i3 Improvements/repairs Thoughts?

Have lately had a spate of bad luck with my i3.

I did my initial calibration/setup printing in PLA, and got acceptable results with it, but pretty much maxing out layer height at 0.33mm (For a 0.34mm nozzle) Naturally this doesn't produce the smoothest sides or curves, but it was good enough for the pieces that I made that way, including new legs and bearing/endstop holder for the Y axis. At the very least, my new parts are far superior to the crap that my kit came with, functionality wise, though not particularly pretty.

I then went on to start with ABS, and got some of the typical print failures, many resulting from the tendency of the ABS to curl. I started messing with ABS juice and the like, and again managed to get acceptable results, but now I wanted to improve surface quality, as many of the ABS parts that I wanted to print are for use in my car. ABS is the material of choice because of some of the extremes in temperature that happen in the car. From -45 degrees in wintertime to whatever temperature dark coloured things get to in a closed car in the sunlight, which I know to be hot enough to make lesser plastics deform badly, but I know that many parts in the car's interior are already made of ABS, so my hope is that it will be fine to use that. Anyway... I changed my slicer settings to use 0.125mm layer height, with a first layer of 0.25mm. I got some nice prints, but for larger things, it still likes to curl, which is seriously annoying. (Will have to make a box for the printer, and try higher bed temperatures.)

In the meantime though, my printer has developed some nasty quirks, which seem to stem from questionable design choices.

1. the smooth rods for the X axis aren't held firmly in place by anything but friction. They press through the right hand side carriage piece and into the left hand piece where the motor resides. On large pieces where there is a lot of fill (Using 45 degree rectilinear fill) the printer has developed a tendancy to start to eject the top smooth rod out of the side, until it falls free of the left hand side piece, and the result is a really shitty print. I have not ever allowed both rods to fall free, but if I were to go eat a meal while printing, as I have done in past, it certainly could happen. It appears to have done some damage to one or more of the linear bearings on the extruder carriage, as this now sometimes makes noise during the print, and has also started massive numbers of missed steps. Well, LM8UU bearings are cheap, and more are on their way. The smooth rods as yet appear undamaged. My proposed solution will be a cap that goes over the right hand side fitting and screws in, thereby holding the rods securely.

2. The belt for the X axis tension is a royal pain to adjust. When I made a new bearing holder for the Y axis, I made the design such that a zip tie is used to adjust tension on the Y axis, allowing me to set it up, and make minute increases in tension until it is adequate, but not too tight. (If too tight, cut the zip tie and start again It should also be trivial to use 3mm screws here, too.) I like the idea, and it works well for me. The tension adjustment for the X axis is a `7`shaped hole, Where you set up the belt at the top of the 7, then move it to the bottom once the belt is locked in place. In short, this sucks. the only way that I have had acceptable tension here not to have a sloppy X axis makes the belt tighter than I really want. the result is that it pulls on the smooth rods enough to give them a subtle bend, which I can`t see as being long-term good for the printer. In addition, it has a second effect which is seriously annoying, in that raising the Z axis quickly (Like the 5mm raise at the beginning of the Gcode from slic3r) can pull the smooth rod upwards. Even worse, this takes the Z axis homing limit switch with it, resulting in me having to re-calibrate the axis from scratch each and every single time. Therefore, I will make further adjustments to the X axis bearing holders to also have a tension adjustment similar to my Y axis adjustment, though probably with screws. Probably the same screws that hold the endcap which keeps the smooth rods in place, thereby killing 2 birds with 1 stone.

3. All of the problems associated with the one above have made me lose a lot of patience for adjusting the endstop on the Z axis. Trying to make minute adjustments to the position of the Z endstop is tiresome, and difficult to do precisely. In addition, since the Z axis smooth rods are basically held in place by gravity, they are prone to moving. Repeating the idea of using caps to prevent movement of the rods is a partial solution, and probably enough to not have repeat problems, but I would still like to be able to make very small precise adjustments to the Z endstop. »The current design uses a clamp to attach the endstop to the vertical smooth rod (The same method used to make the endstop for the Z axis, for that matter) where the switch gets hit by the X bearing holder. My thought would be to make the switch holder a part of the motor mount, elevated enough to engage the system. Another switch holder could be incorporated into the rod holder at the top, for an upper limit switch, preventing problems with builds that are taller than the printer can allow. I would add a threaded plunger to the bearing holder that can be adjusted with a screwdriver for precision control of the bed height. (At present, I set the switch to be as close as I can, then adjust the hotbed height screws for fine tuning. then more fine tuning to make it all level again. A major pain.) This should allow easy precision bed height adjustments.

4. The Z axis doesn't always stay plumb. Because the acryllic frame only has 2 attachment points to the Y axis frame, during the rectilinear fill, it visibly wobbles, which I can't help but think is having an adverse effect on overall print quality. The older Mendel design had this taken care of by triangulating the top mounting solution, though presumably had other issues which resulted in the new design. I'm not entirely certain what the correct solution to this is, but I plan on adding a diagonal stabilizer (or two) to the frame to keep it more rigidly attached to the printer frame. Making that adjustable will also mean that I can assure myself that the Z axis movement is exactly perpendicular to the X/Y plane.

5. What exactly IS perpendicular to the X/Y plane shifts too easily. The threaded rod frame of the printer twists too easily. I'd already addressed some of this with the new feet. By making the holes that the threaded rod fits through longer, it adds some rigidity to the frame design, but it still seems to be inadequate, in that the frame can still twist by several milimeters (As opposed to centimeters with the original design.) this is more than enough to throw off the bed levelling, even with small events such as moving the printer on the desk. The short axis of the Y bed seems to be okay, as the design uses a pair of threaded rods, and they are distant enough to make for a rigid rectangle. (Well, nit perfectly, but very close.) The issue is the long axis, which has onely 1 threaded rod per side, and then the smooth rod, which is not firmly held enough to enforce structural rigidity. I suppose that I could make the frame taller, and add a second threaded rod under the first, but this doesn't seem like the correct approach. Instead what I will probably do is incorporate a stiffening element to the box which I need to build anyway in order to help with keeping the heat up to prevent the ABS from warping during prints. I think that if I build the box to have several stiffening ribs along the bottom, I can keep total frame twist under 0.2mm, which will be good enough to not cause problems with bed levelling (<0.05mm difference corner to corner in nozzle to bed height)

So here's a question: Are other people having similar problems? Are these unique to the i3 kit that I bought, or are other kits made with similar designs? How standardised are the rod sizes and placements in these printers? How about the acryllic frames? Motor mounts? When I design these parts to solve these issues, is there interest by other people to get ahold of my designs?

---

MBot3D Printer
MakerBot clone Kit from Amazon
Added heated bed.

Leadscrew self-built printer (in progress)
Duet Wifi, Precision Piezo parts

Hello. The problem with the X-axis rod falling out is probably due to the wrong length rods being installed. The i3 kits typically include three pairs of smooth rods; two for each axis.

Typical lengths:

Short (320mm) is for the Z-axis
Medium (350mm) is for the Y-axis
Long (370mm) is for the X-axis

I suspect if you measured the Y or Z axis, you'll find the long ones installed there...

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

I am having a few of those problems my self with m i3 kit. As i learn more and more what everything does and how it works, i see design flaws everywhere. Last night i mounted the printer to a piece of melamine (laminated particle board) I ran zip ties through the board and over the threaded frame rods in the front of the Y frame, through the holes underneath the Z axis motors, over the threaded rods in the back of Y frame and through the slits in the plexixglass where the Y frame slides and mounts to the acrylic frame. I have almost ZERO movement now and that is pushing on it harder than the machine itself would make it move.

My Z smooth rods were a little too long and protruded out the top of the acrylic frame by 1/4inch. Had to drill a recessed hole into the cap that goes over it to fit over the exposed rod to be able keep it in place. Small fix, but come on!

Now a new problem that arose last night, the Z rod nuts, are not staying in the Z bearing holders. once the carriage begins to go up far enough one side slides up and off the nut. So i have to expoy the nuts in that hole, for now until i can remake those holders and make the hole tighter. If the expoy ends up working fine, then i won't bother with replacement.

I don't know about other kits, i'm assuming they should just because of the nature of the device and it's function, but i have no where to mount the filament reel. That's my next step which will be easy, but still....it's part of how this thing works, why was there no design for it?? Stupid.

I haven't even done a test print yet and i've had it for 3 weeks now, but i am valuing the learning experience. I'm learning more than if i had just received a fully assembled printer.

@Mindrealm: Definitely not the problem, as I measured all of the rods and installed them as per instructions, it is that the rods aren't held in strongly enough, and they're slipping out. Issues with the linear bearings are making things worse, for sure, but it's not a good design.

@Badkitty: Yeah, I just popped through your thread. It's different if the nut spins in the holder, but if not, I don't think that I'd epoxy the nut in place. There are reasons why you'd want the nut to be able to come out the bottom.

Quote
SupraGuy
2. The belt for the X axis tension is a royal pain to adjust.

Have a trawl through thingiverse for x belt tensioners (like [www.thingiverse.com]). They let you tighten the belt by just tightening the screw on the end and also place the strain along the x smooth rods which stops the squeeze of the whole x axis. There are similar things for the y axis as well.

Quote
SupraGuy
3. All of the problems associated with the one above have made me lose a lot of patience for adjusting the endstop on the Z axis.

You definitely need a screw based z stop adjuster like [www.thingiverse.com]. I just knocked up my own very quickly and superglued it to the side of the x carriage. Fraction of a millimeter adjustment by just turning the screw a fraction. One of the most essential and useful upgrades.

I'm not sure of the best way on the other issues, I built a box frame i3, and the big wooden box (despite being ugly) is much more rigid than a single plate like some of the i3 models.

Hi, I bought a I3 acrylic kit. I report the same problems with my kit. I changed a lot :
- Xends with clamp style rods holders, belt tensioning system, screw-style Z enstop adjusting.
- No wobble Z transmission (floating nuts and stiff couplings)
- 6mm rods for transmission (corrects some Z artefacts caused by the 5mm rod pitch)
- Custom shims to hold the Z rods into the acrylic frame.
- Geared extruder with 3 fans 1x 40x40x20mm for the hotend; 2x 40x40x10mm for parts cooling.
- Custom X carriage with endstop mount, fans/accessories terminal and cable management.
- Stiffer Y motor mount with motor shimming (I3 rework parts)
- Y belt tensioning system (I3 rework parts)
- Modified Y feet and frame to allow a base plate mounting (bolts on the feets, metallic zip ties on the frame)
- Stiff, heavy and flat base plate with holes to maintain the whole I3 Square. It's a torsion box filled with sand to absorb vibrations.

The list above does not include all the kit's components that have been replaced (Bearings, hotend, bigger motors...)

Both was necessary improvements. Now working on the enclosure, with fumes filtration and external cooling air and electronics. I should have bought a Rework I3 aluminum version at the start, with a most of problems already curedt (but not all of them !). It's a huge amount of time (and filament!) spent on the printer improvement. IMHO there's a lot of things wrong with the Prusa I3. Actualy, there's too much things you want to change. The Prusa I3 needs a reprapers workshop to fix anything wrong.

---

Collective intelligence emerges when a group of people work together effectively. Prusa i3 Folger (A lot of the parts are wrong, boring !)

Thanks, Opus. That's pretty much the design that I had in mind for the X axis tensioner and cap arrangement, which I was also going to integrate with the X endstop holder, in order to make it more easily adjustable.

I've also ordered a set of opto endstops, I had intended to set up a plunger to push the switch at the travel limit with the length of the plunger determined by a screw. I figure though that if I make the plunger modular, it's easy enough to modify it so that instead of pushing a switch, it triggers the opto-endstop as well. I don'y have the endstops here yet, so a V1 will have a mount for the microswitches and V2 will mount/trigger the opto endstops.

I'm not clear on what kind of Z artefacts the 5mm rod thread pitch will cause. At 4000 steps per mm, that's pretty clean for the increments that I use, or plan to use. I was using 1/3mm layer heights, which I know could cause a problem, since 4000 doesn't divide evenly by 3, but then neither does 3200, which the 1mm thread pitch of the 6mm rods uses. My curent slicer settings are for either 1/4mm layers, or 1/8mm layers, with a 1/4mm first layer. Practically all of my models are done with metric measurements, so this gives me something very close to actual measurements. 0.1mm layers is about as fine as I PLAN to go, but 0.05mm is likely possible. (I bet that would take a long time to print, too!)

I'm giving serious thought to a geared extruder, but I'm going to try the E3D V6 hotend with the drive that I have, which is a direct drive first. The hotend that I have doesn't seem that great, and the mount for it isn't very solid, allowing it to move more than I like.

My Y motor mount was okay, seems stiff and solid enough, though I've been giving thought to a welded angle iron bed which would change just about everything with the Y axis drivetrain, and make it super stiff, keeping the X/Y plane exactly where I want it. If I do that though, I'll probably also make a metal Z axis support, and bolt it all together so that nothing moves that I don't want to. Seems a little extreme though. I have been thinking though that if I were going to do a steel frame, I'd probably change to a design with a moving gantry, and then move the bed up and down for the Z axis. this would also be more in-line with doing a mini CNC router, as I could make a swappable module for either a heat bed for 3d printing, or a vacuum table for routing. Swap out a head cartridge that either has an extruder or a spindle, and voila. the biggest trick there is ensuring that all of the holes drilled in the steel are precisely aligned, but I bet that I could make a jig that would work with my drill press to do so. (I'd probably 3D print it. )