Sunday, November 24, 2013

Oversized Prusa I3 finished and then some

It has been so long since my last post. I have been very busy with work and play that I never got around to updating the blog. I have made it the home page on my browser to remind me. So here is my oversized I3 story in one brief summary.

My I3 has been finished for months and has been well used. As a matter of fact its my preferred printer now. It has an oversized print base. Currently configured with a 300x300 build plate. It can actually accommodate a 400x300 build plate if necessary. However to make this oversize version needed some modifications to the original build.

When I first built it I used the standard (at the time) 8mm steel smooth rods. It was quickly discovered that with the extra lengths in my build, this allowed to much flexure, which affected print quality. The quality was poor. So... back to the drawing board.

I managed to source very high quality (strong) 12mm smooth rods and bearings from a local CNC parts shop. So I reprinted the parts to handle those. These rods are very rigid, very sturdy and very heavy hehe, the two rods on the X carriage weigh over 1Kg.
Anyway this rebuild with all of the upgraded steel rods and threaded rods work a treat. It is extremely ridged.
This thing is big, it is hard to tell from the photo, but I had a friend's standard I3 here for a while and it is only slightly wider than the carriage and can almost fit through the gap  :-) It looked so cute!.

In the original build I also discovered two other issues which I resolved when I rebuilt it:
  • Moving the much heavier build plate around (at speed). There is always some print job that has fast back and forth fills that just find the right frequency to want to shake the printer apart or at least make it skip a step in the Y direction ruining the print. I adjusted the Acceleration and XY Jerk settings in the firmware to reduce the intensity of the direction changes to help resolve this. This increased the print times a little but the top speed is still very fast so on big objects t flies. Another option would be to upgrade the Y stepper to a nema 23 but I have not got around to that.
  • In the original build, looking at the original prints under a microscope (as you do), it appeared that I was getting some slight movement of the x carriage that matched the thread of the z threaded rods. Basically as the rods turn, the thread can put pressure on the X carriage and case a very slight shift affecting the print quality up the Z.  I was not sure if this would carry across to the new build but to make sure I redesigned the mounting of the x carriage to decouple it from the threaded rod.
    This was done by adding a separate lifter that holds the bolts that ride up and down the Z threaded rod. The X carriage just sits on top of these and rides up and down. Any lateral movement of lifters does not translate to the X carriage as they are not hard coupled in any way to the carriage. If there is any minor X and Y movement caused by the threads, it does not get translated by pushing or pulling the X carriage.
    This also has the advantage of limiting the damage that can be done in the dreaded event of the Z end stop being missed. If that happened, the hot end can be pulled into the Print Bed with some substantial torque causing major damage if you are not quick enough to stop it. However with this design, the X carriage is simply resting on the Z lifters so in this case they can ride further down the threaded rod and the X carriage will simply rest on the print bed and not be pulled down onto it. This also means in an emergency, I can simply lift the X carriage by hand if necessary as it just rids the smooth rods.      
The image below shows the Z lifter on the threaded rod. The carriage simply sits on the lifter and is not attached at all. I have also included a couple of prints at .2mm layer height. They have not been cleaned up.


Sunday, May 12, 2013

Prusa I3 Starting to come together

After some woodwork on Saturday and some assembly on Sunday, the basic shape is there.
Got some help from Stuart (Auzze) who dropped in with some bits and helped with assembly.
This is wider than the average I3 able to print 300mm wide instead of 200mm

Wednesday, May 8, 2013

I3 Printer Build Begins

Here the Rostock is busy printing parts for my new I3 printer. Parts are being printed in PLA at .2mm layer height and 50% circular fill.

This will be a slightly modified I3 as it will handle a larger 300 x 300 XY build area and 200mm in height.

Sunday, April 7, 2013

Yet another statue

This is another model from Thingiverse.  The model took about 9 hours to print and was printed at 0.12mm layer height and the model is over 1100 layers in height. Once again really happy with the rostock both in reliability and quality.
Hmmm I am beginning to detect feelings of insecurity as I appear to be surrounding myself with worrior women :-)

Sunday, March 17, 2013

Redid that model Full size

Well I liked that particular object so much I did it again this time its full size about 180mm across and 130mm high. Printed at .15mm layer height it took 11 hours to print.
This is as it finished so it has not been clean up yet (The supports are still attached to the sword handle)
I am pretty chuffed.

Saturday, March 16, 2013

Rostock 0.15mm layer height

Had a go at a model using .15mm layer height. This is a 50% scale of the original model. The original model would have taken about 8 hours to print so I scaled it down quite a bit. This took just over 2 hours. I would like to try the full size one as there would be alot more detail.
Pretty happy with the quality.. might have to give .1mm a go next

Sunday, February 24, 2013

Minor mods to Rostock

Hi All,
Just thought I would post some minor updates that I made to my original Rostock.

Pulley Case Covers: I found that anytime I put anything down on the desk it would seem to be attracted to the moving pulleys and belts at the base of the rostock while I was in the middle of a print job hehe.  So I made up some simple covers for the motor mount at the bottom of each axis.
The stl can be found on thingiverse:

Kossel Effector: I decided to replace the original rostock effector with the newer simpler design from the Kossel. This image shows the original design sitting on the print bed with the new, more simple one from the Kossel already fitted floating just above.
The new effector design is lighter, allows a more unobstructed view print job in progress and allows the effector rods to be attached in a more simple and secure fashion.
If you do use this part for the standard rostock you must make sure that the spacing between the effector rods at this effector end is exactly the same as it is where they on the X,Y and Z axis. The scad file from the kossel will allow you to adjust that dimension if necessary. I just used the standard spacing as the spacers from my Heim joints allows it to fit perfectly.

Herringbone Gears: Finally, just because I was board, I replaced the standard gears on my extruder with herringbone gears. These have less play and seem to be a bit quieter.
I just used the standard parametric herringbone gear openscad file that you can find on thingiverse adjusting the gear ratios for the ones I use on my wades on this Rostock.

Monday, February 11, 2013

Rostock printing real things :-)

Today I actually designed and printed something other than a trinket or a part for a printer. I wanted a better way to adjust the vertical angle of my home theatre projector. The commercial mount has the projector mounting bracket hanging and pivoting from the one point. Unfortunately this poses a problem as the strange mount on my projector means it is slightly back heavy. To add to this I have it mounted on an articulated arm so I can fold it back to the wall when not in use and then swing it out to optimal position when i use.
To make alignment to the screen easier, I used the free creo elements cad package to design an adjustable bracket and then printed out the parts. The images below show the results. It works a treat.
It adds stability to the projector as it is not mounted from two points on the bar and can quickly be adjusted if necessary by turning a thumb wheel to raise or lower it.

This is the design in the Creo Elements cad package.

This is the end result looking from the back of the projector. The bottom nut is a nylock nut that holds the threaded rod and in the top clamp section there is a standard nut in the thumb wheel adjuster with some washers top and bottom.

Saturday, February 2, 2013

Rostock more fine tuning 0.15mm layer height

Another big day today.
  •  I added the LCD controller to my Ramps board - as well as being able to monitor or control the printer, I particularly like how this can be used to slow down or speed up the print job as required.
  • Added mmall LED posts to the front axis to light up the print area - These are attached to the lower part of the Front two smoothrods. The carriages never come down that low and they generate nice highlights so you can keep an eye on the print job. 
  • Switched to KisSlicer for slicing the stl files - This seems to make much nicer prints, it is much better at handling shallow curves and seems to generate much more logical pathing. its also allows you to print the outer most perimeter slowly and then all the other very fast.
  • Started printing at 0.15mm layer height  - and this really starts to allow details to show in the models. Really happy with these results. The .4mm layer height I used to print at seems so far removed from what I am getting now. It is still not perfect as I am still learning how to tweak KisSlicer.
The image on the left shows the new LCD controller and the print area lighting. The other shows the owl from the other day (printed at 0.2mm) and now the the ape and the funny looking gargoyle at 0.15mm.

Tuesday, January 29, 2013

Fine Tuning the Rostock ...Success!!

I spent this long weekend resolving some issues on the Rostock.
Two issues that were causing intermittent problems were:
  1. The Teflon tube on the bowden occassionally working its way loose during a print.  
    This was a pain. If I tightened the clamp up too tight it would crimp the tube and restrict the path of the plastic filament. I tried a few things like filing the outside of the tube to provide more grip, adding some tape, etc.
    Finally I had a eureka moment when I suddenly woke up at 3am Saturday moning. I Then couldnt sleep, so I got up to try and implement the idea. I added a threaded nut onto the Teflon tube and redid the connection so the night is locked inside and cannot pull out. Success
  2. The extruder sometimes skipped when feeding filament during parts of the print that involved laying alot of plastic quickly. Basically did not have enough torque. I redesigned the gears shrinking the drive gear a little and increasing the size of the secondary gear to increase torque. 
Basically this resulted in the quality of my prints going from this:
To this..

This last print was at .2mm layer height abd too about 4 hours. I had to vary the speed at certain points to allow the filament to dry on the smaller diameter parts

Sunday, January 6, 2013

Rostock 3D Printing Vases

Well I printed one of my longest jobs on the rostock. A couple of vases. All up they took about 5 hours or so.
I have printed them in plain white as they are translucent and I intend to put some of those LED flickering tealights in them.
The one on the right is after doing more tweaks to the unit. I added sides and crossbeams to increase the rigidity of the unit which has decreased vibration during printing substantially.


Wednesday, January 2, 2013

Rostock Finally working Accurately- YAY!

Well I have been slack of late and never got around to debugging the Rostock printer.
I have had a long outstanding issue where prints were not perfect dimensions in X and Y
If for example I printed a large square 200mm x 200mm it would not be perfectly square. It would be out by a couple of mm in 1 corner.
I tried all sorts of measurements but could not work out the issue.
As I had a few days off over new years weekend, I decided to bite the bullet and fully rebuild it.

I totally dismantled it and cut new base and top boards. Assembled and squared everything up making sure everything was square and all rods are exactly the same length from the print centre, both Top and Bottom.
As mentioned in a previous post, I also replaced the printed rods and joints with manufactured Heim joints.
This time it all works very accurately.
Things that helped that I did not have as an early adopter for the first build:
  • A good set of precise dimensions for the base and top boards. Previously its was all angles and distance from centre etc. Very difficult to get perfectly accurate. This set of dimensions lets you locate all the key points just by measuring from the top left of a rectangle board. .  I still calculated the build centre point which is 175 In from the left and 175 in from the back. Lines going out from the centre point at 120 degree angles intersect the midpoints between each of the whole pairs.
    An important note about the diagram is that the holes on that are marked as 8mm. This is wrong they are only 3 or 4mm and are used to secure the plastic parts that hold the rods. The rods do not go through the board they are on the outside edges of the board.
  • A good calibration guide:  This is an excellent source of information that clears up so much of the black magic. Thanks minow :-)
  • A second try :-)  My experience from the first build made the rebuild easy. Basically it was done in a day.  
This time around I did not add the threaded rods for support. I am going to add some sides on the left and right sides that extend a couple of inches beyond the the back of the unit. I will then add a threaded rod X across the back for stability.

I will post an image once I have completed the build.