Pocket Printer Upgrade - hybrid roller slides
Derived from
Description
The print volume is 120mm W x 100mm D x 65mm H. This could be expanded a little by getting longer extrusions.
Pocket Printer on Marlin - youtu.be/lAwCmbKS-yQ
I have comb activated, so it travels around the cord cutout on this iPhone dock at a very rapid travel speed.
Watch this printer and my Aluminum Mendel thingiverse.com/thing:16076 print side by side here youtu.be/JysNv3Hi67k
I posted a video at youtube.com/watch?v=4tpOokI6_Z0
I wanted to see if it could print vertically. It is printing a stretchy bracelet - thingiverse.com/thing:13505.
I hope this inspires others to expand on this design.
Instructions
This is printed in PLA. Drilling and tapping small parts is necessary to build this project. A drill press is nice, but it could be done by hand. The sliding surfaces fit well with my printer settings, but you can use a new, sharp file to remove a little material if necessary. Take your time with fitting the slides, a good fit without too much friction is needed. The x and y inner slides can be adjusted externally for a perfect fit. I use a very light film of silicone lube on the slides. I tried many different designs on the slides, but this one really works the best.
The exploded view and BOM should provide enough info to make this project. I zipped the stl files because there are 22 total. I also zipped the entire Inventor project. I did not provide all the limit switch or power supply and switch mounts.
4/24/2012
I upgraded the slides to a hybrid roller version that is easy to print and assemble and has low friction and good precision. All the stl files are in the zipped folder. I was unsuccessful in simplifying the pulley arrangement by combining x and y movement, so I added a small, lightweight stepper for the x axis. This motor weighs only 4.9 ounces, compared to 12 for the standard steppers, so moving mass is still low. The motor should be run near its maximum current of 500 mA, so I recommend making the x motor mount out of aluminum plate to dissipate the heat. A drawing with dimensions is included in the zipped file. I also upgraded to the Misumi belts and pulleys. I think the smooth action is worth the extra cost.
4/27/2012
An exploded view drawing of the hybrid roller slide has been added below. The BOM is on the exploded view drawing.
You must be logged in to post a comment.
Dave - FYI, I just completed my first successful print on my Pocket Printer derivative. I'll post the details soon.
I'm having two issues with the z-axis.
1. The lead screw nut pulls out of its seated position slightly while descending, making it impossible to get a consistent z-axis home. I'm seeing about a half-millimeter (max) play due to the nut unseating, making calibration and subsequent printing impossible. If I tighten the self-tapping nut hold-down screw enough to keep the nut seated, the nut hold-down screw puts a torque on the nut, orthogonal to the lead screw z-axis, and the lead screw loads too heavily due to friction. I very nice break, BTW. Occasionally, I find a M8 1.25 nut that is so far out of tolerance, it makes a poor candidate for the lead screw nut. This is not the case. I hand-selected the nut. I am considering using JB Weld epoxy to permanently bind the nut in the seated position, and not use the nut hold-down screw. I would hold the nut in place using a M8 1.25 bolt and temporary opposing nut, to make reasonably sure the lead screw nut stays on axis while the epoxy sets and cures.
2. The linear slides want to "stick and walk" slightly during both ascending and descending the x-axis, essentially causing the extruder to wobble roll about its y-axis -- if it had a y-axis. This also impedes x-axis homing and subsequent calibration. There is some play in the slides, in yaw, pitch, and roll -- not a lot, but I think enough to prevent calibration. I have yet to prove the walk is a calibration show stopper, because problem 1 above, impedes accurate measurement. I applied a lite dose of 3-in-one oil, but it does not help. I experimented with different loads, both positive and negative. The wobble is not caused by the weight or moment of the extruder and extruder arm.
What did you do to fix these problems? How do you increase the size of the slides to take out the wobble? I considered using JB Weld (or other brand) epoxy to add dimension to PLA bearing surfaces, then filing them to fit more tightly, but the coefficient of friction between epoxy and the Aluminum extrusion is prohibitive. I used waxed dental floss on the x and y-axis to solve a similar problem, but I think the geometry of the fixed z-axis slides makes it impossible to wrap dental floss.
Help me Obi Wan. You're my only hope? :-)
1. Maybe a second screw opposite the first would help.
2. Check that the z inner slides are snug. Also check that there is no wobble in your threaded rod or coupling.
3. You might try a better quality bearing, or use two thrust bearings. To get started, you should be able to push the z down when in the home position. It is all up from there, so backlash shouldn't be an issue.
Good Luck
I finished building my Pocket Printer a few days ago. I am calibrating it now. I have a question about xy axis scaling. Here is my thinking.
1. Both x and y steppers are 200 steps/revolution.
2. Both x and y pulleys are 15 mm diameter.
If I did the math correctly, I come up with 200 step per revolution divided by (Pi x 15 mm) pulley circumference ~= 4.2551 step/mm
That is a HUGE difference between the default Marlin scale of:
DEFAULT_AXIS_STEPS_PER_UNIT {78.7402,78.7402,200*8/3,760*1.1}
To confuse matters more, when I advance "10" units from Printer Interface, the Pocket Printer responds with an actual displacement of about 12mm, which is pretty close to what it should be, suggesting my calculations and reasoning are way off.
What am I missing?
I will be installing the xyz-axis optical limit switches soon. I am assuming a right-hand coordinate system, where the optical limits switches close at position zero.
Here are my assumptions, Let the y-axis points north and south, east is on the side of the y-axis motor, west on the z-axis motor side.
x-axis - positive counts toward the west, zero origin on the far east
y-axis - positive counts toward the south, zero origin on the far north, closest to the z
-axis vertical extrusion.
z-axis - positive counts upward toward the ceiling, lowest relative point is zero origin.
Does this seem reasonable?
FYI, I temporarily scrapped my optical sensors, and replaced them with smaller mechanical micro switches. I was running into too many space constraints with the optical switches. The circuit boards are too large. From what see from others, this is a common issue.
This is a little confusing. Let me explain this way: When at the origin, the extruder is in the SW corner of the build platform according to your terminology, at the extreme left of the platform. The build platform is to the right and towards the z axis. Positive x the platform moves left, the extruder moving to the right relative to the table. Positive y the platform moves toward you away from the z axis, the extruder moving towards the rear of the platform.
What distance should I raise the heated aluminum print bed plate from the PLA x-axis slider plate? I'm concerned the bed heater might warp the PLA x-axis plate.
Also, I'm considering using a glass mirror plate over the aluminum plate. Thoughts?
Use at least 1" long screws to attach it. It must be high enough so that the extruder will reach the bed when z is at minimum. Heat will not be a problem at this height, at least with PLA bed temps. I have very good luck with the aluminum/mirror print bed. Be sure to ground the aluminum plate if using a 120 V heater.
I modified the x-axis slide assembly. The bearings pinch the y-axis rail far too tightly, and when I did manage to get the bearing posts to flex enough, the bearing posts that share the rail guides lift and rotate the guides enough to significantly bind the slider. There is a larger bending moment on these posts, because they are longer than the opposite posts. The opposite posts have to be shorter to clear the x-axis motor mount. I think the longer posts will always bend farther sooner than the shorter posts. I drilled out all four bearing fastener holes, and elongated the bearing mounting holes on the short posts, increasing the spacing between opposing bearings so that the overall pinching force is just enough to keep positive pressure against the rail. I used longer fasteners with locking nuts on the x-slide mount. It's temporary, until I can print an improvement.
I'm curious if anyone else has run into this, and what the fix was.
The y-axis idler pulley appears to be about 5mm too high. It will not align with the y-axis belt. The 5mm offset is too much to cut from the end support, so I'm going to drill new blind extrusion holes in the end support, and drop the end support enough to bring the pulley back in line. I'll have to add 5mm pads to the rear feet to level the printer. Has anyone else run into this, or did I make a mistake in the assembly?
I started aligning the z-axis tonight. Do any of you have a favorite technique to align the z-axis? I'm still experimenting, but I made the following "obvious" observations.
Errors are most noticeable when the z-axiz slide travels toward the highest and lowest positions.
The fastener holes in the PLA parts are not perfectly located. Some reaming is needed I think.
The flex motor coupling hides errors. I think it's a good idea to pre-align with the threaded rod float
ing or hovering over the coupling bore.
The z-motor shaft and the z-axis threaded rod and bearing receiver are mounted on different rails, and the rails are neither coplanar nor perpendicular in practice. How do we deal with this? I squared things up during assembly the best I could, but I wonder i
f it's enough.
The z-axis motor mount may not be flat .
My threaded rod is black steel with sharp threads. I cleaned up the threads with a die. Because only threads contact the inside do the flex coupling, it slips easily. Some clutch action might be desirable to prevent disasters, should binding
occur, but since we do not have position feedback, slippage could cause z-axiz errors or even print head crashes, as the advance fails to match the print. So, how do we prevent z-axis slippage? Thread filler?
How do we measure the threaded rod parallel and perpendicular planes? When they are severe
ly off, I can feel the extra drag at min and max height, while turning the z-axis motor with my fingers, but that's not a very good measurement.
Ideas?
To align the z axis I like to get it in about the right position by measuring. Run the z all the way to the top, loosen the screws on the bearing bracket/handle and move to position where z screw moves freely. Ream holes in handle if nececcary. Run z all the way to bottom and adjust motor. Repeat if necessary. I have stopped using the spiral cut aluminum couplings because the last 5 I bought had the shaft hole bored out of alignment. They all wobbled when installed on the motor without the leadscrew connected. I use either the resin or alumunum couplers from Misumi with good results. They have set screws which grip the leadscrew better than the clamp style.
Hi All.
I'm making progress building my new Pocket Printer, but I'm stuck on a problem with x-axis slide binding throughout the range of motion. The x-slide mount pinches all four bearing on the sides of the 20-mm Misumi extrusion. It looks correct, but I think the pinching force is too great. Nothing els
e is binding that I can see - certainly not the PLA rail guides. Loosening the guides make no difference. I experimented with a single rail with same binding problem.
I see no easy way to reduce the bearing pinch force, assuming that's the problem. I suppose I could fill the bearing fastener holes,
and redrill and tap, but I can see in my mind's eye, the drill going where it wants to, with or without my drill press. Perhaps I could remove the bearings on one side, and kludge a fixture to reapply an adjustable reduced bearing pinch force.
Any suggestions will be appreciated.
Thanks in advanc
e,
Ron
I have a question about the horizontal extrusion attached to the back side of the z-axis extrusion. I'm not sure what to call the backside extrusion, but the rear feet are attached to it, so l'll call it the "apron" for now. Feel free to correct me.
How does the apron attach to the z-axis extrusion? In other words, how do we make a perpendicular connection with extrusion?
David, your vertical printing experiment at (youtube.com/watch?v=4tpOokI6_Z... looks interesting. I have a future need to print unsupported horizontal surfaces.
What did the print quality look like up close? Was the bracelet functional?
We often dealt with gravity and print nozzle misdirection in the 2D printer world. We compensated with time-of-flight, drop mass, drop viscosity, firing time, etc.
In our new 3D world, do you think we could compensate fo
r gravitational trajectory by managing viscosity and x-y position, or do we need a more clever solution?
I did't spend too much time wth vertical printing. It was more just wondering if it would work. The print quality was equal to horizontal. Since the extruded plastic is adhering directly to a solid layer, I don't think gravity plays much of a role. It later occured to me to try a bridge test, but I have never taken the time to do it.
I now have most of the parts to build my new Pocket Printer hybrid.
How exactly does the z-axis extrusion fasten to the base (y-axis) extrusion? (I have the coordinates right, correct?). It looks like a butt joint, which implies I must use brackets. I don't see any brackets in the photos or drawings. It looks like Entropy85 used brackets in his derivative. I would
think we need brackets both fore and aft, to insure a stable joint.
Can you please advise?
Yesterday, I finished ordering the last of the parts for my new Portable Printer, including the controller and power supply, plus selected spare items. I just have to click my shopping cart "ship" button for the Kapton tape and a set of metric tap and die . This really is too much fun.
BTW, I was fortunate enough to get the two Raspberry Pi computers I ordered. The "Pi" is a business card sized computer that can run Debian Linux. I will eventually run my 3D printers using Pi computers.
I have a few questions about the print bed.
It looks like many people apply Kapton tape over the surface of the print bed. Kapton tape is a technological marvel, and justifiably expensive. Do you guys have a favorite vendor that delivers Kapton tape at a great price?
Do you ever have issues at the seam, where the wide strips of tape meet?
For example, do you get undesirable print artifacts when the printed object touches the tape seam, or is it negligible?
Some people suggest using glass plate over the surface of the print bed. I'm not sure that is a good idea, as the glass may not conduct heat very well.
Your thoughts would be app
reciated.
Kapton on an aluminum bed with a silicone heat pad underneath has worked well for me. It is expensive, and easily damaged. The seams do leave small mark. My vendor has always been Ultimachine. I saw a suggestion a few weeks ago for printing on a mirror. It seemed like the metallic coating could help heat transfer between the aluminum plate and the glass. A local hardware store cut me a few pieces, but I never really tried it until tonight. The results of a few small print jobs were quite good, so I may use this for a while. It is a perfect durable surface, and with the mirror you can really see the first layer as it is going down. I clipped it to my bed with some binder clips. I have also tried a piece of plexiglass from a storm door. The part stuck so hard I could hardly get it off. Maybe it would work if it were not heated. The glass definitely adds mass, so we'll see how it works on some faster prints.
I would like to order additional Misumi die cast brackets to keep in my inventory, mainly for test fixtures and other kludges. An assortment might be nice. For starters, what Misumi part numbers do you guys use for the Pocket Printer rear support? The term "die cast brackets" gets zero hits on the Misumi site.
Extrusions are too cool. I thought up a hundred uses last night when should have been sleeping. :)
I have some questions about filament.
I plan to use the recommended Prusa Stepper Plastruder Kit for 1.75mm filament. It appears to me that narrower 1.75mm filament might provide smaller streams/drops per step, resulting in better resolution. I helped design hp Color DeskJets, and small drop mass, among other things, helped us produce
high resolution photo quality prints. Your thoughts on 1.75mm filament would be welcome. Why do you guys prefer 1.75mm filament?
Do you ever switch to thicker 3mm filament? If so, why do you use wider filament, and how do you configure the 1.75mm extruder to accept 3mm filament? Or do you?
Who is
your preferred vendor for filament?
When I started printing, I chose to go with 1.75 mm filament for the very reasons you stated. I have never tried 3 mm, so I can't speak to how it works. I'm sure people are getting good quality prints with it. If you want to make your own extruder, my guess is that it is easier to hob the bolt for the larger filament. I have always gotten my filament from Ultimachine.
Here are several BOM questions for a Pocket Printer I'm building. I want to extend the length of each axis to create a larger print envelope. I'm an electronic/firmware engineer, not a mechanical engineer, so please bear with me :)
I how can I locate the end-stop datum points on each axis? Once I know the end-stops, I can verify the current print envelope, and calculate my desired envelope, given the longer extrusions.
BTW, I'm thinking of adding approximately 50mm to x, y, and z axis.
I assume the cantilever beam (part 23
in the exploded frame) allows the calibrated placement of the print head (extruder) nozzle to rest over the center of xy bed, when both x and y axis are at 50% full travel. How much longer should I extend the cantilever beam to guarantee nozzle placement at the extended xy bed center?. I guess I ne
ed datum for part 23 as well.
Now for the fun part - the belts. Now I studied statics and dynamics under Dr. Meriam, the man who wrote the text books, and I did pretty well. But my point is, that was a long time ago.
The y-belt looks easy. Do I simply add twice the extended y-axis length to the be
lt circumference?
The x-belt looks complicated. Would I add four times the x-axis length to the belt circumference?
I'm comfortable with firmware, so after I get my Pocket Printer working, I would like to try the alternative "more efficient design" depicted in the pulley arrangement. Could you sim
ilarly help me with the alternative pulley lengths? The Y-axis looks the same in both configurations. True? The X-belt looks very different. Not sure where to start, and I pulley block dimensions are a little fuzzy to me.
Thanks in advance,
Ron
If you wait until tmw, I will post my pocket printer derivative. I've extended the frame to fit a 200x200 Prusa PCB print bed.
Ron,
The measured travel on my current build with the extrusion lengths listed in the BOM is: x = 120mm, y = 100mm, and z = 70mm. The extra length you get should add directly to the travel on each axis. The beam for the extruder will have to be extended the same amount as the y axis. The extruder plate allows quite a bit of adjustment for fine tuning.
I never got the "more efficient design" fully implemented, so I strongly suggest that you use the hybrid roller slide upgrade, and my x and y belts are about 465mm and 525mm respectively. If you are adding 50mm to each axis, each belt will be 100mm longer. If you buy the suggested Misumi GBN9942GT-60 belts, they are 994mm long and are cut to length required. I got both my belts out of one, but 994mm is the longest in this series, so you will need two. For your larger build area I also suggest using the Pololu #1209 stepper.
I am anxiously waiting for the new BOM! lol I purchased all of the extrusions so far. I'm making my build area a little larger so I can take advantage of the Prusa PCB HBP. Is the X axis motor a NEMA 17 or 14? If 17, what is the torque on a small motor like that?
I'll work on the documentation this weekend. The x motor is a NEMA 14, from Pololu - #1208. It is 28 mm high with 14 oz-in holding torque (compared to 44 oz-in for the standard motors). You might want to use the Pololu #1209 which is 36 mm high with 20 oz-in holding torque with your larger platform. That motor weighs 1.4 oz more, but is still about half the weight of the standard motor. You would have to increase the height of the feet 10mm to have clearance for the larger motor.
FYI youtube lets you rotate videos. When logged in, go to your video page with the sideways video (when you are logged in)"
http://www.youtube.com/watch?v...
And click "enhancements" at the top. There is an option to rotate!
This is really a professional job. BOM and drawing is top shelf. We are going to use this as a project at the HS STEM lad where I am a mentor. Well done! Looking forward to the CNC mill design.
Is it possible to save this as a file which can be opened in Inventor 2008?
Thank you
:)
Would a long drive shaft and a sliding pulley (dragged with the X motion) work for the Y?
I like your idea of dependent coordinates and handling that in the firmware. It should be a trivial change. For linear motion, you could just hack the positions with a post process of the gcode:
#!/usr/bin/perl
while(
&
lt;
&
gt;)
{
if(/^(.+\s+X)(-?\d+(\.\d*)?)\s+Y(-?\d+(\.\d*)?)(\s+.*)/) {
print("$1$2 Y".($4 - $2)."$6\n");
} else {
print $_;
}
}
:)
A post process of the gcode sounds like a good idea. I am afraid I do not know how to implement your script. Is there a fairly simple explanation of how to use it? I can do a little programming in the Arduino environment, but that's about it. Would it be possible to translate your script into C and insert it into Marlin or Sprinter? I nearly have it working in Sprinter, but it X drifts off slightly on each layer so it doesn't print quite vertical. Maybe its a rounding error when it calculates its position.
I love the design of this, looking at getting a 3D printer for my schools engineering workshop and this works in a similar way to our CNC milling machine. Would be great as the students can see how this works which isn't possible on the mill and get an understanding of both mechanically.
You know, if you could add an automated build platform to this it would be the basis for a phenomenal little manufacturing facility :-)
This design is begging for a dremel or router attachment to make it a mini mill. I could use a small sawdust producing device, and now I have a good start. Thanks, Dave!
Is it possible to eliminate one of the Z steppers? You should be able to mount the Z in the middle somewhere.
I love the shape of this printer. That photo of it being carried by hand is really awesome! This really seems like the ultimate in portability.
I am probably asking for too much, but did you happen to measure the weight and power consumption of the machine?
This is very high on my list of printers I would love to have given enough time/space/money.
Hi could you please add a DXF of the extruder plate so that this can be cut on a laser cutter?
Much appreciated for your Pocket Printer.
Would there be a significant disadvantage to mounting the x motor to the x axis (or y on the y) rather than having it stationary with the pulley system? You mentioned increased resistance, would this be similar to the resistance of increasing the weight? Just curious. Love the design.
Also, in the firmware, if you knew of much the x would change based on y G-codes, you can have the firmware calculate it's own G position for x. i.e. x needs to move back 1 for every 5 that y moves forward so when the firmware gets a 5mm movement for y it also moves the x back 1. You would have
to have a linear relationship between x and y movements though.
I think keeping the moving mass low is critical for high speeds. The accelerator firmwares are great at reducing jerk, but it is still an issue. I notice a big difference between my mendel and this printer running the same file. The mendel will shake the workbench, the pocket printer is much smoother. Watch the video of the pocket printer on marlin, with fast travel. Also I want to minimize forces on the pla slides.
Conceptually it is pretty simple to combine x and y, but it is pretty difficult in practice. I will keep working on it so that the friction can be reduced in the belts.
Fabulous printer, now this is the one I want to make after I commission my Prusa, it was going to be a Medel90 or Wallace, but this has to be the one. When can you add the missing parts that you mentioned? Limit switches, Power supply and switch mounts. I like the way the controller is hidden behind the printer with the PSU above it and a nice power switch on the side, this is really cute and I want to make one!
This is great! We have been working on an HBOT design that I believe would work for your machine. We have a friend that helped us with the correct firmware
you can check out the Hbot prototype here: http://www.phlatforum.com/view...
&
amp;t=3495
as for the firmware you will need to talk to Jovian on the Phlatforum, unless he finds you here first :) You could always shoot him a PM there.
Great job on this cool machine
This is looking so awesome! Can you make a video with it? Would love to see it move around too!
Maybe add a belt to it so you can carry it with you on your sholder? Add a battery and you can carry it with you while it's printing!
See Joris: http://blog.ultimaker.com/wp-c...
Looks like a great printer!
Makeslide is another good choice that integrates the slides into the extrusions.
I cant see how the x axis moves left to right acuratly when the y axis is moving forward and backwards. does the firmware cause the X motor to turn in order to compensate for the movement of the y?
License
More from dkennell
More from 3D Printer Parts
Hi, great printer, one of the best frame designs out there I think. I am designing a derivative to allow for big extruders so I can print in new materials and have 3 extruder heads. I am hoping to fix the extruder heads and put Z movement in the bed. Is there any way you could upload .STEP files of your Inventor parts? Solidworks user here. Thanks!
OK, I added a zipped folder of the STEP files. Good luck.