If there are rights of passage in home built 3D Printers, at least one of them appears to be the printing of the “Zombie Hunter” (See: http://www.thingiverse.com/thing:69709) by Ola Sunberg.
Here is my attempt to print the Zombie Hunter at 60% size on my RepStrap One printer. I’m using White ABS 1.75 filament at 0.2 layer height and 235c for extrusion and 115c for the bed. Using a .4 nozzle and E3D Hotend driven by a QU-BD Extruder with an improved filament drive gear and minimalist upgrade from the folks at Deezmaker.
My printer is not optimized for speed yet so this print took 4+ hours.
The lighting does not do this justice for two reasons. The lighting picks up the filament layers and makes them appear much more pronounced than they are. And second, the design of this piece has intentional wrinkles and imperfections. It’s really quite impressive.
This is also my first serious attempt to use the OctoPrint server on Raspberry Pi. After a few false starts and crashes (I suspect power problems, not the software or Pi) It got it to work.
@Brian_Evans I’m more interested in learning and understanding the issues that affect the speed of the print. Right now, took at 10mm horizontal slice across the nose and I am making prints at .1, .19, and .29 at 100%/200%/300% speed of my Slic3R Profile – just to use as a benchmark. Then, I will experiment with speed, acceleration, and Jerk. One challenge that I have is that this is a RepStrap – and a mass heavy one at that --so I have nothing to compare to except myself.
Nice print. Pretty impressive for a repstrap. If you have a lot of bed deflection due to large mass being thrown around, lower your acceleration. My frame is not very rigid so my acceleration is dropped all the way down to 75 mm/s2 for all categories in slicer. Ignore jerk, it’s pretty much just the largest initial acceleration, so it only helps for the first step from rest. And don’t mess with speed and acceleration in firmware like I originally did, just change in the slicer you’re using. Although I’m probably the only one who was silly enough to keep changing this in firmware. If you’re acceleration is low enough, you can crack up your speed pretty high, the only limit being sufficient layer cooling. One thing to note, if you’re speeding up your print with the print controller, it’s probably overriding minimum layer time set by the slicer, so you may have insufficient cooling between layers. I also noticed in repetier host, the estimated print time is quite and underestimate. I think the reason is that the max speed is increased while the max acceleration is maintained. With such a low acceleration, the ramping up in speed takes the longest time, and I think repetier simply divides the remaining time by a scalar value. Anyhow, your print looks much nicer mine, so you’re doing something right, good luck
@Tim_Rastall No, it is a home built machine using whatever parts I could find, many from past projects. It’s quite unorthodox - in fact, the Z Axis is driven by Rack and Pinion because that is what I had to work with. The design is modeled after a CNC machine, however, if you look closely, there is a lot of wood involved. My Z Axis is heavy and a lot of mass to move a round. I need to shore up the chassis to keep it from disintegrating at higher speeds.
@Eric_Moy Thanks for the feedback on acceleration. Here’s what I am using currently, and I welcome feedback from you or anyone else…
First, know that I am using Repetier Firmware on an Azteeg X3. I drive that with Repetier Host.
In EEPROM, my settings are:
Maximum Feed: X & Y = 200; Z=40
Homing Speed: X=40; Y=30; Z=35 <-- this seems to work best for me.
Max Jerk=15
Max Z Jerk=5
Acceleration:
X=100 <-- this is the heaviest mass
Y=200
Z=50
AT one time I played with Acceleration as high as 600 but found that I missed steps and, since I am still trying to wrap my head around Acceleration and Jerk, I set these back to the above.
As for testing increasing speed, I had been doing it in repetier using the multiplier (up to 500) but that does not really tell me the actual.
I prefer to find workable settings for the Firmware EEPROM and then do the rest in slic3r so that I can document what I learn.
@Eric_Mack I’d suggest using your repstrap to build a reprap of some sort and transfering the electronics and hardware across. That is if you want to continue improving things and not to say what you’ve built there isn’t impressive, it certainly is but I think the heavy linear motion gear you are using will hold back your print speeds. (Some machines run at 9000mm/s/s). If you want to improve ridgidity, aluminium plate or t-slot is cheap and easily machined, you could swap out the wood parts for metal I imagine. Also, for $70-80 you could get hold of 6 x 10mm linear shafts and some sintered bronze bushings that would let you seriously reduce the weight of your moving parts.
@Tim_Rastall wrote: " I’d suggest using your repstrap to build a reprap of some sort and transfering the electronics and hardware across. That is if you want to continue improving things."
Where would the fun in that be? Seriously, I may do that at some point but when I set out on this journey I had a specific goal and that is the journey - I fully expect bumps and bruises along the way and I believe that I am learning a great deal more from this path than I would have learned had I started with a traditional design…
Yes, my current design is very heavy and yes that will absolutely limit the speed at which it can be used. What I am after right now is not so much the top speed but the learning of of the factors that affect speed all play together, including the mass of each axis, the stepper, current, speed, acceleration and Jerk - and a host of other things I have yet to learn about.
I’m grateful for the people like you and others that have posted feedback that inches me forward in this learning journey…
@Eric_Mack fair enough - I suppose I was projecting my own goals there, sorry about that. For me, I build a printer, to build a better printer or to facilitate a rapid prototyping upgrade process of the existing one - the Mendelmax I built 18 months ago is very nearly unrecognisable and almost entirely bespoke. At some stage I think I’ll be happy enough with print quality on the new bot I’m making to start churning out some of the models I’ve produced but the whole design/print/assemble/improve process is so rewarding I’ve kind of put actually making non printer related stuff on the back burner :). I noticed from your pictures you’ve got some slight Z banding, have you looked into a possible cause for that?
@Tim_Rastall I’m curious about the z banding as well. I though mine was z wobble at first on mine but it wasn’t the same pitch as my z threaded rod. BTW @Eric_Mack , kudos on using a rack and pinion, I dig the Frankenstein/millennium falcon vibe on your bot. I believe a reduced the banding on mine considerably by properly tensioning my x and y belts. My y carriage is still wobbly but I’ve found some fixes online that I need to try.
As for the speed multiplier, I think it is pretty accurate. I belief it changes the peak speed of each move, but the peak speed is also limited by the acceleration,ie, my acceleration is so low in slicer that the peak speed on small features is pretty darn slow, which is perfect. I have not going jerk to do anything useful, as I mentioned before. I’m testing as follows:
Keeping my acceleration low, I’ve been increasing speed with multiplier first, then in slicer. Eventually, I will start slowly bumping up my accel to the limit I get to much vibe or I skip steps.
I’m pretty sure you’re skipping steps is me an artifact of accelerating too fast, but even at slow accel, there will be a speed limit. Also, as I mentioned before, if your small layers are printing too fast, you’ll get bad warping and curling. I’m using pla at home so I can assist with a fan that I’m kludging together, not sure if that’s a good idea for abs, as it’ll probably give you horrible warpage.
