Best z sensor fornglas bed



  • So heres my question.
    After revewing all the documentation, what i got out of everything was with a glass bed the capasitive sensor would trip the most reliable. But this was at the expense of accuracy.. i used the information on the duet website to make my decision
    Now later on in the game i am wondering if i could have made a much better choice.
    Please give me your input and what would be the best sensor for my aplication i did not care for the bl touch. To compicated from my piont of view at the time. And would rather stay away form it and the bent pins knowing my experience level
    Im with-in what seems to be about 40 microns of accuracy. is that an acceptable tolerance for a 3d priters first layer??? If not whats the best resolutoin if i dont care about the cost of a sensor
    Thanks for all your input. Im open to all suggestions 🙂



  • @ziggymanpopo said in Best z sensor fornglas bed:

    ............. Im open to all suggestions 🙂

    You did say that you are open to all suggestions so in that case, you could consider doing what I did. Basically the hot end can "pivot" upwards when the bed touches the nozzle. So the nozzle itself is the probe and the physical (mechanical) stop is also the electrical switch. https://somei3deas.wordpress.com/2020/04/18/my-6-input-mixing-hot-end-part-7/

    My conclusion at the end of that post was (quote)

    "I now have a Z probe / homing switch that is arguably as precise and repeatable as they come, has zero offset between the physical nozzle position and the switch trigger point in X, Y and Z, has no hysteresis, does not need any deployment mechanism, does not need any signal conditioning, and is self compensating in that the physical end stop is intrinsically linked to the electrical trigger point."

    You can see a quick video of it in action at the end of this post https://somei3deas.wordpress.com/2020/04/28/my-6-input-mixing-hot-end-part-8/



  • That had been a thought
    Im installing linear rails and a henera h oi t end uf it ever comes 🙂 fir the same ides a guy could go back to the micro sw. on the frame or mount a simple limit sw on the hotend mount as an option ??? This would also cute inconsistency in your z offset correct??



  • @ziggymanpopo said in Best z sensor fornglas bed:

    That had been a thought
    Im installing linear rails and a henera h oi t end uf it ever comes 🙂 fir the same ides a guy could go back to the micro sw. on the frame or mount a simple limit sw on the hotend mount as an option ??? This would also cute inconsistency in your z offset correct??

    I'm not sure that I interpreted your abbreviations correctly so I might have misunderstood, but a "conventional" switch will have some movement. So there will be an offset between the switch trigger point and when the nozzle is just touching the bed. Such an offset can potentially drift for a umber of reasons. My method uses fixed plates which are both the mechanical stop and the switch contacts so there is no offset. If the hot end is seated, the switch is in one state. As soon as the hot end moves off it's mechanical seat, the switch goes to the other state.



  • Historical testing I saw:

    Capacitive proximity sensors - huge variation in trigger distance with ambient temperature and humidity.
    Induction sensors - poor repeatability device-to-device, but acceptable repeatability with one device. Best performance with a ferrous surface (so steel).
    Snap-action mechanical switches - better than 0.025mm (25 micron) repeatability with a single switch.

    Basically the cheap micro-switches that everyone uses for endstops are actually far more accurate than we even need. The issue is more often the mechanical bits between the switch and the nozzle (the further apart they are, the more potential of in-accuracy).
    The big reason to use snap-action switches is the snap-action bit, they are basically a spring inside and as a result will trigger faster than a conventional switch and in an electrically more reliable way.

    I run printers with induction sensors and allow for any triggering variation with a thick first layer (0.25mm), but I also use a steel bed. I've never felt the need for anything else, my first layer success rate is high enough that I don't care.

    I loved glass to print on, but I'll not go back unless I can get some of that special high-iron glass they could make if I could afford to order a couple of tons.



  • I was thinking of trying a fixed sw mounted just behind the mounting plate.. using a micro sw on the mount would do the same as a sensor as fare as bead leveling and compinsation ???

    Having the accual micro sw
    Trip on the glasswould be a bad idea cause it would have to protruded past the hotend. Hence forth the aray of senors thay now have on the market. This protruding sw. would catch the prnit and pull it off the build sirface.

    Im looking at the diferances in the triger hieght and wonder if i should just live with the inconsistency
    I like the idea of using the tip of the hotend to trip the micro sw. Or any other sensor for that fact.

    It might be time to come up with a new alternitve
    For printing on glass and z sensors. Since i will have to redesign my carrage completely this would be a viable alterative and cure the problem
    This gives me a nice alternative. Thanks for the input. Ed..



  • So its safe to assume that the capasitive swnsor would work fine. I w oi uld jusy have a larger error factor.
    I would think enough guys arenusing them right ???
    Just sayen for now. ???



  • So thanks for all the input
    I managed to putball the pieces together and have the printer running consistent 1st layers.
    Everyone who chimed in contributed to a bit of the picture in the true spirit of this site.
    Once again i have had success at one more strugle to understand... thanks again for all the help
    The last piece i was missing was the actual hieght the capasitive sensor gives the best readings.
    2MM was what i found out.
    Thanks all...... Ed........


Log in to reply