Jump to content
Ultimaker Community of 3D Printing Experts

Recommended Posts

Posted · Hall effect sensors

Hey guys

I looked into leveling beds, and one sensor you might want to check out is a LVDT. I did a lab back in college where we used one to measure thicknesses of playing cards and found the thinnest. I don't know if the software/ hardware would be compatible, it needs to use an alternating current to work.

Also I saw the video of a delta printer using sensors to "level" the platform, I really like how those work since the entire system does not need to actuate the build plate.

 

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    Hey

    I have read about LVDT and found this paper:

    http://www.ni.com/white-paper/3638/en/

    You could replace the core in figure 3 with a magnet that gets attracted to the buildplate and is held back by a spring (could be acquired from a ballpen).

    But this will lead to building ones own sensor and that is a thing i want to avoid for now. If there is a preset solution i first want to use that, but i will keep LVDTs in mind if Hall-O turns out to have insufficient resolution.

    Edit: I have integrated daids fanmount, but my printer is jammed. So i will not print this out today. :-

    This is what i want to do:

    https://www.youmagine.com/designs/daids-fanduct-with-hall-o-magnet-mount-for-autoleveling

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    Published it. Now you should see it.

    Edit: Printed it. The sensor comes really close to the bed, which is nice. Will start to take in-situ measurements as soon as i found that **** little magnet that came with the board.

    And i found that autoleveling is already implemented into marlin. Nice!

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    Okay, i have everything up for the test.

    Materials and Setup:

    - I have the Hall-O attached to https://www.youmagine.com/designs/daids-fanduct-with-hall-o-magnet-mount-for-autoleveling#!design-flagwhich in turn is attached to my http://www.thingiverse.com/thing:94678.

    - The Hall-O is sitting on the bottom of the fanduct, the magnet is glued to its inside with hotglue.

    - I have replaced the aluminium sheet by a 2mm ferromagnetic metal-sheet i had cut out by hand. Note: This sheet is very uneven. In the process of cutting i used some old metal shears that were too weak for this sheet. I had to straighten things out with a rubber-hammer.

    - Use heating-PCB with the metal sheet.

    - measure the output Voltage with a cheap http://www.mastech.com.cn/html/en/products-va18b.htm

    - reset the z-endswitch.

    Process:

    - Ievel the bed so, that the nozzle touches the bed.

    - start autozero, and measure output.

    - drive down the bed with steps of 0.01mm and measure the output.

    I have made first measurements, which are not too promising. The change in output voltage is about 0.01V per 0.01mm on the first three z-steps starting from nozzle-height. This is close, but could work.

    What is more of a problem is repeatability. As Daid already pointed out for a wokring autoleveling-procedure you need to have reliable results that are repeatable. When i drove down to about 2cm (where the output-Voltage didn't change anymore) and drove up again to the same position as before the output was changed as much as 0.03V.

    This isn't very promising.

    I will make more measurements and check if inaccuracies in the sensor-mount are playing a role in this. But if not then the Hall-O propably isn't good enough for this purpose and i will have to look for another way of measuring the bed-distance.

    What a friend of mine mentioned a cool http://www.ti.com/tool/ldc1000evm which they claim should be accurate to the sub-micrometer area. What does the community think? Comments are desired.

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    I need to sit down and read your last post more seriously, but it sounds like you have some really juicy and valuable tests and data for us in there.

    Also could you post a picture of your apparatus? I think there are more than a couple of us interested in your setup.

    I love that you are trying to innovate!

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    I ran another test today, will upload pics and data as soon as i find the time, but now we have a "meeting".

    Have tightened the Fanduct. The accuracy seems to be better now. Maybe it is works with hall-sensors after all...

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    So finally I get to do this:

    Here is a picture of the whole assembly:

    Mechanics:

    I have redesigned daids fanduct, to take up the Hall-O-board into a cavity on the bottom and also take up a small magnet on its inside (see pic) thus keeping magnet and sensor-fiel at a constant distance.

    magnet

    (rotate this image 180°. The black Thing is the fan, the silver thing is the magnet)

    I replaced the aluminium sheet of my heatbed with a piece of scrap metal (ferromagnetic) from my workshop (I only attached it with three screws to the wooden board, instead of using all ten screws). You can see how bent it is.

    Bed warping

    The hotend was unmounted during the test to get the sensor as close to the bed as possible without touching it.

    height over plate

    Electronics:

    I powered it with the 5V powersupply of a former phone-charger. The signal was measured with a multimeter of this type: The whole setup gave me an signal of 1.73V which remained very stable because the magnet had a fixed distance from the sensor field.

    mounted sensor voltage

     

     

    What i did:

     

    My goal was to asses the relation between signal voltage and bed distance. I moved the bed up until it was close to the fanduct/sensor (see pic above), which caused the signal-voltage to drop from 1.73V to 1,408V (first series) or 1,29 respectively . From there on i moved the platform down in 0.1mm steps using the ulticontroller option for moving the axes. i noted the signal-voltage after every step.

     

    I did four series of measurements. After the first series i had to readjust the z-endswitch which caused the signal values to drop.

     

    Data:

     

    hall O Values

     

    hall O Table

    Y-axis: Signal voltage

    X-axis: distance of bed in mm

    Conclusion:

    The first two steps from 0 to 0.2mm ate the most important ones. Here we have a signal drop of more than 0.02V per step. Gr5 pointed out that the arduino could measure differences of 0.005V precisely, so according to that we could split each step into four steps of 0.025mm each.

    Theoretically this should be enough to precisely level the bed without having to touch it.

    The setting with the fixed distance should give out enough resolution.

    Further steps:

    As Daid said: resolution is not as important as repeatablilty. Mounting the sensor to the fanduct does not really ensure a stable environment for the measurement. I mean if a failed print gets pinched between bed and moving printhead the whole assembly will get tilted and from then on give wrong results. Also I want to eliminate the need for a fan on the printhead by using a crossflow fan.

    Therefore I will try to mount the sensor to the printhead i am designing. This should increase repeatability.

    Also i will try to get my hands on an oscilloscope to see the form of the output signal. Perhaps changes in printhead-directions or heating-current have an impact on the signal.

    So far

    Jump

     

     

     

     

     

     

     

     

     

     

     

     

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    Anyone measured the backlash in the lead screw?

    Seems like on a system with a lead screw z-axis, you wouldn't want to be switching directions back and forth during the print. Basically because any backlash from the lead screw might negate the benefit of auto leveling.

    I like the auto leveling with springs and clamps approach with the modification of a second set of strong springs for safety. Takes up some space in build height, but while your busy extensively modifying your machine, you could get some taller side panels laser cut, along with some longer rods.

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    I am trying to get a fully automatic bed-leveling because i want it to be able to print remotely. I want to set up a slicing server where you can send you .stl file and have it printed without even being there...

    Like a regular printer.

     

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    Great idea!

    I'm very interested in this becaus I also started working on a semi-auto leveling using a similar approach than in a linked video of this thread.

    How do you plan to level the bed - only via software or mechanically??

  • Link to post
    Share on other sites
    Posted · Hall effect sensors

    Software only. I think this is enough. At least the guy in this video above shows that it is possible.

    I don't know if autoleveling is implemented into my Marlin, but it is possible to do so and has been done already:

    https://www.matterhackers.com/news/automatic-printer-calibration-update

    Also Gcode commands G29-32 refer to autoleveling-procedures.

    http://reprap.org/wiki/G-code

     

  • Link to post
    Share on other sites

    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now
    • Our picks

      • Talking additive | The 3D printing podcast
        Why should we be the only ones asking questions? Join us and ask Jabil all your questions on September 22nd 5pm CET
          • Like
        • 3 replies
      • Ultimaker masterclass: Optimizing your Ultimaker Cura workflow
        Save your seat for either broadcast on September 23.
        What will you learn?
        · Best practices for iterative print preparation and every lesson you should learn from each print
        · The right way to use per-object settings and when they are most useful
        · Easy-to-use resources for anyone who wants to develop their own printer definitions, plugins, or print profiles
        · How to optimize print profile settings and whether to “keep” or “discard” changes
        · When is the right time to export your drawing from CAD? (Based on Ultimaker Cura’s surprising power as 3D control software)
        · And a whole lot more tips and tricks!
         
        How can I join?
        This free masterclass will take place twice:
        1. 11am CEST (5pm SGT, 5am EDT)
        2. 5pm CEST (11am EDT, 8am PDT, 11pm SGT)
          • Like
        • 0 replies
      • Do you use 3D printing at work? Let us know
        It doesn't matter if you are using 1 Ultimaker or 10, there is inspiration in everything. We're looking for ...
        • 1 reply
    ×
    ×
    • Create New...