Bed tramming

Discussion in 'Calibration, Help, and Troubleshooting' started by Sarah Nicholson, Jun 28, 2016.

  1. Sarah Nicholson

    Sarah Nicholson Well-Known Member

    Joined:
    Dec 20, 2015
    Messages:
    323
    Likes Received:
    115
    I've been getting some nice single nozzle prints out of my BigBox and decided it's time to move on to try dual extrusion. However I'm finding it really hard to get the nozzles perfectly level, which leaving aside the awkwardness of the adjustment mechanism has highlighted that my bed really isn't that level to start with. While the mesh levelling will cope with this to some extent, obviously the less adjustment it's having to do the better, and when it comes to dual prints it's clearly much more critical.

    I've spent far too much time with calipers trying to measure the height at each corner and get it perfectly level but no matter what I do it doesn't seem to help too much, those standoffs are difficult to adjust precisely and even when held with a spanner tend to move slightly when the screw holding the PCB down is tightened up, ruining all my careful adjustments.

    So I'm curious how other people are doing this and if anyone has any hints for a good way to get the bed as level as possible prior to mesh levelling.
     
  2. Ephemeris

    Ephemeris Well-Known Member

    Joined:
    Nov 25, 2015
    Messages:
    388
    Likes Received:
    241
    I use a small machinist square sitting upright on the bed (like a capital L). I go to each corner, position the X-rail over the square's blade and watch the gap between the blade and rod.
    One thing though, the PCB is pretty flexible so any attempt to adjust all four corner screws is likely to generate a saddle shaped bed. (Which is why I really wish E3D had gone with a three point suspension and possibly a thicker PCB.) So what I suggest is you leave one corner totally loose, applying no force on the PCB, until you level and lock down the other three corners. Then gently adjust the last corner.
     
    #2 Ephemeris, Jun 28, 2016
    Last edited: Jun 28, 2016
  3. Sarah Nicholson

    Sarah Nicholson Well-Known Member

    Joined:
    Dec 20, 2015
    Messages:
    323
    Likes Received:
    115
    Hmm that's not a bad idea, I've got a small square somewhere I'll have to give it a go. Thanks for the tip!
     
  4. Henry feldman

    Henry feldman Well-Known Member

    Joined:
    Sep 5, 2015
    Messages:
    969
    Likes Received:
    391
    You prefer this over a dial gauge?
     
  5. Ephemeris

    Ephemeris Well-Known Member

    Joined:
    Nov 25, 2015
    Messages:
    388
    Likes Received:
    241
    A good engineer is a lazy bastard trying to find easier ways to get things done ;)

    The square is handy, requires no setup and eyeballing the gap is quick and more accurate than I need.

    Besides, I'd have to walk down two flights of stairs and back to get my dial gauge.
     
    #5 Ephemeris, Jun 28, 2016
    Last edited: Jun 29, 2016
    Archania, Stefan and mike01hu like this.
  6. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    Getting the nozzles level is a pain but provided you can achieve it the general bed levelling (MBL) will be OK but it is critical to get levelling right to get consistent results. I have yet to invoke the second nozzle but it is level with the primary one and I check both when mesh levelling occasionally. I use a 0.02 or 0.03mm feeler gauge when levelling so that I don't need to mess around with Z-offset as the difference is too small. I get excellent first layer results, so rarely have failures due to the first layer other than today when I had a power failure or another occasion when the memory was not restored properly. On the latter, I have found that, for whatever reason, levelling drifts and the memory data is wrong, so I have taken to doing a mesh level at the start of a session; I suspect it could be minor transitions of one or other of the Z-drives and differential expansion rates due to the position of the printer i.e. one side close to a sunlit window but then it's more likely to be me.
     
  7. Sarah Nicholson

    Sarah Nicholson Well-Known Member

    Joined:
    Dec 20, 2015
    Messages:
    323
    Likes Received:
    115
    That's pretty much what I'm doing, but if I level the nozzles with my feeler gauge in the centre of the bed so they are both say 0.2mm from the glass, if I then move X to 0, the left hand nozzle is scraping the glass, hence why I need to fix my bed adjustment first. If I do the mesh levelling with just one nozzle I can also tell from the offsets that I've got wild variation in bed height, but so far despite fiddling with it multiple times I just can't seem to get the damn thing level.
     
  8. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    731
    Likes Received:
    93
    Some difficulties first. Depending on what you are measuring, say if all the rods are respectively, level (X rods) perpendicular (X wrt Y) level with (X rods wrt bed) etc. there is no reference to start with. Mesh leveling gets over most of this so quite handy but it does not solve the initial problem of where do you start measuring from?

    Next is that although the frame is reasonably rigid, the gap at the front means that the two X-Carriage rods are a stressed part of the frame instead of relying on the frame for support. Moving the BB, possibly on a not totally level surface may both twist the frame and put stress on the X rods. After a couple of months of printing I found that the rods were slightly loose in the X motor and Idler brackets.

    Part of the commissioning is to twist the frame until the gap between the front and the front X rod is equal, and then to tighten things. As lots of screws loosen with the constant vibrations, then whatever is holding the printer together to keep this distance equal almost certainly comes slightly loose altering the shape of the printer as if you had to bend it a little to get it right it will probably spring back into the wrong position once screws loosen a little or when you hold the frame to move the printer. To this end I have added a brace between the top front corners. This magically has lead to the X-Carriage moving much more smoothly than before, when it was already quite smooth.

    I'm now working on a way to get the bed more accurately placed prior to mesh levelling etc. The part of commissioning that makes things more difficult here is that at one stage the threaded rods are adjusted by hand, making any prior accurate adjustment a waste of time.
     
  9. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    Sarah, you have hit on one of the problems that I came across with bed levelling. What I forgot was that if I moved the head by hand then the auto-levelling mechanism does not come into operation and the head/bed gap changed and clashed in some places , so after an auto-home all bed movements should be made using the LCD so that the levelling data is applied to those movements. Try this to see if you get better results, I did and do not worry so much about the physical level any more.
     
    Old_Tafr likes this.
  10. R Design

    R Design Well-Known Member

    Joined:
    Oct 7, 2015
    Messages:
    944
    Likes Received:
    190
    My approach:

    1) Level the entire bed as per commisioning by turning the lead screws and measuring height from floor on each side next to the lead screws.

    2) Autohome => IR sensor is at sense point front left corner. Then disable steppers and move carriage IR sensor to front right corner, adjust screw until point where red light is at on/off point. Proceed to right back corner, then left back corner, then back to home corner. Continue to "circle" the bed until results become consistent.

    3) Bear in mind that there is real "sag" in the rods and therefore what we have done is not a true bed levelling, but that it's better than it was. MBL takes care of the rest.

    4) Sit quietly and pray for 5mins that MarlinDevTeam are working on automatic MBL so that we will be free to fiddle with the mechanics of the bed without having to set time aside to go through a whole calibration!
     
    Old_Tafr and Razar like this.
  11. Old_Tafr

    Old_Tafr Well-Known Member

    Joined:
    Mar 7, 2016
    Messages:
    731
    Likes Received:
    93
    Would you consider doing the initial levelling down from the X smooth rods (which is finally where the bed has to be level with) rather than up from the frame?
     
  12. PsyVision

    PsyVision Moderator
    Staff Member

    Joined:
    Sep 4, 2015
    Messages:
    798
    Likes Received:
    248
    Do you know if there is any code available for this yet? I had a look a week ago and it appeared not still.
     
  13. mike01hu

    mike01hu Well-Known Member

    Joined:
    Apr 22, 2014
    Messages:
    957
    Likes Received:
    167
    That is what I do but from the carriage using my digital calliper depth gauge and ensuring that I do not press down on the carriage as this will distort the readings. I have not done this levelling for a while as I find MBL provides all the adjustment needed and it is easy to see if there is any major shift in mechanical level during the MBL process; I would only do mechanical levelling if I had made a change to the bed and drive or noticed a major difference. It can be easy to knock against a lead-screw coupler and shift the bed level but not by an amount that MBL can't handle, unless you have the 8mm pitch lead-screw!
     

Share This Page