Printed Parts Not square

[Pyropainter]

Hello

I have been trying to dial in certain aspects of my Rostock Max v2 for quite some time now and seem to be stuck. These include flatness of the bed to the machine and the ability to print objects that are square or as programmed.

First some printer history. This printer was purchased used and included Tricklaser carbon fiber arms.
I discovered that the bed base that the Onyx heater mounts to was not flat so I manufactured one from 1/2 MIc 6, I also have a piece of 1/4 mic 6 covering the heater. A cork spacer insulates the heater from the bed base. I also made an end effector from aluminium. This machine also has trick trucks. Also at bed height and toward the top to the printer I manufactured some braces that fit between the towers. They bolt to the machine using the slot in the tower extrusion. As viewed from the top my printer has taken a hexagon shape. Upon rebuilding this machine I am sure that the towers are square in both directions with an error less than .1 degrees. Through further reading you may discover otherwise.

Using a digital travel indicator with the center of the bed being zero I get the following results. Negative numbers show bed being lower positive numbers show bed being higher. At the towers I see in mm X=0 Y= .02 Z= -.001 the reading at positions opposite the towers X Opposite= -.09 Y Opposite= -.14 Z Opposite= -.10. Unsure if this is reasonable or not.

To avoid having to slice an object, wait for it to print, then check the results I have modified a ball point pin so that is spring loaded. Thus turning my printer into a plotter. In Matter Control I programmed a macro that draws three lines. One line from each tower base to a position directly opposite (length of each line, 190mm) . The result is an asterisk (looks like a Mercedes Benz symbol duplicated and rotated 60 degrees). I can now measure the angles between each tower and the angles between the tower opposites. This angle is 120 degrees. I have made 120 degree template on a CNC machine. After plotting the asterisk and checking with the template this is what I am seeing. Placing the template on top of the X,Y and Z tower locations appear to give me 120 degrees. While placing the template on the locations opposite each tower I see the following X Opposite greater than 120 degrees (roughly .4mm gap over the 115mm length of the template) Y opposite less than 120 degrees (roughly .4 mm gap over the 115mm length of the template) Z opposite greater than 120 degrees (roughly .4 gap over the 115mm length of the template). Something about this does not make sense to me....

I have tried working with the 210, 330 and 90 degree tower locations in the eeprom knowing that it would effect bed leveling and it does and not for the better.......

Thank you for reading this through. My hope was that I could initially answer any questions that you may have of me. Feel free to critique my methods and procedures.

[Qdeathstar]

I don't think any of that will help us get to the bottom of your issue, your tolerances on the bed, if in millimeters look not perfect, but workable. Maybe your arm length is set a little short.

How about a picture.

[Pyropainter]

Thank you for your input Qdeathstar. All numbers in my original post are in millimeters. However, I did notice one mistake where I described the points as indicated by the digital indicator. The reading at the Z tower should be .01 mm not .001 as I indicated in my original post.

I am now including a few photographs of rough drawings I created that will hopefully help someone to help me to me sort out my machine's discrepancies. I have tried to explain what each photograph is meant to represent. I am having trouble coming up with he proper words that will explain what I am trying to say. It is my hope that each image is worth a thousand words......

Fist a word of warning. I haven't quit figured out how to include images or photos within a post. This is my first time. The urls should lead directly to an image as described in the description prior to the url but I believe it also includes to the two other images that I will later talk about in this post. I hope it does not lead to too much confusion. I did use photo bucket. I may need your help with this as well.

This first image is of the 120 degree angle at each tower. The 120 degree included angle is created by lines generated by plotting a line from the other towers and though to its respective opposite.http://i1249.photobucket.com/albums/hh5 ... jrdncc.jpg

The next image is of the 120 degree angle at each tower opposite location. The 120 degree included angle is created by lines generated by plotting a line from the other tower opposite positions and its respective tower position. In this example I am using the Z tower location as a reference because its opposite opposite included angle generated by lines plotted form the X and Y towers through to the center of the heated bed was equal to 120 degrees. I do realize that my logic may be corrupt but I am not sure where to start as am learning as I go. http://i1249.photobucket.com/albums/hh5 ... hbzmwy.jpg

This last image is of the desired tower positions as indicated in black and of the tower actual positions as indicated in red.http://i1249.photobucket.com/albums/hh5 ... wnulcw.jpg

I have been pondering this issue for some time now and have not come up with a solution that will eliminate the angular discrepancies without creating discrepancies elsewhere. I have considered rebuilding this machine taking greater care to make sure the towers are closer to being square. As I mentioned in the beginning of this thread I am sure that the towers are vertical to within less than .1 degrees. I have machined a square with a base that is 1 inch wide by 5 inches (this keeps it from tipping over without having to hold onto it)
and is 12 inches tall. At 12 inches and .1 degree tolerance we would see a gap of about .02 inches at the top or bottom of the square. I am sure that I am seeing less than .01 inches. I have considered machining a square with a larger base and a height that will reach to the height of the Trick Trucks at their home position.

I do thank you for any and all input. At this point I am willing to do just about anything.

[Qdeathstar]

have you printed anything or is this just a measurement orgy? post a picture of your print.

[Pyropainter]

I have printed some parts. This printer does print some amazing looking pieces when the material is dialed in. The most impressive thus far was a Voronoi/Voronol Human Skull, there was some internal stringing but the results were amazing. there was also the Yoda and a Dragons head. There have been a few practical parts as well. Including a fixture that flattens 1 inch od by 3/4 inch id silicone tubing, so that it can be cut cleanly, this is combined with a arbor press fitted with a razor that will cut the tubing about 3/8 of an inch long. I finally broke the fixture after about 400 pieces. I have to say that I was rather impressed with the end result (creating 400 silicone rings that is).

As far as accuracy I have printed a couple of the "Asterisk" delta printer tests as well as a basic square box and have been less than impressed with the end result. I believe that the results reflect that of the plotted images as shown in an earlier post. Now, I am not looking to bad name our gracious host or the Rostock Max v2, I am looking to make my machine as accurate as it possibly can be. Also, I like learn. Maybe someone can learn from my experience also.

I am willing to learn form someone who has go down through the path that I seem to be going.

Maybe I am trying to set goals for which this machine is not capable. But I see people printing parts for printers that are capable of printing accurate parts which in turn are used for 3d printers or hardware such as gears for extruders. This machine should be capable of doing the same. This machine has built in parameters in which I am trying to tune in order to achieve such prints. I am looking for someone who has had the same issue as me and has learned from their experience and is willing to share their experience.

I have been watching and learning from this site for about a year or more now and find it to be an invaluable resource. I do not think that a day goes by that I do not learn something new.

Your help is appreciated,

Thank You.

[Mac The Knife]

What numbers did you come up with when you measured the cube you printed?

[Qdeathstar]

Why won't you post images of your prints?

[Tincho85]

I was going to suggest checking for tower lean but you said it's good. Then I read that you made a custom effector.
Have you measured it?? If it was made with a badly calibrated cnc he could be the one giving you problems.

Also, I'm in for the request of pictures. Sometimes the say so much more than words. Maybe others can see things you missed.

[Pyropainter]

I should be able to post a picture or two this evening or in the early morning hours. Probably print something so that it is hot off of the printer.

Thanks

[Pyropainter]

Tincho85, I do understand your concern regarding my cnc machine not being properly calibrated. I work in a family owned machine shop. If I remember correctly we laser calibrated the mill on which my parts were produced 2 summers ago. I do thank you for giving me that thought to consider.

Qdeathstar, You mention that I should post some photos. I have thrown out anything you may consider helpful in diagnosing my dimensional issues. As I am typing there is a 50mm x 50mm x 8 in the works.
I do plan on printing something that will show the angular deviations that I have mentioned in earlier posts. I do not believe they will be available earlier today as it is getting late (early). I stayed up way to late last night. If you keep reading you will see a couple of prints showing quality of a printed part. Thanks for your help as well.

The first image of a dragon head.
http://i1249.photobucket.com/albums/hh5 ... w9td8q.jpg

Second is image of a corner bracket that will fit into 1 inch square tubing. You will notice that the inside corners of the legs do not look very good. I believe this is where the printer moves from one leg to the other.
http://i1249.photobucket.com/albums/hh5 ... bv6dgr.jpg

To all, have a great weekend

[rubiks24]

ccavanaugh had posted (I can't remember where on the forum) but a guide to getting dimensionally correct prints. The reasons you are not getting perfect 120 degrees is due to either the squareness of your arms and or tower lean. Take a look at this PDF that goes through a pretty good diagnostic procedure as well as how to fix the issue.

https://github.com/ccavanaugh/Max-Metal ... f?raw=true

[Pyropainter]

Thanks rubiks 24 for the information. I quickly scanned the pages and found the information where he talks about his method to keep the towers parallel to each other. The last time I semi rebuilt this printer I built a platform from Mic 6 aluminum that fits on the printer in place of the melamine platform that the the Onyx heater sits on. This allows me to check tower lean towards the center as well as left to right with a square made from 1 inch thick aluminum with a base of 5 inches and a vertical reach of 12. Being able to check the left or right tilt of each tower should allow me to keep the towers parallel each other. An inside corner is cut out at each towers position of the platform to give me access to the center of the plate as well as one side of each tower. It sits on 3, 1 1/4 inch spacers that are fastened in the printers existing holes. My aluminum base plate that the heater uses these three positions when attached as well. The following is an image that will hopefully explain. The image is not great as I needed to use a flash to get a better picture, this material is rather reflective
http://i1249.photobucket.com/albums/hh5 ... 9xbcuh.jpg

I am offering an image prior to any explanations of a 50mm square x 8mm tall cube that was printed last night. To help clarify the explanation that follows. http://i1249.photobucket.com/albums/hh5 ... vonfuc.jpg In this print I am seeing the following deviations. All sides appear to be flat with the X axis side that runs from corners 3 and 4 being the exception. This side bulges out slightly. The vertical sides appear to be parallel. The horizontal sides do not. This could be an effect of the bulge. As far as being square. It is not. Corner 1 is less than 90°. Corner 2 is greater than 90°. As far as corners 3 and 4 go I would have to say they are close to 90°.

If the followers of this thread are not fully frustrated or bored to tears please continue on. At the beginning of this thread I mentioned that my printer looks like a hexagon when viewed from the top. This a view at the heated bed, the view just below the micro switches is the same. The basic dimensions to build these were obtained from somewhere on the SeeMeCNC site. The center section is 1 inch tubing the elbows are solid and key into each tower. They were produced in hopes eliminating the problem I am still having now. Roughly a year later. Did it help with the problem, not sure. Did it make it worse, I don't think so. I do not always have as much time as I would like to work on this so things are going slow. That is why I am asking for help.
http://i1249.photobucket.com/albums/hh5 ... y5ampk.jpg

I have been following this site for some time now and do enjoy it and have learned a lot.
Thank you to everyone who is following this thread.

[rubiks24]

Yep, you really need to read the guide in detail and focus now on the parallelism of your arms. Print those calibration prints in PLA and show us the results.

[Pyropainter]

Thanks to all for now. I will be playing machinist this week and free time will probably be limited.

[lordbinky]

One of the things I did the last time I went all perfectionist on re-aligning my Rostock V1, I made a holder that I zip tied a cheap laser pointer to (and repeatedly made sure all the other efforts were squared up as I could possibly make it). Then I slapped a toothpick into a glob of clay (squared and centered equally from the towers as best I could, which was also not the exact center of the print bed), and painstakingly tightened down the bolts on the towers keeping the tip of that toothpick glowing while I ever so slowly making sure they didn't twist (and they did, often, until I got it all tightened up). Now I'm slowly getting a metal max put together and trusting Brian's work outright and hopefully avoiding the process again until I decide to resurrect,beautify, and further frankenstein my first molten plastic spewing love.

Another habit I developed when calibrating, using the unheated hotend to gently tap on carbon paper with the nozzle.

Hard work on the ultra fine tuning in a configuration that isn't your final setup (such as using pen holders instead of the hotend, etc) can hide factors causing printing irregularities and not translate into a rewarding effort unless you enjoy a trip down rabbit holes. That isn't to say sanity checks on the movement and operation of the machine aren't helpful, it can certainly shine a light on many significant issues, but there's nothing like testing the real world setup especially determining print irregularities.

[Pyropainter]

Thank you for the advice.