IN DEVELOPMENT Inductive sensor and PEI aluminium plate

Discussion in 'Guides, Mods, and Upgrades' started by Alex9779, Aug 8, 2016.

  1. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    A few weeks ago I ordered an inductive sensor to use it with my Clever3D PEI coated aluminium build plate.
    I had issues with the IR some other people had to, the homing position was not constant.
    So here a small update what I used, how it works and some new "issues".

    I ordered two inductive proximity probes type "LJ18A3-8-Z/AX" on eBay from HongKong. They have an 8mm trigger distance and need a supply of 6-36V. At least this is on the spec...

    Some said that probe is too big but I think I just got the right one for what I need and it was very easy to get it working. Reading some guides, all said I would need a voltage divider because the signal of the probe is just the supply voltage but the endstop inputs on the RUMBA are max 5V.
    Reading some other Arduino guide I found some people using the sensor with just 5V, so I decided to try that.
    And I can tell you it works. I just used the existing cabling for the IR, had two switch the PINs in the plug but anything else... I crimped a plug to the sensor and used PINs to connect the plugs...
    The trigger distance to the aluminium is about 2.5-3mm, not much but enough, actually the distance from the nozzle to the bottom of the fan duct on the "Dual Titan" is less...
    I have no idea if this is so less because of the low voltage, using some steel it is a little more but far away from the specced 8mm so I assume yes...

    I modified the carriage to fit the sensor to it in the front right corner, on the pic you see the temporary mounted probe with the printed new part...

    IMG_0728.jpg

    The probe is working great so far but I had one new "issue": the trigger distance is temperature dependant!
    Not much but from "cold" bed (25° C) to "warm" bed (50° C) +0.1mm to "hot" bed (80° C) about +0.2mm difference... So I changed my start script so that I first heat the plate to 50° C then I home. And I levelled the bed at 50° C... And I changed my scripts in a way to wait always for the bed temp to be reached even on cold downs...
    I realised that when I cancelled one print and immediately started a new one and the Z distance increased so the first layer didn't stick anymore...
    Doing it this way gives a very constant Z homing position...
     
  2. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    Ah, so thats why its waiting in the 50C :p

    Would be interesting to see if other sensors are not that temperature sensitive. Is it so the temperature of the material sensed that changes the sensitivity or is it the temp in the sensor it self?
     
  3. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    I think it is the temp of the sensed material... It also happens if I move the sensor off the bed and heat it and the try it again... the sensor is still cold...
     
  4. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    Oh and the wait to get to 50 was in there before to have the bed at that temp when starting to heat the hotends so both reach their end temp at the same time... I just change the order when I home...
     
  5. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    I just noticed the extra wait to 50 when I reinstalled my Mac and retrieved your latest profiles.

    getting a head start from 50 helps but the alu bed still need more time to reach 85C than the hotend to reach 240C :D

    That temp difference is interesting. Wonder why.
     
  6. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    I have no problems when doing XT prints with 245/80... nozzle and bed reach the temp almost at the same time...
     
  7. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    166
    There appears to be no easy answer to electronic level sensing as all electronic devices are going to have some temperature, reluctance or light dependency. Perhaps the answer is to go back to limit switches . . . or not:confused::(
     
  8. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    @mike01hu well now that I know it it is easy to get around... I just level with the bed at temp. I do that because to get to 50° C after a print is much faster than to get to room temp from 80° or higher...
    And I chose 50° as the temp because all bed temps I print at are higher... For now my lowest bed temp is 63° C for PLA so I could gop for this temp...

    IMHO that temp dependance is much easier to handle than lighting conditions if it is that what cases the issue with the IR... I must say I have pretty steady lighting, LEDs in the printers no windows around the box and still I had issues with it. Not every print but it seemes the Z home point travels after each subsequent print until it is totally out of range. But it never gets back. Thats the strange thing with this issue... If it is environment related then there have to be good and bad homings but it gets worse each print...
     
  9. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    The funny thing is I have had more consistent homin with the red aluminium bed than glass. At times 9 of 10 prints are spot on.
     
  10. fpex

    fpex Well-Known Member

    Joined:
    Oct 18, 2015
    Messages:
    621
    Likes Received:
    58
    I have not tried the red yet ... Because my glass with kapton almost never fails!
     
  11. Tom De Bie

    Tom De Bie Well-Known Member

    Joined:
    May 7, 2016
    Messages:
    191
    Likes Received:
    33
    I've found some articles which explain why the inductive probe is temperature dependent. It's quite technical, but maybe someone might find it useful ;)
    http://www.schneider-electric.com.hk/documents/energy-efficiency-cup/Inductive-proximity-sensors.pdf
    http://www.ab.com/en/epub/catalogs/12772/6543185/12041221/12041227/print.html
     
    Stian Indal Haugseth likes this.
  12. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    Thanks @Tom De Bie,
    Someone here might be able to create some formulas estimating the changed perception in distance.
     
  13. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    166
    Thanks for the heads up. Yes, my concern was that the error always increased and never went back under static environmental conditions and I raised the issue some time back on whether the problem is a digital rounding error somewhere that accumulates. My software writing days are well past now so I have to rely on others to hack the issues. As @Stian Indal Haugseth says, I have less of a problem with the red bed than I do with glass and I have not had the time to investigate screening the sensors at the moment.
     
  14. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    @mike01hu good to hear you have the same impression on red. I think this might be that it reflects more closely to IR colour and not so many other colours so there is less "noise".
     
    mike01hu likes this.
  15. R Design

    R Design Well-Known Member

    Joined:
    Oct 7, 2015
    Messages:
    918
    Likes Received:
    183
  16. fpex

    fpex Well-Known Member

    Joined:
    Oct 18, 2015
    Messages:
    621
    Likes Received:
    58
    For homing this should not be a problem as it is done cold. Same for mesh levelling. So when would the bed be warm the the sensor be used?
     
  17. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    After a cancelled print the bed is warm... waiting for it to cool down to room temperature takes long... So I decided to do leveling and homing at 50° C this does not take so long... I am still thinking to raise this to 60° C because 63° C is the lowest temp I am using at the moment for PLA... but 60° is a little too hot for example for Edge to get off the bed after a print... At 50° C all materials I tested are easily removable from the plate...
     
  18. Stian Indal Haugseth

    Stian Indal Haugseth Well-Known Member

    Joined:
    Sep 11, 2015
    Messages:
    573
    Likes Received:
    95
    Not being an expert in materials I believed aluminium can warp expand differently with different manufacturing techniques. I believe cast aluminium is less prone to warp than pressed sheets. So ideally I would level at the temp I'll be printing on.
     
  19. fpex

    fpex Well-Known Member

    Joined:
    Oct 18, 2015
    Messages:
    621
    Likes Received:
    58
    what about swapping the bed? the PCB will cool fast
     
  20. Alex9779

    Alex9779 Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    2,405
    Likes Received:
    731
    If you have a look at the chart in one of the links above showing the temperature dependency you will see that starting from 50° the correction grows exponentially.
    So then you would have to re-level it you switch materials that requires a different bed temp...

    Could be a solution, you got more bed, I have one so no option for me... And the clips are not so easy removable and right after a print the bed is still hot no touchable with the bare hand...
     

Share This Page