The "Missing" Rostock MAX Build Step

[Bill Havins]

I continue to "tune" my Rostock MAX. I am hoping to be able to produce very precise prints. Despite having read every post on the forum that has to do with "calibration" and related issues (including re-setting "delta radius") I have not been satisfied with what I am getting from the printer.

Today I pulled my Rostock MAX apart. As I did I watched for component "fits" that might influence accuracy and repeatability. What I concluded is that the three vertical columns (the aluminum extrusions) were neither perpendicular to the build surface nor were they "plumb" relative to the build surface. So, I decided to correct these issues and discovered it is not easily done.

The short version of the story is I loosened every fitting on the top and the base. I then tightened the screws that hold the three "Tri Supports" to the Upper Base Plate. I then attached the "Snowflake" and the Onyx Heater to the Upper Base Plate. I then attached the build glass to the Onyx Heater with binder clips.

Next I used a large engineer's square to ensure the aluminum extrusions were perpendicular to the build glass. This required many adjustments to the button head screws that hold the extrusions to the Tri Supports.

At the same time I tried to ensure the extrusions were "plumb" to the build surface. This was very tricky because I could not simply "register" the engineer's square to the Upper Base Plate and hold it to the side of the aluminum extrusion; the Upper Base Plate is not flat at its periphery when the Tri Supports, the Snowflake, the Onyx Heater and the Build Glass are attached. The best I could come up with was to use the engineer's square like a "winding stick" while it was on the build glass (google "winding sticks" to read about this technique - it's a real pain in the keester but works pretty well).

So, the three extrusions on my Rostock MAX are now perpendicular and "plumb" to the build surface (the glass plate on top of the Onyx Heater"). Now I'll put everything else back together and, if I'm done before midnight, see how it all works.

Perpendicular and "plumb" - the missing step that just can't be done any other way. You can't rely on the Rostock MAX parts to register the extrusions square. Laser cutting leaves an approximate 5 degree angle on cut edges; if you are expecting parts to assemble square... And the flat head screws/lock nuts with their less-than-perpendicular pull on parts..who boy! You just gotta' get out the squares and check the alignment of the extrusions. Software weights just can't correct for extrusions that are not perpendicular and "plumb."

Bill

[Eaglezsoar]

I read what I could about winding sticks on google and am still confused. Is it possible to write up something that is easier to understand and helpful to all Rostock owners to allow them to align these extrusions?

Carl

[Bill Havins]

Carl,

Here's a quick explanation. In this first photo the engineer's square is resting on the build plate (the glass) and its vertical leg is oriented perpendicular to the Y Axis face. If you zoom in in the photo you'll see there is no gap between the vertical leg of the engineer's square and the Y Axis face - the Y Axis, then, is perpendicular to the build plate (the surface of the glass).

To test for "plumb" one would normally want to move the square to the side of the Y Axis. But, the build plate (the glass) does not extend around the side of the Y Axis. An obvious alternative is to simply place the square on the Upper Base Plate with the vertical leg against the side of the Y Axis. But when you do you get this with the square to the left of the Y Axis:

And when you put the square on the right side you get this:

When you get a gap at the top of the square on both sides its obvious the Upper Base Plate is not flat at its periphery (especially next to the Y Axis).

"Winding Sticks" are a technique used by woodworkers. When you're working long wide boards you want them flat from end to end. If they twist or "wind" you won't be able to build a flat or stable surface (imagine a four-legged table that has a "winding" top - you'll never get all four legs to rest on the floor at the same time).

To test for "wind," a woodworker will use a pair of sticks; each stick must have parallel sides. The woodworker will put one stick across the board at the far end, and one stick across the board at the near end. The woodworker will then sight across the top edge of the nearer stick, looking to see if the top surface of the winding stick at the far end of the board is in the same plane with the near winding stick. If the top surfaces are in the same plane the board is obviously flat. But, if one corner of the far stick sticks up above the surface of the near stick it is then obvious that the board has a twist in its surface (it "winds").

We can apply a variation of this technique to test for "plumb" of the Y Axis. In the photo below the square is on the build surface and is oriented parallel to the face of the Y Axis. By positioning the square so a thin sliver of light shines between its vertical leg and the side of the Y Axis we can determine if the Y Axis is plumb. If there is more light at the top or the bottom of the sliver then it is obvious that the Axis is tilted (is not plumb). When you look at the sliver of light in the photo you'll quickly see that the gap at the top and bottom are the same. This estimate of plumb can be made more convincing by moving your head slightly until the edge of the square just "touches" the edge of the Axis. When it is perceived to touch, moving your eye (not your head) up and down will confirm that there is no gap at the top or bottom.

Humans can detect as little as .1mm variation with this procedure. It's remarkable but true.

I wish I had a measuring device that might allow me to check for plumb in a more reliable manner. But, this is a whole lot better than what I had.

Okay...school's over for the day.

Bill

[Eaglezsoar]

Now I see what you are saying. Thank you for the lesson.

Carl

[Jimustanguitar]

So how are the results? Can you detect a noticeable difference in print quality, accuracy, and repeatability?

(I did see your post about the gauge piece, but was that before or after the squaring?)

[Bill Havins]

The results were worth the effort. Take a peek at this thread: http://forum.seemecnc.com/viewtopic.php?f=54&t=1871

Bill