1 (edited by sk8kims 2014-02-13 00:55:48)

Topic: circle dented

hi
I received yesterday solidoodle4. so I printed PLA circle sample object.
but circle looks dented. how can I fix this problem?

Post's attachments

IMG_0982[1].JPG 1.84 mb, 2 downloads since 2014-02-13 

You don't have the permssions to download the attachments of this post.

2

Re: circle dented

Belt tension mate, make sure it's equal between the Y and X axis, but not too tight.

This is a crowd funding thing that I'm running: http://www.gofundme.com/bvi140 It's for pretty selfish reasons tongue

3 (edited by kirill 2014-02-25 23:32:40)

Re: circle dented

Helllllppp,

I have exactly the same problem with new 4G Solidoodle. Exactly the same deformation of circles. I passed last 2 weeks to calibrate the axys whit a good improvement of general print quality but with Circles the issue is not changed minimally.

The very strange thing is that with different shapes print seems to be good (2nd image ABS),
but with circle is unwatchable and unusable. (1st image ABS)

I don't know what try else..:(

Post's attachments

20140224_003553.jpg
20140224_003553.jpg 1.2 mb, 2 downloads since 2014-02-25 

20140225_142240.jpg 1.24 mb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

4

Re: circle dented

Those parallel straight edges on the circles are a tell-tale sign of loose/uneven belts.
You mention you've calibrated the axes, but have you performed all the instructions from the solidoodle site: http://www.solidoodle.com/calibration/? Even extruder calibration, bed levelling and z-stop calibration may help you out a bit there.

When you fix it, print another vase the same and hold it against the old one to compare the shape - although it does look pretty schmick already, I'd guess yours is probably a bit squished compared to a perfect one!

Good luck, have fun! smile

SD3. Mk2b + glass, heated enclosure, GT2 belts, direct drive y shaft, linear bearings, bowden-feed E3D v5 w/ 0.9° stepper
Smoothieboard via Octoprint on RPi

5

Re: circle dented

This is another comparison image, the 2 objects are printed exactly with the same axys calibration and only little extrusion difference, tell how the z axys can be the problem..?

Post's attachments

20140226_080500.jpg
20140226_080500.jpg 1.39 mb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

6

Re: circle dented

Thanks for suggestion Grob i surely continue with the calibration tests but the thing that let me perplexed is that I already tryed to tensioning more belts with the result of Y axis bend and no results on circles. I saw some other 4G solidoodle prints and seems everybody have the same problem with circles, at this point i'm afraid of something of more subtle.. i hope to mistake.

7

Re: circle dented

I am finally closing in on getting my SD4 to print perfect circles, and here's what I found:

I printed and installed the pillowblock bearing mod on my Y-rod; I don't think it is necessary on a SD4 (I didn't notice any difference), but it certainly can't hurt, especially if you're going to really crank up the tension on your belts.

Your photo indicates to me that you have some slack in your X and Y belts.  (I had similar-looking circles, but my boxes and other geometric shapes came out fine)  I kept reading posts on here, that said to tighten the belts, but not specifically where or how.  I finally figured it out on my own.

Start with your X belt, the one that moves the extruder left and right.  Using the Manual Control tab of RH, use the arrows to position the extruder in the center of the build plate.  Now use the arrows to move the extruder left and right the maximum single step (50mm if I recall) while watching the belts.  As the extruder moves and stops, any slack that's present will accumulate in the 'return' side of the belt.  When you command the extruder to move the opposite direction, look for a shudder (like the twang of a guitar string) in the belt (the former 'return' side that is now the 'tractor' side).  That 'twang' is slack that needs to be eliminated by tightening the belt's hex screw at the bottom left rear of the extruder.  Make small adjustments to the screw, and repeat the process until the shudder is gone. 

Once you have the shudder removed, loosen the set screws of the sprockets, and run the extruder left and right by hand to even out the tension throughout the three belt segments (upper long segment between sprockets, and the left and right segments between the extruder and sprockets.  You want to remove any slack, have even tension throughout the belt, but not overtighten (which may cause your X stepper motor to stall).  It may take a few attempts to get it right.

If you do this first, the bulges in your prints should move from their present orientation of 10 o'clock - 4 o'clock to 12 o'clock - 6 o'clock.  This will indicate that the remaining slack is in your Y-axis belt.  You are on the right track.

Use the same procedure to eliminate the slack in your Y axis belts, starting with the left one.  Move the extruder forward and back with RH, and observe the belts for slack.  I recall that there is a screw front and back for adjustment.  The added complexity of Y-axis is that you need to tune two belts, and they need to be of similar tension and aligned (not keeping the extruder assembly crooked) for everything to work properly.

To summarize belt tension:  If the bulges are 10-4 o'clock, both belts are loose.  If the bulges are 9-3 o'clock, it's your X axis.  If the bulges are 12-6 o'clock, it's your Y axis.  (the reason both belts will cause a 10-4 and not a 2-8 is because Slic3r runs circles counterclockwise).

This was beginning to frustrate me, especially since my other prints were turning out so well.  My circles aren't quite perfect yet, but I'm very close!  A few more tweaks, and they'll be perfect.

I hope this saves you some butt-pain.

8

Re: circle dented

Thank you very very very much, finally I have some detailed and clear points to work on! Next days I will work on your suggestions, meanwhile for a little satisfaction can you post a photo of your improved circle quality..? Thanks!


knowack wrote:

I am finally closing in on getting my SD4 to print perfect circles, and here's what I found:

I printed and installed the pillowblock bearing mod on my Y-rod; I don't think it is necessary on a SD4 (I didn't notice any difference), but it certainly can't hurt, especially if you're going to really crank up the tension on your belts.

Your photo indicates to me that you have some slack in your X and Y belts.  (I had similar-looking circles, but my boxes and other geometric shapes came out fine)  I kept reading posts on here, that said to tighten the belts, but not specifically where or how.  I finally figured it out on my own.

Start with your X belt, the one that moves the extruder left and right.  Using the Manual Control tab of RH, use the arrows to position the extruder in the center of the build plate.  Now use the arrows to move the extruder left and right the maximum single step (50mm if I recall) while watching the belts.  As the extruder moves and stops, any slack that's present will accumulate in the 'return' side of the belt.  When you command the extruder to move the opposite direction, look for a shudder (like the twang of a guitar string) in the belt (the former 'return' side that is now the 'tractor' side).  That 'twang' is slack that needs to be eliminated by tightening the belt's hex screw at the bottom left rear of the extruder.  Make small adjustments to the screw, and repeat the process until the shudder is gone. 

Once you have the shudder removed, loosen the set screws of the sprockets, and run the extruder left and right by hand to even out the tension throughout the three belt segments (upper long segment between sprockets, and the left and right segments between the extruder and sprockets.  You want to remove any slack, have even tension throughout the belt, but not overtighten (which may cause your X stepper motor to stall).  It may take a few attempts to get it right.

If you do this first, the bulges in your prints should move from their present orientation of 10 o'clock - 4 o'clock to 12 o'clock - 6 o'clock.  This will indicate that the remaining slack is in your Y-axis belt.  You are on the right track.

Use the same procedure to eliminate the slack in your Y axis belts, starting with the left one.  Move the extruder forward and back with RH, and observe the belts for slack.  I recall that there is a screw front and back for adjustment.  The added complexity of Y-axis is that you need to tune two belts, and they need to be of similar tension and aligned (not keeping the extruder assembly crooked) for everything to work properly.

To summarize belt tension:  If the bulges are 10-4 o'clock, both belts are loose.  If the bulges are 9-3 o'clock, it's your X axis.  If the bulges are 12-6 o'clock, it's your Y axis.  (the reason both belts will cause a 10-4 and not a 2-8 is because Slic3r runs circles counterclockwise).

This was beginning to frustrate me, especially since my other prints were turning out so well.  My circles aren't quite perfect yet, but I'm very close!  A few more tweaks, and they'll be perfect.

I hope this saves you some butt-pain.

9 (edited by knowack 2014-03-01 03:53:57)

Re: circle dented

Here's a print of a 35mm pirate coin.  Even using my crappy white filament (I'm trying to use it all up), I'm only .1mm off from being perfectly round at this size.

Not yet perfect, but certainly good enough for most purposes.

Keep the faith!  You can achieve good circles!

Post's attachments

coin.jpg
coin.jpg 1.82 mb, 2 downloads since 2014-03-01 

You don't have the permssions to download the attachments of this post.

10

Re: circle dented

Yeah your circle is exactly what I want but:
..desperation, I passed all day on this with no results.. especially no improvement.

I tryed to tighten progressively the X from very low to "normal", I tryed to watch when the belt "twanging" and now the X seems to be clear but in the photo are the results, the bulges never changed ....:((

What is wrong..?

knowack wrote:

Here's a print of a 35mm pirate coin.  Even using my crappy white filament (I'm trying to use it all up), I'm only .1mm off from being perfectly round at this size.

Not yet perfect, but certainly good enough for most purposes.

Keep the faith!  You can achieve good circles!

Post's attachments

CircleCross.jpg
CircleCross.jpg 294.16 kb, file has never been downloaded. 

printIssues.jpg 302.81 kb, file has never been downloaded. 

QuadCross.jpg 319.66 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

11 (edited by knowack 2014-03-02 03:23:56)

Re: circle dented

Ok, I'm not the expert here, but this is just my opinion:  What those images show doesn't look like slack to me...especially since that geometric image (with the squares) exhibits a similar 'shift'.  That looks like something is binding during movement.  I don't know if it's a belt/pulley issue, or if there's a burr or other imperfection on a rail, or something else entirely.

If I were troubleshooting something like that, I'd first print a few layers of the circle, then kill the print.  Next lower the print bed a few mm, and use the arrows on the Manual Control tab of RH to position the extruder to those locations where the deformities occur (looks like 11 o'clock and 5 o'clock to me).  Note the X and Y position of the carriage components at those locations. 

I'd first make a visual check, to ensure nothing is snagging on something at those positions.  Then I'd start checking belts, pulleys and rails at that position (isolating them and then moving them manually if necessary) to see if you can determine where the binding is occuring.

Another possible troubleshooting method would be to print the same circle scaled up or down, to see if the problem occurs at the same X Y coordinates, or print the same-size circle off-center on the print bed (i.e. shifted toward one corner), again to see how the X Y coordinates play out.

It seems to me that if that if the problem is some sort of binding, it would occur along the same X axis or Y axis bisecting the print bed.  If there's no discernable, consistent line or location where this happens, there may be something wrong with a motor itself.

I may not be explaining myself adequately.

12

Re: circle dented

Ok KnowAck it's clear , one important thing is the "loose" of 1mm that you can see in my examples is equal with all scales. After this analysis solidoodle's support suggested to me to check my .stl projects.
In fact what's happen so far is that I only tryed to print my own Autocad 2011 designs. The only one not mine design is the Vase that i posted before which was printed really not bad. I will test tomorrow some circular designs from thingiverse.

Hoping.......:)


knowack wrote:

Ok, I'm not the expert here, but this is just my opinion:  What those images show doesn't look like slack to me...especially since that geometric image (with the squares) exhibits a similar 'shift'.  That looks like something is binding during movement.  I don't know if it's a belt/pulley issue, or if there's a burr or other imperfection on a rail, or something else entirely.

If I were troubleshooting something like that, I'd first print a few layers of the circle, then kill the print.  Next lower the print bed a few mm, and use the arrows on the Manual Control tab of RH to position the extruder to those locations where the deformities occur (looks like 11 o'clock and 5 o'clock to me).  Note the X and Y position of the carriage components at those locations. 

I'd first make a visual check, to ensure nothing is snagging on something at those positions.  Then I'd start checking belts, pulleys and rails at that position (isolating them and then moving them manually if necessary) to see if you can determine where the binding is occuring.

Another possible troubleshooting method would be to print the same circle scaled up or down, to see if the problem occurs at the same X Y coordinates, or print the same-size circle off-center on the print bed (i.e. shifted toward one corner), again to see how the X Y coordinates play out.

It seems to me that if that if the problem is some sort of binding, it would occur along the same X axis or Y axis bisecting the print bed.  If there's no discernable, consistent line or location where this happens, there may be something wrong with a motor itself.

I may not be explaining myself adequately.

13

Re: circle dented

I had the exactly same flat-ended circles in Y as your Pictures show.
In the end it was the Belt-Tension of the Y-Axis but it was the vertical one.

I had to loosen the 4 screws for the Motor behind the left cover and strengthen the Belt-Tension.

Now I'm quite happy with my circles.

14

Re: circle dented

kirill wrote:

Yeah your circle is exactly what I want but:
..desperation, I passed all day on this with no results.. especially no improvement.

I tryed to tighten progressively the X from very low to "normal", I tryed to watch when the belt "twanging" and now the X seems to be clear but in the photo are the results, the bulges never changed ....:((

What is wrong..?

knowack wrote:

Here's a print of a 35mm pirate coin.  Even using my crappy white filament (I'm trying to use it all up), I'm only .1mm off from being perfectly round at this size.

Not yet perfect, but certainly good enough for most purposes.

Keep the faith!  You can achieve good circles!


Excellent pictures! Did you put a pen on your extruder or how did you make them?

I have an SD3 but I'm also working to perfect my circles right now and have it nearly dialed in.  Most of it was belt tension - the circles video helped.  some was pulley lubrication on the shoulder screws in the front idlers: 3-in-1 oil helped there for now until I can replace with pulleys fixed to shafts on bearings.  I think the rest is probably vref adjustment on the y-axis combined with speeds and acceleration to overcome the friction/sticktion during the direction change.  I don't think it's missing steps but that it doesn't have enough immediate torque, sticks, and then jerks into the intended position.

I'm sure that my alignment is good and that my belt tension is proper - vref and lubrication seem to be the next best bets.

SD3, RUMBA, 360W power, ABS: Glass bed + Aquanet Extra Super Hold Hairspray, Anti-backlash Z spanner, Repetier Host + Slic3r

15 (edited by jagowilson 2015-03-02 05:27:44)

Re: circle dented

Wow what an old thread. I just wanted to point out that If you want to fix your circles ditch the short motor belt and go direct drive. My printer is insanely accurate now. Prints before and after direct drive look like they came from different machines. Best of all printed parts fit perfectly together too, right off the machine. Goodbye circle problems and Y backlash. If I were only allowed one mod, direct drive would be it. It made a bigger difference than even the e3d.

The short motor belt is just terrible and sloppy. It is far too susceptible to error if everything else in the machine isn't perfect. Direct drive with the right coupler works even with fairly poor alignment of things. Here is the mount I am using at this time. It requires a 5-to-6 mm oldham coupler and a new 6mm Y rod (the stock one will be too short).
http://www.thingiverse.com/thing:593006

16

Re: circle dented

Thanks - have been considering direct drive more and more recently. I put it off because I didn't want to drill any holes but that adapter looks great and I'm more comfortable taking my printer apart and putting it back together than I was before.  I'll post back if I try it out. I must have found the old thread in a google search and didn't look at the post date.

SD3, RUMBA, 360W power, ABS: Glass bed + Aquanet Extra Super Hold Hairspray, Anti-backlash Z spanner, Repetier Host + Slic3r

17

Re: circle dented

Do it sooner rather than later. You will be stunned at the difference.

18

Re: circle dented

Claghorn's desgin does not require drilling into your frame. Just grab a few M3x16-20mm screws and use the current Y motor mount slots, then mount the motor

Printit Mason and Printit Horizon printers
Multiple SD2s- Bulldog XL, E3D v5/v6/Lite6, Volcano, Hobb Goblin, Titan, .9 motor, Lawsy carriages, direct Y drive, fishing line...the list goes on
Filawinder and Filastruder #1870.....worth every penny!

19 (edited by jagowilson 2015-03-03 02:06:21)

Re: circle dented

Indeed. Here are some pictures for reference. I will need to drill his part out a bit so I can get better alignment with the coupler, or go with a different design that allows for more alignment adjustment on my SD4. Despite the pretty ok alignment I have going right now (as you can see it is not great) I am still not experiencing any significant lash on my Y axis. The oldham coupler is very forgiving.

http://www.soliforum.com/misc.php?action=pun_attachment&item=8126&download=0
http://www.soliforum.com/misc.php?action=pun_attachment&item=8127&download=0

Post's attachments

image.jpg
image.jpg 990.12 kb, file has never been downloaded. 

image.jpg
image.jpg 1.06 mb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.