axis.
    I home the printer to the front of the bed (bed is leveled against the gantry) so that when I home my z axis, I always have the perfect Z height.
    In your hypercube setup, it is better to have the piezo on the nozzle or attached to the carriage.

  • I just tested it with 1 piezo under the bed, but the 3 others were connected, and laid on the floor.

    And yes, I homed Z and touched the bed with my finger. Only when I touched, I got the signal, not while normal moving.

    The first test was good, tomorrow I will print the rest for the corners and I test it again.

  • Test was not good, I do not get reliable answers from the piezos

  • @hevilp said in Underbed piezo with PCB 1.2:

    Test was not good, I do not get reliable answers from the piezos

    Yes that is to be expected with a moving bed ๐Ÿ™‚

  • @whosrdaddy

    I will build a voron, there should the underbed setup be perfect or?

    Any hints? I made 4 piezos in parallel order.

  • Yes, but make sure you find a way to mount your piezos.
    If you moun them in the middle you must be aware that they are heat sensitive and triggering range will change.
    My bed is mounted on 4 TPU feet corners and the piezo is located in a sleeve in the 2 front feet. My bed is mechanically flat and is leveled to 10ยต range so I only use the piezos for homing purpose...

  • I would mount them in the 4 corners.

  • @hevilp : I tried that in the past and found that this setup broadens the trigger range (which is not what you want ๐Ÿ™‚ ) but maybe you will be more lucky than me...

  • But your yt video looks fine.

    Why does the values are bad then?

  • In my video, I only have 2 piezos in the front feet (+2 "pickup" piezos mounted on the frame)
    So sensitivity is less at the back of my bed. This means that I can only do a mesh map for the first half of my bed (but I don't use G29 since my setup leveled). This setup gives me 10ยต range on homing (a nd I only home when everything is heated).

  • Okay ๐Ÿ™‚ thanks

  • I am working on doing this myself. There are some challenges.

    Because the piezo sensors rely on movement, the larger the base that you can cover within the sensors, the better. I had a relatively small base for my bed size, and that ended up with unacceptable results.

    So say that a piezo sensor requires 0.1mm of movement to trigger a response that the board can pick up. (Other options, like speed play a factor, too, but assuming that probe speed is constant, it will come down to movement.)

    If you probe directly over top of a sensor, you'll need 0.1mm of deflection to register a touch. for anything that is outside of the polygon defined by your sensor locations, there will be an axis of deflection, probably made up by the two most distant mount points. We'll assume zero bending of the bed itself, probably close enough for the forces and spans involved on most 3D printers. For my old setup (I'm changing it to be better) it might be possible to probe 80mm away from a sensor, which in turn could be 80mm away from the axis line. In that case, it would require 0.2mm of movement to get the same signal. Actually, since it tried to use an axis of deflection in between 2 sensors, I found that I needed over 0.35mm of deflection in one corner, which was a clearly unacceptable deviaton over the surface of the bed. In addition, probing in the centre of the 3 sensors that I was using needed an even smaller deflection, almost zero, which would otherwise be almost ideal.

    For temporary use, I decided to use a 3 point probing system, directly over each of the 3 underbed sensors, which resulted in useable leveling.

    I am currently constructing a solution where I am stiffening the support, and placing sensors under the head bed mounting scews. 4 in a 209mm square, under a 220mm by 275mm heated build platform. It will still be possible to probe outside of the sensor polygon, but I will probably define my mesh leveling to stay within that polygon in order to achieve the maximum consistency in probing. I don't have results yet, I'm still in the actual build process for this.

Log in to reply