Rostock Max Calibration

[ccholas]

I am a first time 3D Printer user. I received my Rostock Max 3 weeks ago and have been assembling and calibrating every since. I have calibrated the print end (using a depth guage) within 0.08mm at the center, x,y and z points at 5mm above the heat plate surface (I have glass on top of the heat plate). I have measured and set my DELTA_DIAGONAL_ROD, DELTA_SMOOTH_ROD_OFFSET , DELTA_EFFECTOR_OFFSET and DELTA_CARRIAGE_OFFSET to match those measurements. I have ensured that the belt tightness is correct on all three ends. But when I lower the head to z=0 I get see variations between the center, x, y and x locations. And I notice that when I re-home the head I get variations in the points by as much as 0.1mm. I can adjust the screws at the x,y and z points. But it appears to be a water balloon effect. I barely change the screw height on X and Y,Z are then off. I am not sure what to do from here. Is it possible my heat plate is not flat? I have a 1/8" aluminum plate coming to sit between the heat plate and my glass. But why would the head Z height change between re-homing (G28)?

[mhackney]

A couple of things:

Don't calibrate directly to the Onyx, it really is not all that flat. The glass will be flat though. Also, you may be chasing a problem with limit switches not triggering at the same point consistently. I'm working on a solution for this but take a look at where the screw head hits the switch lever. You might be having a problem there. I tested this by moving close to a tower and then homed and went to 0 and measured. Repeat this and the measurement would be off by up to .3mm. Do this on each column. Chances are one is problematic but maybe all are. There needs to be a flat surface for the switch to reliably trigger.

Test this and report back your results.

[Bill Havins]

Understand that I don't have my Rostock MAX yet, so I may not fully appreciate what is happening as described here. But I wonder if optointerrupters might work better in this situation than mechanical switches. Using an optointerruptor in place of each axis limit switch might yield a level of accuracy equivalent to that of an optical encoder.

Mechanical switches, especially those that are thrown or switched by the head of a common screw, are going to show a "switch-point variance" that may look pretty ugly as mechanical tolerances go. Optointerrupters, "keyed" by fixed "flags" on each bearing plate, might yield a "switch-point variance" that is pretty tight (i.e., +/- .01 mm or less).

Again, I don't have a Rostock MAX yet (it's ordered) so I may be expressing an ignorant opinion. Just a thought.

Bill

[mhackney]

Many RepRapers use optical home switches. I have then on my H-1s. I intended to swap out my mechanical switches on my Rostock but frankly, they have been pretty good! Once I had it dialed in and added the aluminum heat dissipator that also keeps the Onyx flat I have had no issues with repeatability to put down the first layer of the print.

[Polygonhell]

FWIW I tried optical switches on my other printers and ended up going back to mechanical ones, I found the optical switches less accurate.

[mhackney]

Opticals can be very repeatable. Many of us use them on our CNC machine tools where metal swarf and coolant fluids create a harsh environment. Mechanical switches (much higher quality ones made for this purpose) are also effective. The opticals on my milling machines and printers are repeatable to about .0001" - the limit of my ability to measure it.

[ccholas]

Thanks for the recommendation. I did a couple things since the post. I placed a 1/8" thick piece of aluminum between the Onyx and the glass. I have that all secured down now. I figure that this will distribute the heat evenly and hopefully will reduce the variance in the heat plate. Second, I am trying to find a flat head screw that fits the bill. I think that you are right, the home position seems to vary because of the way the screws hit the switches. I think I found a good solution and will try tonight.

[mhackney]

I've been using the stock setup with mechanical switches and stock adjusters now and it seems very stable. I suspect it was more the issue of the Onyx warping when I was having problems. Since adding the aluminum plate I don't have warping and my Z is spot on every time.

[ccholas]

I bought some 2" 6/32 taper headed screws, 5/16 aircraft cable spacers, and some lock nuts. The spacers have a counter sunk side to them. I placed the taper headed screw heads on the counter sunk side, and then used a 6/23 locknut to tighten it on against the screw head. The surface is flat and hits the home switch the same way each time. I do still see variation on the first Z point just after it homes to the top. That can vary as much as 0.3mm each time. But after I cycle the print head around I find it flattens out.

Also, I placed a 1/8" Aluminum heat plate between the heat bed and the glass. My hope was that the aluminum would heat evenly and reduce the twist and bend of the FR4/PCB bed. I secured it with wider paper clamps and taped the edge before I clamped with Kapton heat tape. The bed and the print head are level now. The heat bed is not level and changes over temperature. Would have been nice to have a reference to this in the installation manual. Also, a deeper explanation of what the different measurements for the system mean and how to measure would have been useful. May have saved me 3 weeks of calibration and frustration. I am hoping to write up my procedure this week and submit it to this site.

On to the next problem.