Full Printrbot plate. Bet you'd love to see slic3r settings.

Show off your projects and talk about them

Full Printrbot plate. Bet you'd love to see slic3r settings.

Postby FalconFour » 2012-Aug-Mon-13-Aug

Last night my self-sourced Printrbot (glory of open source architecture! :) ) became a daddy. Finally.

526565_4129926640431_225984383_n.jpg


A full plate of the latest revision of all the Printrbot parts except 4 - the two extruder gears and the two pulleys, since they would be printed best with unique settings versus the rest of the static parts.

A LOT of work went into this. First, about 2 hours of fine-tuning the part placement within the 200x200mm bed (I wanted to avoid getting right next to the edges which tend to be unstable, so I packed them quite tightly into about 187mm square. Many revisions later, I finally got all the parts arranged. Then, in the perpetual setup-and-calibration process of many failed prints and various other unrelated prints, I found all kinds of problems with Y-jerking and slop, extruder slipping/grinding due to over-speed, extruder width slop, X tilting slop, whole Printrbot wobbling and causing slop, Z homing (just implemented, was always doing a manual "home" every print with little trouble), uneven printbed, and PARTS. CURLING. EVERYWHERE.

So a short list of the problems and solutions, from the most elementary problems to the most useful and "why was this so hard to Google" problems...
  • Extruder slipping
    The extruder idler design is the most frustratingly poor design of all the parts on the printer. Horrible. Crap. Can't say enough bad about how horrible the extruder idler is. The screw hole is always printed with too much layer slop and I have to Dremel the hole big enough to fit the screw, and it inevitably weakens the already-piss-poor hinge design. And the bearing never fits in the slot either around the "spacing" bumps, so I have to Dremel those out too. Both the professionally-printed one I got with my plastic parts, as well as the accidental replacement I got from Brook directly (I was missing an X-idler, but got shipped an extruder idler - good thing, since that first one broke and I already printed an X-idler using zipties to hold the axis together)... both snapped across the layers under the reasonable tension against the hobbed bolt. I've on my third extruder idler now, and last night's extruder behavior and a small crack in the idler looks like I'm going to be using the one I just printed here and going for the 4th.
    As a result of the sloppy idler design, I have tons of trouble with tension - too much, could break the idler (again)... too little, it slips the filament. Also had trouble with the hobbed bolt being squeezed too tightly - at first I didn't know I needed to screw it on loosely and lock it in with a lock-nut. But that idler... please, someone, please design a stronger idler!
  • Base wobbling slop
    This is an effect caused by the "jerking" around alternating corners in gears, narrow infill, etc. In my setup the bed is large and bulky, and the extruder is also heavy under the weight of the stepper. So when it jerks around, the whole machine (and even my desk, a bit!) tends to wobble. In turn, that affects the inertia of the movements and slops the hell out of the print. I've resolved this by attaching printrstands to the base - and, well... adding the 4th base threaded rod (long story - been running most of its life with only 3 threaded rods, recently went and added the 4th when I added the glass bed).
  • Slow bed heat-up
    rugh. Still trying to track down the cause of this problem. The bed heats up comically slow - and can't even reach 110C if I don't put insulting napkins over the bed to allow it to heat up. With insulation during warm-up, it takes 10-15 minutes to warm the bed up. I measure 11.85v right at the wires going to the bed, and about 12.05V at the contacts of the power supply (not the Printrboard end, but the lugs I installed on this custom power supply that bolt down the supply wire). The power supply is a sloppy piece-of-crap 250W Bestec power supply, but the Marlin firmware doesn't appear to do any voltage-checking and heater-switching to associate a crap PSU giving out good voltage with slow bed heat-up. The supply wire to the board does get a bit warm (very hot at the board end), and I'm starting to think it may be a weak solder job. Anyone else got input on their heat time? Anyone need to use insulation to get up to 110C?
  • Z homing
    OK, I still can't figure out where the Z endstop is supposed to be mounted and triggered. It still eludes me. As you can see above I mounted it on one of the base rods on the front side, and used some junk tape to crap-space the bottom of the base to trigger it. It sucks, bad. I still do manual adjustment each time to make sure the Z rods are level. It would be nice if I could make it automatic, but with the steppers wired together and still having trouble with the right-side Z rod straining a little on each rotation (I think the coupler is uneven), I still wouldn't trust it to stay level between prints. Still a WIP, but where the hell do I mount this thing?
  • Uneven print bed
    Ugh. Seriously, how is it possible to get a level print surface using a wood built platform that loves warping? Mine was printed mirrored (bought it on eBay), so I first assembled it with the rough side up so the Y belt clips are on the right side. It looked like crap. It eventually started warping with a curve along the Y axis (pulling the board/rods together), and when I "refreshed" my 'bot with the 4th rod and glass bed, I moistened the board, stuck it in the (large) toaster oven for a few minutes, then set it on the (hard) kitchen floor and sat on it for a minute or two. Perfectly flat now, but really? That has to be a common problem. Why not metal or something that doesn't warp on its own? Adding the glass bed helped a lot with the curve in the PCB heated bed (always used Kapton tape, though), but I still get a little variation from one end of the Y axis to the other. Nothing in X because I level the Z rods manually, which compensates for any tilt in the X axis... but nothing I can do about Y!
  • Y belt tension
    Is it just me, or are the Y belt clips just lacking something... "clippy"? They slip a ton if you jerk the bed around a lot, and you can't usually get a proper amount of tension out of the thing since you can only get it to stay put on the increments of the belt teeth - so no fine tuning. Solution? I printed an adjustable belt tensioner. Broke the damn nut-hole (why can't these nut holes just be universal and round? really?) and snapped off the top while trying to fit a nut in there... used a zip-tie to hold it back together after Dremeling it to the right size. Other than that, it almost totally fixed my Y-slop, but it still has a little play I can't get rid of without breaking the now-fragile tensioner or slipping the belt out of the flimsy, smooth Y clips. Gravity affects the Y axis "bias" - if the bed is near the end of the Y axis with the platform hanging way over the edge, gravity causes the Y to have a "+ bias", where it's inclined to pull forward against any Y-belt sloppiness. When it's near the back (near 0, with the platform retracted), it tends to have a neutral bias that inherits whatever bias it carried with it from the inertia of the previous move. So things printed near the end (near the Y200 side) will come out cleaner than something printed near 0, because of the "+ bias" that reduces the slop. To compensate near the end of this print in the picture, I placed one laptop 2.5" hard drive face-down/board-up under each front leg of my Printrstand to slightly tilt the 'bot backward and give it a "- bias" throughout. Slop fixed.
  • Extruder width slop
    Now this is an ongoing problem. Can't tell why, but even though I measure the right amount of filament being pulled into the extruder while calibrating and checking, I never get valid prints when measuring the filament diameter accurately. I measure 2.87mm with my (Harbor Freight, but mid-grade "class") digital calipers... but when I enter 2.87 into slic3r, it always blows too much filament into the print, blobbing everywhere and turning a 40% fill into 60-75%. I ran this print with 2.97mm setting, which seems to work, but it's been an unpredictable crapshoot. I can't tell if it's because of my hand-drilled 0.6mm nozzle (started life as an unsuitable 0.35mm tip I didn't know was not the standard), or if slic3r's math is just off a bit. I HAVE seen some significant bugs and odd behaviors in slic3r, though...
  • Odd behavior in slic3r (0.8.4)
    Figure it deserves its own point. First, the math in the extrusion width is a bit off, or my calibration, or something's slipping, but... yeah, above. But most concerning is a strange behavior I always see it doing on every print: random movements and "dotting" at the end of each phase. It wastes time and often damages the print with rapid, seemingly random and pointless (no plastic actually touches the print) movements after each layer - it jumps from point to point retracting/advancing the filament rapidly just as it finishes the perimeter or fill phases of most layers. I imagine it works in order of large to small "operations", but that's one of the bigger problems I've had - during many of these movements, the head travels rapidly in a straight line to its random destination instead of traveling slowly from part to part doing fills or perimeters. Hence, it often crashes into edges of parts as it travels, either skipping the X or Y axis steppers, or snapping the part off the board. Very, very, very frustrating when it's not something it should be doing to begin with.
    Another odd behavior is the experimental - but potentially very helpful - "Arc" movements (G2 and G3 in Marlin). Instead of slic3r writing (and computing - it's already SLOW!) G1 commands for every point along an arc, it uses an "Arc" command to have the firmware handle the curve itself. It tends to do a cleaner, smarter job of doing the curve since the firmware knows what's under its own hood. I'd love to use it, but slic3r places many arcs in the wrong places! The preview in Pronterface shows lines curving in totally inappropriate places, and the print agrees - perimeters intersecting each other, some lines drawing outside the part entirely. It's experimental, but this issue hasn't come up in any searches I tried to do. Rather frustrating since it could be rather beneficial in G-code creation time and in code efficiency.
    Finally, it would be quite nice if slic3r offered a way to have the code slow down around "jerky" movements like small infills - I don't want to slow down the whole print because some parts have narrow infill! But I resolved that (to a large extent) with a later fix...
  • X tilt slop
    So, the extruder unit is heavy. And it's all held together with 2 bars on either side and 2 bars as a basic tension mechanism to keep the unbalanced X-carriage straight most of the time (it's very unbalanced - all the weight of the thing constantly induces the carriage to twist - only thing keeping it from doing so is the 2-bar setup). So what happens when X jerks around or the filament tugs on the extruder? The tension placed on the carriage shifts and X tends to tilt, affecting the print height in tiny variations (and hence, causing sloppy perimeter appearance). Haven't been able to fix this one in hardware, but I did find a way to compensate in software with the same fix as above (later).
    It would be nice if the extruder motor and gear were mounted center along the X carriage so the weight would be balanced, but of course that would require a redesign of the extruder itself (maybe some playtime in OpenSCAD with the extruder design from source?). As it is now, it could probably be band-aided by creating a sort of "extension" for the stepper to be mounted further back and have its shaft extended out to the small gear of the current extruder using an extension shaft of some kind. That would at least balance the heaviest part, and resolve the X-tilt problem for the most part.
  • Parts curling, breaking off board, etc
    Oh god, this has been the biggest thorn in my side since day 1. Even with Kapton tape and Gates of Hell bed/nozzle temperature (up to 110C bed/230C nozzle), I always seem to have trouble with multiple-part prints. From the start I intended to be able to print a full platform of Printrbot parts simultaneously (I'm really quite a low-income geek, was hoping to make up some of the 'bot cost by printing and selling parts on eBay)... but I couldn't even get 4 parts to print without at least 1 part breaking off and ruining the print (or, as I learned, stop at the very beginning of a new layer -> pause -> reverse 2mm filament -> run M114 -> see the current Z position -> re-slice the file with everything in exact same positions but the broken part removed -> edit gcode -> find the current Z position in the file -> cut off everything before "G92 E0" just above the line with the current Z position -> save -> open modified file -> resume print now with the broken part excluded from the remaining job).
    FINALLY, I found a godsend-status solution in slic3r: brim. It adds a suction-cup like extension to the perimeters of the first layer, firmly gluing those obstinate parts to the bed. It worked so well with the above full-platform print, after cooling on the bed overnight, I spent a good 3 minutes with a razor blade trying to snap the build off the board without taking the whole 'bot with it! Also learned the extraordinary importance of de-greasing the bed before printing with acetone. That made a night-and-day difference in stickiness. With a 5mm brim and a razor blade handy after each print, I believe I've 100% resolved all "stick"-related (curl, breaking) problems.
  • Jerking around tight corners
    The biggest problem to date - the firmware and slic3r's blatant disregard for how it takes the "perimeter speed" and "infill speed" settings and blindly applies them to all motions large and small. So if I have a part that's got a long, square outside perimeter and a gear-tooth inside (maybe a Spirograph?), I would have to trade-off the outer perimeter speed - which the filament and heater are more than capable of ripping through quickly - for not slopping the ever-loving crap out of the gear-teeth on the inside with excessively rapid jerking movements. So I'd have to slow down the whole print such that it'll take those teeth slowly. More than tripling the print time it would otherwise be able to achieve if it were smart enough to slow down around the teeth and take the square perimeters fast. Fortunately, I seem to have found a solution in the Marlin firmware.
    First, "acceleration", code M204. The value seems a bit arbitrary - expressed in m/s^2 (meters/sec per second) - since the numbers have a somewhat unpredictable effect on the motors. Sometimes it will slow down around tight corners, others it will slow down around long moves but still jerk around small movements. The syntax is S=printing acceleration, T=retract (extruder?) acceleration. T is unclear to me - I don't know what constitutes "retracting" or how it knows which extruder to apply to if it only applies to the extruder stepper(s). But "S" is the acceleration for motions. I set mine to 500 for a good smoothing effect, and 350 to add a silky smooth (but significantly slower) motion to all axes.
    M201 is a more refined command that sets the "max" (huh?) acceleration for each axis - X, Y, Z, and E. So if you only want to smooth Y (as I do), I'd skip doing the M204 and go with a M201 Y350 instead, and only smooth Y.
    But... you may notice... it ignores acceleration completely at the most critical time! It still whacks carelessly around short segments, what the F? Well, that is a job for "XY Jerk". Sure enough, it has a separate, unique and independent speed setting for segments it detects are short and jerky. "M205" (advanced settings), "X" (XY Jerk max mm/sec). Default in Marlin is 35mm/sec for jerky movements. WAY out of line for such a bulky platform! I run "M205 X10" to limit small segments to 10mm/sec, and ALL my sloppy segments disappeared. It now does exactly as I dreamed: accelerates movements to smooth (eliminate) vibration, motor skipping, and the Y belt jerking loose... AND it will run a fast straight line while slowing down for jerky movements. Literally like a dream.
    Bonus tip: "M220" (build speed multiplier) is a real-time speed adjuster. S=speed in percent. So, "M220 S75" will slow all movements to 75% of the specified speeds in the G-code, if you find it's running too fast. Or if you want to see how far you can push your speed settings, something like "M220 S150" will run at 150% speed (to the coded limits of your motors - which by default there are none, so watch for skipping, ESPECIALLY in the extruder, which mine skips steps over 40mm/s).
    On an axis-by-axis basis you can set hard-coded limits using "M203" (maximum sustainable feed rate). So to prevent 150% speed from over-speeding the extruder (it will slow down the movements if they make E run too fast, but otherwise will still run 150%) - you can run "M203 E40" to limit the E-stepper to 40mm/s.
    By default mine didn't come with EEPROM save/read support, but I compiled it in when I hard-coded all my changes. For you, you could either enter these in Pronterface in real-time, or add them to the "start GCode". To check if you have EEPROM support, run "M503" to display the current running parameters - they're blanked out if EEPROM support wasn't enabled in your firmware. If you have it, you'll see a nice list of all the current running tweaks mentioned above (all the defaults and anything you changed). "M500" will save it so it loads every time you start the machine. "M501" reverts to the stored settings or the defaults (as it is when it powers up, just without restarting). "M502" reverts to factory defaults for run-time, which you can save with M500 if desired. However, if you DON'T have EEPROM support, you'll just get no response from M503, and all the other M50x commands will do nothing.
  • Bonus: improve hot-end temperature accuracy with Marlin PID autotune and auto-calibration
    Here's an easy one. Apparently Marlin uses some slick algorithms for keeping the hot-end at a steady temperature despite the long response time in the thermal feedback loop (that is, it heats... temperature doesn't rise for 5 seconds... it shuts the heater off... temperature keeps rising for another 5-10 seconds... etc). You can enable and tune this automatically, but be warned: the firmware has a very sloppy implementation of the routine and, by default (I made changes in mine to stop autotune after "S" loops, or when another command is received), it locks up the whole machine when you start tuning it - requiring a board reset to get back to work again. It shows you the results after every loop but it loops endlessly until reset. You can run the PID auto-calibration with the "M303" command - syntax is S=temperature for calibration. So I would do "M303 S220" to calibrate for my common running temperature, 220C. Start it while cold, and it'll run through several power settings on the head and see which one works best to keep it in tune. When it's done with a loop, it will display a list of 3 different "profiles" to choose from - I choose the "some overshoot" profile. Note the Kp, Ki, and Kd values after letting it run about 5 loops (it will play with the temperature and dump a list of parameters 5 times). Take those numbers and add them to your Start GCode section, using the M301 command. Its syntax is just P, I, D, for the Kp, Ki, and Kd values it gave you during auto-calibration. Then, reset your Printrboard (disconnect Pronterface first!) by hitting the reset button or unplugging USB and switching off/on the power supply. Now you've got customized temperature control for ever so slightly nicer prints and less extruder drama. :D

Hope this helps some folks a lot. Maybe this should be split/copied to multiple topics for searchability. Maybe added to the Wiki. I dunno... I can write good documentation but I'm bad at integrating it into the doc structures (Wiki, etc). Do with it what you like, whatever will help people out.

Now, who wants to see that STL and the config.ini? :lol:
  • 3

You do not have the required permissions to view the files attached to this post.
User avatar
FalconFour
I'm an end stop.
 
Posts: 58
Joined: 2012-Jun-Wed-03-Jun
Location: Fresno, CA
Reputation: 5

Full Printrbot plate. Bet you'd love to see slic3r settings.

Sponsor

Sponsor
 

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby mdfast1 » 2012-Aug-Mon-16-Aug

Awesome post. I've read it all once. Now I need to go home and digest it all and read it again. Printing some feet tonight, and working on Bed Leveling.
  • 0

Printrbot Simple Metal - RED -
UBIS 13

Printrbot LC+ - Operational

Printrbot Original- Donated
GT2 - 20 Tooth Pulleys
GT2 Belts
Brass Threaded Rods
Spur Gears from drawcut
.35mm nozzle
User avatar
mdfast1
Layer 650 of 1234
 
Posts: 669
Joined: 2012-Feb-Thu-12-Feb
Reputation: 12

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby plexus » 2012-Aug-Mon-18-Aug

Well, either you are having a lot of problems that I'm not having, or you are being much more picky than I am. not that being picky is a problem. i can see with 3D printing that there is a lot of potential quality that can be achieved. maybe I just feel ok about the results I am getting. yes i have some little problems here and there but this is exceeding my expectations. so im happy. and i can see more room for improvement of the hardware that will likely result in better prints. but at the same time i understand that Brook's goal was to make (one of) the cheapest 3D printers on the market at the moment. I think he achieved that with some sacrifice in quality.

But the info in your post is fantastic and offers a whole other detailed level of tweeking possibilities! thanks for that! its very useful and interesting and will help anyone who is trying to squeeze the most out of their printrbots.
  • 0

User avatar
plexus
Site Admin
 
Posts: 2117
Joined: 2011-Dec-Mon-16-Dec
Location: Toronto, Canada
Reputation: 95

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby FalconFour » 2012-Aug-Tue-00-Aug

Still playing around with odds and ends. At this point I believe the M205 X10 command may be the single most valuable undiscovered (or at least, under-implemented) Marlin firmware feature in existence. It really does clean the hell out of sloppy corners, and I was able to print 2 copies of the large extruder gear (the unbelievably complex one) - one to replace the worn gear on my own printer causing E-slop, and one for sale with the part set. In the same batch I also ran 2 10-tooth pulleys and one modified shortened one as a spare for my Y axis (motor shaft doesn't reach the full length of the standard one, so to get the screw-down to fit, I shortened the model). All 6 parts printed with minimal, repairable issues... and these were all gears. Printed in about an hour. Near the end as it was doing the tops of the pulleys, I slowed it down to 25% speed and the quality shot through the roof to what I could only describe as "near divine quality":

2012-08-06 22.05.27.jpg


Now, time to try some artistic work. :mrgreen:

Oh, and here is that elusive STL file and slic3r config for the full part plate.
printrbot-parts-layout.png

Includes:
  • From the official Printrbot "thing":
    • pb-BeltClip-x2.stl (2 total - split, flipped, rotated, and rearranged as needed)
    • pb-Y-Bearing-Guide.stl
    • pb-Base-v2.stl
    • Base-w-endstop-mnt.stl
    • pb-Xtruder-Mount-v2.stl
    • pb-Y-BarEnd-x4.stl (4 total - split, flipped, rotated, rearranged)
    • E-Body.stl (recently updated with other parts)
    • E-Idler.stl (also updated)
    • x-motor-v3.stl
    • pb-X-Idler.stl (that one almost slipped through the cracks, required a full re-do to fit into the plate!)
    • pb-X-Carriage-v2.stl (modified with a warp-resistant and short-hotend compatible snip of the bottom tip of the bearing sleeve, so it can get closer to the bed. still not updated with endstop screw holder, wth Brook? I don't use the X/Y endstops so I forgot to add the adapter part)
    • pb1-y-stop-fix.stl (honestly don't know why this one is in here. going to remove it before the final publish to Thingiverse... probably replace w/ X endstop)
  • 2x Z couplers with independent screw plates for motor and rod parts. (big smack-on-the-forehead "duh" on that one, why shouldn't they be separate parts? Also split, rotated, and rearranged the components to fit.)
... 22 parts total, all in one print. Yay-uh.

(Another side note: yeah, I'm the only person I know that has a 3D printer, and this is also my first 3D printer... I'm just a geek on a low income, so I'm the kind of guy to take a Corolla and turn it into a Cadillac - well, if I could afford a new car, haha.)
  • 1

You do not have the required permissions to view the files attached to this post.
User avatar
FalconFour
I'm an end stop.
 
Posts: 58
Joined: 2012-Jun-Wed-03-Jun
Location: Fresno, CA
Reputation: 5

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby Mooselake » 2012-Aug-Tue-14-Aug

Impressive post, will study it in much greater detail when I get some spare time!

Check out drawcut's spur extruder gear. He's a professional gear designer (I'm an amateur low knowledge gear dinker with Gearotic Motion) and has given the extruder gears the attention they deserved. They're on my list of things to print.

Kirk
  • 0

Modified KickStarter Classic Plus 7/2012
KS Thingybot Delta Pro 10/31/16
User avatar
Mooselake
My next printer is...
 
Posts: 3614
Joined: 2011-Dec-Tue-23-Dec
Location: Moose Swamp, Upper Michigan, USA
Reputation: 177

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby drawcut » 2012-Aug-Tue-18-Aug

Very nice work. I'll have to take some time to digest all this, for sure. I have played a bit with accelerations and jerk after another member here posted about it in their blog (Lincomatic, I think). But I really haven't had much Printrbot time lately between work and multiple car issues. Hopefully I can get back to PB things soon.

Edit: Thanks for the shout out Mooselake. There'll be something extra in your check this week. ;)
  • 0

Printrbot original. Major mods: Y axis extended to ~8", Z extended to ~8.5". 5mm SS Z threaded rods w/ flex couplings. E3D hotend. Purchased Acetel gears. Glass bed with Elmers' purple glue stick for most prints. Top of Z axis rods have added cross structure similar to a Prusia i3.
User avatar
drawcut
ACME leadscrew
 
Posts: 386
Joined: 2012-Mar-Mon-19-Mar
Location: USA - Wisconsin
Reputation: 16

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby plexus » 2012-Aug-Wed-00-Aug

@FalconFour can we see a closer up better quality shot of that big herringbone gear?
  • 0

User avatar
plexus
Site Admin
 
Posts: 2117
Joined: 2011-Dec-Mon-16-Dec
Location: Toronto, Canada
Reputation: 95

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby FalconFour » 2012-Aug-Wed-14-Aug

Here.

I'm still in "watching the machine throughout the whole print... GO FASTER, DAMNIT!" mode, so print quality suffers for speed. Once I find a way to place the printer apart from me, maybe network it in my bedroom's attic-space-accessible (vented) closet, I'll be able to slow it way down and get better quality. But this is 2 gears and 4 pulleys printed in under an hour with the modified "jerk" and acceleration parameters.

Photo Aug 08, 8 58 00 AM.jpg

(best quality I could fit under 256kb for the forum. BTW, why is that limit so low? Web host trouble? Storage is dirt cheap these days, it'd be nice to be able to post larger attachments, like larger STLs as in the build above...)

I 100% plan on printing that spur extruder gear. I love that it foregoes fancy curly-que circles in the complex back-pattern for a simple "spoke" design. Makes a damn ton more sense and looks a WHOLE lot quicker and easier to print. And the gear design makes a lot more sense too. Can't wait to try it.
  • 0

You do not have the required permissions to view the files attached to this post.
User avatar
FalconFour
I'm an end stop.
 
Posts: 58
Joined: 2012-Jun-Wed-03-Jun
Location: Fresno, CA
Reputation: 5

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby plexus » 2012-Aug-Wed-15-Aug

FalconFour wrote:Here. [snip]


Why are printing with such coarse layers and fill? just curious. i've been going the other way and trying to print as finely as possible. i can't really do 0.1mm layers without aberrations. but 0.2mm layers are going ok. i am going to upgrade the quality of the parts on the Z and that should help.
  • 0

User avatar
plexus
Site Admin
 
Posts: 2117
Joined: 2011-Dec-Mon-16-Dec
Location: Toronto, Canada
Reputation: 95

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby FalconFour » 2012-Aug-Wed-19-Aug

I use 0.5mm layers, because of the 0.6mm nozzle and the "OH MY GOD HURRY UP ALREADY" factor. I've been nervous about peeling/curling parts for so long it just became a habit. Plus, whenever I get the layers smaller, it seems to start slopping, extruding too much plastic (though my E-steps are calibrated and the thickness is set - I seem to have to enter 2.97mm into slic3r, for example, for what I measure as 2.87mm filament). And when too much is put down, it leaves bumps hanging up that the nozzle crashes into while traveling. Thicker layers resolve that, it seems...
  • 0

User avatar
FalconFour
I'm an end stop.
 
Posts: 58
Joined: 2012-Jun-Wed-03-Jun
Location: Fresno, CA
Reputation: 5

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby SolidVis » 2012-Aug-Tue-19-Aug

FalconFour wrote:... "OH MY GOD HURRY UP ALREADY" factor. I've been nervous about peeling/curling parts for so long it just became a habit. Plus, whenever I get the layers smaller, it seems to start slopping, extruding too much plastic (though my E-steps are calibrated and the thickness is set - I seem to have to enter 2.97mm into slic3r, for example, for what I measure as 2.87mm filament). And when too much is put down, it leaves bumps hanging up that the nozzle crashes into while traveling. Thicker layers resolve that, it seems...


First, thanks for the experience you are sharing!
Second, me too, I observed these vibrations on the base and extruder when printing above 20mm/s. If I have time I will balance the extruder (symetrical) by redesigning the upper portion of the printer.
http://www.flickr.com/photos/solidvis/7784799532/in/set-72157629898373934
Last for now, it took me 5h to print that gear and I replaced all of them with higher res ones to improve the quality. Especially the extruder gear. Now it turns very smoothly and feeds smoothly.

p.s. can you spell "sugar water". Sticks excellent. ;-)
  • 0

SolidVis
I'm an end stop.
 
Posts: 69
Joined: 2012-Jun-Sun-18-Jun
Location: QC, CANADA
Reputation: 0

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby augspark » 2012-Sep-Thu-22-Sep

This thread is awesome, thanks for all this great info everyone. I am just barely getting started myself (maybe only on my 6th print), so I have this post open while I try to figure everything out. So far, I have a lot of the same issues as the OP - but overall I am quite pleased. SolidVis - could you tell us a little bit more about how you have things set up to produce such excellent quality? That picture of the big gear is amazing! Is that the natural ABS that shipped with the printrbot?
  • 0

Auginator
Kickstarter backer #1212
Printrbot Classic • Wooden Extruder • Spur gear for PB Wooden Extruder (http://www.thingiverse.com/thing:29366)
User avatar
augspark
Loading filament...
 
Posts: 14
Joined: 2012-Jan-Wed-13-Jan
Location: New York City
Reputation: 0

Re: Full Printrbot plate. Bet you'd love to see slic3r setti

Postby SolidVis » 2012-Sep-Fri-17-Sep

augspark wrote:... SolidVis - could you tell us a little bit more about how you have things set up to produce such excellent quality? That picture of the big gear is amazing! Is that the natural ABS that shipped with the printrbot?

It is the ABS shipped with the printrbot indeed. But you have to:

1. Tighten well the printer belts. Any loose belt will bring the printer in resonance at (higher) speeds.
2. Lubricate well the shafts.
3. With M201 X350 Y350 and M205 X15 find the maximum speed you can print without having the bed/printer to resonate. I started from 30 mm/s and went down to 15mm/s.
3. Get rid of that ugly printed gear http://www.flickr.com/photos/solidvis/7283517120/in/set-72157629898373934 by printing a new one.
4. Replace it and print the smaller one.
5. Replace the smaller one.
6. Replace the nozzle with 0.25 and print the gears again by keeping the temperature down to 214 oC.
It took me 6h to print it. The smoothness/roughness of the surface will change if you have loose parts and resonance. Hold down gently for a second the bed/glass or the wooden head of the extruder with fingers (BUT DO NOT BURN YOURSELF) and listen for the buzzing to stop. If yes-tighten somehow. This printer has a lot of weak points, but if you want to learn, you have to start from somewhere.

Sure you can skip some steps - better ask someone to print you a set of higher quality gears.

With PLA I was able to speed up to 45...50 mm/s and still have relatively good results. LET THE PRINTING BEGINS!
  • 0

SolidVis
I'm an end stop.
 
Posts: 69
Joined: 2012-Jun-Sun-18-Jun
Location: QC, CANADA
Reputation: 0


Return to Showcase!

Who is online

Users browsing this forum: No registered users and 1 guest