376 (edited by redbarret 2014-11-29 17:26:55)

Re: Odd circle test results

pirvan wrote:

OK, I haven't posted on this thread before, but I did read (most of) it.

Doesn't seem like you did. For the reasons I mention below.

First of all, it indeed beats a dead horse.  You are making lots of assumptions about what might be wrong, but are unwilling to try any of the fixes that have already been posted.

This is just not true. As I said in reply to n2ri, I have tried the suggestions posted. Yours included.

Secondly, it would appear that most of the SD4 printers can print a fairly decent circle to start with

I don't know where you got that info, but I've talked with like 8 people in this forum and over emails and they all have this problem. In fact I have never seen any SD4 printing decent circle to start with.

and you're spending more time printing silly circles, that printing anything useful or fun.

Again not true. Look, the reason I personally started using the circle test was because I was noticing artifacts in my *real* useful prints and wanted to test how my prints were improving as I was changing stuff on my printer. The circle test uses far less filament and time.

How often do you really need to print a perfect large circle?

If you can't print proper circles, it means you likely have backlash. And if you have backlash, you'll have problems on almost every print you'll make, including issues like gaps between infill and shells.

As n2ri said, these are unique machines, with their own character, there is no single solution.  Yes someone's solution may work for others, but not all.  So the is no ONE SOLUTION to rule them all.

This is just a claim. And I don't agree with it.

You guys are making it sound as if you're looking for the holy grail, but all you're doings going around incircles, complaining that the sky is falling and waiting for some miracle.

No, that is how you're seeing it. What we are doing is discussing the possible solutions. And you are coming and telling is to not complain, that is not what we're doing.

Go out there and make something.

Like I said, I have.

Andstop complaining about the smallest things.

Ugh... this is just, I don't even know what to respond to this. We have spend hundreds of dollars (over 1000 by now with all the upgrades) to buy a printer which can't print anything out of the box properly and come here for help and discuss what the issue might be, and have people like you tell us we are "complaining about the smallest things?".
Yes, again you're pretty rude.

Ultimately if this SD4 printer is not to your liking, maybe you should buy a different one.

Well maybe everyone is not as rich as you to be able to buy another printer if he can't get the one he has just bought to work properly?

"the perfect circle" is your ultimate goal, maybe you should look for one of those printers.

no one is going for perfection here

To put your mind to ease, I did replace the back 6mm rod with stronger one from McMaster-Carr, but it didn't make much difference, I also put a pillow block with support bearings, it made no difference either.  FOR ME, it was the backlash in the small motor belt that solved my problem.

This is why I think you haven't read everything before replying.
The point of replacing bushings with bearing and getting a better rod is to be able to tighten that small belt you are talking about more without causing too much friction which leads to step skipping first right now.

I really don't see where you're getting your facts.
I'm trying everything I can which is suggested.
I've ordered flanged bearings to replace my bushings when they arrive.
I'm going to replace my back rod and suggest others that might be the problem.
Where am I complaining and expecting for a miracle??

Solidoodle 4

377 (edited by nbourg8 2014-11-29 18:43:13)

Re: Odd circle test results

So much drama!  I only stated what I said because the original people who were helping the cause quit posting because of drama.  If it continues, I'll ask a moderator to close this thread.

I will return from my Hawaiian honeymoon cruise tomorrow and start being useful again.

Everyone's individual solution to this problem is very important to us. Yes we know each printer is different but they all have a fundamental foundation under scrutiny in this thread. We understand there is not a universal fix but who is to say no one will have the same solution as another.

Constructive comments are how we bring new people with new ideas to the thread.

378

Re: Odd circle test results

Cool- let me know if you need any help with the lawsy carriage/linear bearings upgrade. I posted a how-to several pages back. It's well worth it.

379 (edited by n2ri 2014-11-29 19:23:50)

Re: Odd circle test results

redbarret if you truly think this statement

"If you can't print proper circles, it means you likely have backlash. And if you have backlash, you'll have problems on almost every print you'll make, including issues like gaps between infill and shells."

then you need to search topics on upgrading you 'Z' axis threaded rod. thats how backlash has gotten tweaked better. dont see how it has anything to do with circles in the "X" "Y" plain though, unless your build plate some how moves side to side.

Solidoodle 2 with Deluxe kit cover & glass bed with heater. and 2nd board SD2 used not 3rd and alum platform not installed yet still wood. also need cooling fan installed to board. use Repetier Host couple vers. Slic3r also have all free ware STL programs

380

Re: Odd circle test results

nbourg8 wrote:

So much drama!  I only stated what I said because the original people who were helping the cause quit posting because of drama.  If it continues, I'll ask a moderator to close this thread.

I will return from my Hawaiian honeymoon cruise tomorrow and start being useful again.

Everyone's individual solution to this problem is very important to us. Yes we know each printer is different but they all have a fundamental foundation under scrutiny in this thread. We understand there is not a universal fix but who is to say no one will have the same solution as another.

Constructive comments are how we bring new people with new ideas to the thread.


that will not be the outcome when a 'thread' has more than about 4 pages length (much less 10x that) and lots of off topic or start over posts mixed into actual fixes. when I 1st joined and searched for help with issues I had. I 1st read short topics related to my issues. avoiding long or really long topics like this because I wanted to start printing something ASAP with my investment.

not spend all my time re-posting on the forum like the affore mentioned members who say they have printed and posted useful things, then next post say they havnt been able to print any good prints since getting their printer.

Solidoodle 2 with Deluxe kit cover & glass bed with heater. and 2nd board SD2 used not 3rd and alum platform not installed yet still wood. also need cooling fan installed to board. use Repetier Host couple vers. Slic3r also have all free ware STL programs

381

Re: Odd circle test results

n2ri wrote:

redbarret if you truly think this statement

"If you can't print proper circles, it means you likely have backlash. And if you have backlash, you'll have problems on almost every print you'll make, including issues like gaps between infill and shells."

then you need to search topics on upgrading you 'Z' axis threaded rod. thats how backlash has gotten tweaked better. dont see how it has anything to do with circles in the "X" "Y" plain though, unless your build plate some how moves side to side.

n2ri, backlash occurs in X, Y, and Z. redbarret is correct. X and Y backlash comes from belt tension primarily, and Z backlash from the fact SD uses a regular nut on threaded rod instead of anti-backlash nuts or lead screws.

As for the arguing and bickering, knock it off guys. If you don't like what someone is posting about, simply don't respond to them.

382 (edited by n2ri 2014-11-29 19:56:03)

Re: Odd circle test results

and just because you havnt seen an SD4 print good circles of the few you know of dont mean theres not more than a thousand that do. how many users are going to search for a Solidoodle forum just to brag on how good of circles it prints?? nobody wastes time with that.

on average companies know that customers that complain of problems are a small percentage (maybe 10%) of those that wont even do that when they have an issue.

but at the same time that figure amounts to less than 5% of total sales most times as low as %1

PS. stop using cuss words in posts if you expect posts not to be edited/deleted or worse. this is not the place for it. @ redb...

Solidoodle 2 with Deluxe kit cover & glass bed with heater. and 2nd board SD2 used not 3rd and alum platform not installed yet still wood. also need cooling fan installed to board. use Repetier Host couple vers. Slic3r also have all free ware STL programs

383 (edited by redbarret 2014-11-29 20:22:51)

Re: Odd circle test results

elmoret wrote:

If you don't like what someone is posting about, simply don't respond to them.

This ^

Solidoodle 4

384 (edited by n2ri 2014-11-29 20:59:12)

Re: Odd circle test results

redbarret wrote:
elmoret wrote:

If you don't like what someone is posting about, simply don't respond to them.

This ^

NO! cussing etc are against forum rules. for good reason. minors plus others on the site dont want to see it. dont matter if your still in school.

Solidoodle 2 with Deluxe kit cover & glass bed with heater. and 2nd board SD2 used not 3rd and alum platform not installed yet still wood. also need cooling fan installed to board. use Repetier Host couple vers. Slic3r also have all free ware STL programs

385

Re: Odd circle test results

n2ri wrote:

NO! cussing etc are against forum rules. for good reason. minors plus others on the site dont want to see it. dont matter if your still in school.

That is incorrect. Here's the rules if you'd like to re-read them:

http://www.soliforum.com/rules/

Further off-topic posts (or portions of posts) will be deleted.

386

Re: Odd circle test results

I think it will be very difficult to really eliminate backlash without a dial indicator, as I suggested at the google group.  When adjusting the side belts, too loose can give you backlash, and too tight can give you backlash.  If you are just making adjustments and eyeballing circles, you could be just hitting either side of the ideal tension.    I shot some video about this with my SD2 a long time ago and never got around to making it into a blog post, so here is a bit of it-

I don't know what mounts would work with a SD4, but here is one I made to slide over the extruder stepper -

http://www.thingiverse.com/thing:69029  The slots were a tight fit for my indicator, so they may need some carving out or adjustment in the file depending on what kind you have.

Here is another mount that goes on the heads of the 4 screws that hold the extruder to the stepper motor.  Again, I don't know enough about the 4 to know if it will work.  This is from a 2012 thread about Y backlash on the SD2.

http://www.soliforum.com/post/1169/#p1169

Here is another relevant bit from a post I made in 2012-

Playing around with mine some more this morning, I found that I can dial the backlash up and down with the tension of the Y motor.  When I got the backlash to .03 with the back belt, I started getting binding at the front of the platform.  I found that loosening the side belts by a full turn relieved the binding without increasing the backlash.  I also found that where a 10mm move command got  a 10.01mm move, after loosening the belts it was 9.93.  It was still consistent, and with low backlash.  I'm not going to quibble with .07mm, the actual extrusion of plastic isn't that precise.

If you really want to understand what is causing backlash and how the tension of different belts affect it, you will need to be able to measure it directly after each adjustment.   I also eventually bought a stepper with higher torque for Y so I could keep the belt tighter without worrying about skipping.  It was a longer motor and required longer M3 bolts.

387

Re: Odd circle test results

Thanks Ian for the post. I'll try to find a dial indicator. As for mounting it, I'll make an STL myself if the one you linked to doesn't work for SD4.

But one question, and I'm sorry if you've said this already, but why will tightening the Y motor belt too tight cause backlash as well?

IanJohnson wrote:

I also eventually bought a stepper with higher torque for Y so I could keep the belt tighter without worrying about skipping.  It was a longer motor and required longer M3 bolts.

That's the reason me and jagowilson are thinking of replacing the bushings with flanged bearings and maybe add another pillow bearing. The idea is to reduce friction and rod bending so we can tighten the belt some more without getting step skipping.

Solidoodle 4

388

Re: Odd circle test results

It's the side belts that can add backlash either way.  When too tight it might actually be skipping rather than backlash.  Microsteps have less torque, you only get the full rated torque of the stepper at the full step.  So the axis might be hard enough to move that all the microsteps get skipped.  Maybe only 1 full step gets skipped and only on the direction change so you don't notice the sound, and it looks like backlash rather than an offset layer.  There was at least one instance when I had the backlash almost eliminated, and when I tried tightening the side belts to get it the rest of the way, it jumped up to .3mm.

Something else to keep an eye on is those idler pulley mounts.  I printed the extra beefy version from Thingiverse, and even they started to become more flexible over time.

389 (edited by jagowilson 2014-11-30 23:19:37)

Re: Odd circle test results

IanJohnson wrote:

I think it will be very difficult to really eliminate backlash without a dial indicator, as I suggested at the google group.  When adjusting the side belts, too loose can give you backlash, and too tight can give you backlash.  If you are just making adjustments and eyeballing circles, you could be just hitting either side of the ideal tension.    I shot some video about this with my SD2 a long time ago and never got around to making it into a blog post, so here is a bit of it-

I don't know what mounts would work with a SD4, but here is one I made to slide over the extruder stepper -

http://www.thingiverse.com/thing:69029  The slots were a tight fit for my indicator, so they may need some carving out or adjustment in the file depending on what kind you have.

Here is another mount that goes on the heads of the 4 screws that hold the extruder to the stepper motor.  Again, I don't know enough about the 4 to know if it will work.  This is from a 2012 thread about Y backlash on the SD2.

http://www.soliforum.com/post/1169/#p1169

Here is another relevant bit from a post I made in 2012-

Playing around with mine some more this morning, I found that I can dial the backlash up and down with the tension of the Y motor.  When I got the backlash to .03 with the back belt, I started getting binding at the front of the platform.  I found that loosening the side belts by a full turn relieved the binding without increasing the backlash.  I also found that where a 10mm move command got  a 10.01mm move, after loosening the belts it was 9.93.  It was still consistent, and with low backlash.  I'm not going to quibble with .07mm, the actual extrusion of plastic isn't that precise.

If you really want to understand what is causing backlash and how the tension of different belts affect it, you will need to be able to measure it directly after each adjustment.   I also eventually bought a stepper with higher torque for Y so I could keep the belt tighter without worrying about skipping.  It was a longer motor and required longer M3 bolts.

This is useful information, thank you Ian. I also have a higher torque motor, but I'm waiting on a McMaster-Carr order that should have bolts that wil allow me to mount it to the frame. I don't anticipate any differences, but more power is always good. I'm pretty pleased with my printer's output lately, though.

390

Re: Odd circle test results

if by some freak of nature Solidoodle does watch these topics. I think the 1st step they should take towards quality checking any SD4s in stock b4 shipping them is to;

do the circle test from this site to see if they need rebuilt 1st.

or just sit back and watch whats left of their company sink.

Solidoodle 2 with Deluxe kit cover & glass bed with heater. and 2nd board SD2 used not 3rd and alum platform not installed yet still wood. also need cooling fan installed to board. use Repetier Host couple vers. Slic3r also have all free ware STL programs

391

Re: Odd circle test results

I am well on my way now to conducting the following modifications:

Glass Bed (Completed 11/30/2014) - Replace Kapton with Glass/Binder clips and adjust Z-stop
     http://i61.tinypic.com/30triw1.jpg
Glass Lid (Complete 11/30/2014) - Replace solid black top with Glass and printed hinges
     http://i58.tinypic.com/sn2iw9.jpg
Z Backlash Well Nut (Complete 11/30/2014) - Add a Well Nut to the Z Axis in an effort to minimize Z backlash.  I didn't experience any but maybe it will help with my initial layer. 
     http://i57.tinypic.com/20qnk9f.jpg
Bed Stabilizer (Pending shipment of rod) - Intended to prevent left/right jiggle of bed during prints of short distance causing shaking of the bed.
X/Y Dial Indicator (Dial on order from Amazon) - Useful for measuring X/Y backlash
Lawsy Carriage (Planned for Tuesday) - Removal of flex and play in x axis rods and preparation for e3d v6 hotend.
Y-axis Brackets (Planned for Tuesday, pending completion of Lawsy Carriage) - I've noticed play in my Brackets; mainly the left bracket.  Upon inspection, these were printed on an awful printer.  This is more for my own appeal.
Y-axis Motor Fan (Planned for Wednesday) - Relocation of rear case fan to the y-axis motor.  OPTIONAL.  Some users report motor heat, I've not yet even checked for it but I printed the parts anyhow.  I'm mod happy at this point.
IanJohnson's Instructions (Planned for Wednesday) - Adjustments to the y-axis stepper motor tension belt.  Ian has provided a set of instructions with video on how to properly tension this belt with respect to backlash generation due to over/under tightening.

392

Re: Odd circle test results

Nice smile
I don't think I have problem with the bed shaking. Maybe I just don't understand the issue very well.

As for me:
1) replacing bushings with bearings - waiting for the 686ZZ flanged bearings to arrive
2) Replacement carriage - waiting for the LM8UU bearings to arrive

Also trying to find a dial indicator locally as if I buy online now it might only arrive in 2015.

Solidoodle 4

393

Re: Odd circle test results

This is the spirit guys! Enjoy the mods! 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

394

Re: Odd circle test results

Here's an update on my work yesterday.

I tried to print side panel mounts with new filament and had a big problem occur.  These new mounts had to be printed with a new role of filament.  This filament was measured with a caliper to be the exact same size as the previous filament and was made by the same company.  This problem started on the 2nd layer of the print and only took 5 minutes to get where it is in the picture.  Well, to make a long story short, here's a picture. 
   http://i61.tinypic.com/fz2oti.jpg

What could be the cause for this?!?

Anyway, this meant I had to take things apart so I decided now is a good time to start my upgrades.

Goal = Lawsy Carriage, Y-axis Brackets

Completion Status = 50%

Summary:
I've dismantled the entire SD4 down to its bare skeleton.  I quickly realized that to get certain items free of the contraption, I had to remove others and before I knew it, the guts were sprawled out on the floor in front of me.  Thankfully, I've done enough troubleshooting to build it by now so putting the pieces back together isn't difficult. My biggest gotcha was having all the right tools for the job.  Having both metric and SAE sets of allen wrenches, socket sets, a hammer and clamp is what I used.

When dismantling the machine, I came across some interesting discoveries:

  • The right side y-axis rod was slightly larger at its ends.  This didn't seem to be a problem as the area which the carriage slides was nearly perfect every inch 360 degrees around with deviations of up to .03mm.  At the ends, this deviation was around .11mm which caused problems with extracting it from the frame.

  • I noticed a slight but very noticeable slip in the x-axis rods in their epoxied holes on the carraige.  I pulled out the entire x-axis from the system so I could witness any loosness in epoxy and sure enough, I was able to pull off the entire LEFT side carriage from the rods BY HAND (no hammer necessary and with little force).  The Lawsy carraige, on the other hand, had to be put on the rods with a hammer and is impossible to wiggle free by hand.  I didn't epoxy it yet but I'm wondering if the epoxy will just squish out the sides due to no room between the rods and carriage.  There is NO movement on that x-axis with lawsy's carriage as a result, it's TIGHT.
       http://i61.tinypic.com/3482a9l.jpg

  • My stock extruder assembly does not fit the hole locations on the lawsy carriage.  This is a big problem for me.  I may end up having to wait until x-mas to put an extruder on top.  I have the e3d v6 with bowden on my x-mas list.

  • I am also concerned about the linear bearings not being snug enough in the lawsy carriage.  They do jiggle.  When I'm done, will they be inherently snug by the completion of the process or should I shove some shims in there?

  • I'm very likely to have some issues with clogging now due to the shavings entering the hotend from the topside.  Will deal with this if I have to.  The new e3d v6 hotend should inherently replace the current hotend and I won't have a problem again.

395

Re: Odd circle test results

Nice to see your having fun with it.
As far as the bearings not being snug most of us just put a single wrap of masking tape around the bearings to snug them up in lawsy's carriage.
I would like more details about the bolts of your extruder not lining up with the carriage though.  Can you post some pictures up you may have something backwards or it might be something with a simple work around. 
Keep up the good work!

Printit Industries Model 8.10 fully enclosed CoreXY, Chamber heat
3-SD3's & a Workbench all fully enclosed, RH-Slic3r Win7pro, E3D V6, Volcano & Cyclops Hot End
SSR/500W AC Heated Glass Bed, Linear bearings on SS rods. Direct Drive Y-axis, BulldogXL
Thanks to all for your contributions

396

Re: Odd circle test results

The bearings being loose in the Lawsy carriages is a good thing.  Get some Instamorph and squish a little bit into the bearing mounts.  Instamorph is a brand name for Polycaprolactone plastic that comes in pellets.  They will melt in boiling water and stay soft at cool enough temperatures to handle with bare hands.   

When I installed mine, I squished a small amount into the back of the bearing mounts and then shoved the bearings in.  When I installed the carriage on the rods, the movement was fairly tight.  For several minutes I ran a heat gun on the carriage to keep the instamorph hot while I slid it back and forth.  Once the movement became really smooth and easy, I took away the heat and kept sliding until the instamorph cooled and set.  This fixed the bearings in exactly the position they needed for perfect alignment.

397

Re: Odd circle test results

wardjr wrote:

Nice to see your having fun with it.
As far as the bearings not being snug most of us just put a single wrap of masking tape around the bearings to snug them up in lawsy's carriage.
I would like more details about the bolts of your extruder not lining up with the carriage though.  Can you post some pictures up you may have something backwards or it might be something with a simple work around. 
Keep up the good work!

I'll post some when I get home from work. ~7 hours from now.

IanJohnson wrote:

The bearings being loose in the Lawsy carriages is a good thing.  Get some Instamorph and squish a little bit into the bearing mounts.  Instamorph is a brand name for Polycaprolactone plastic that comes in pellets.  They will melt in boiling water and stay soft at cool enough temperatures to handle with bare hands.   

When I installed mine, I squished a small amount into the back of the bearing mounts and then shoved the bearings in.  When I installed the carriage on the rods, the movement was fairly tight.  For several minutes I ran a heat gun on the carriage to keep the instamorph hot while I slid it back and forth.  Once the movement became really smooth and easy, I took away the heat and kept sliding until the instamorph cooled and set.  This fixed the bearings in exactly the position they needed for perfect alignment.

I saw your post actually about the Instamorph in another thread on this forum.  It's a very neat solution and one I might try but my problem is no heat-gun.  I'm going to try the masking tape solution first since I can remove it easily if it fails to perform.

398 (edited by jagowilson 2014-12-03 17:11:59)

Re: Odd circle test results

You can use your stock hotend with the standard mk5. If you are more interested in upgrading to the e3d anyway, I'd be more than happy to print you an mk5 for it and ship it out. I'll even throw in a 625zz bearing because you will need it for the tension arm (I ordered 10 and only used 1). Shoot me a pm if this interests you. I'll print whichever you want, a stock Mount or e3d, but seriously, just get the e3d and throw that PEEK junk away. Let me know ASAP because a week from Friday I will be away from my machine until January

Your filament issue there is because the stock aluminum extruder does not allow tension adjustment. You will have problems with the lawsy carriage trying to use it. I couldn't feed tougher filaments like t glase without the full mk5. Using non-Solidoodle filament is impossible on the aluminum extruder without shimming the spring. Been there, done that. It's all junk. I suspect this is because SD filament averages 1.6mm instead of 1.75mm. God knows why they're so under-diameter, but it does make your print quality suffer (in my experience).

399

Re: Odd circle test results

Weird, I have two rolls of Solidoodle filament and they both seem to be 1.73mm.
That filament shredding happens for me when I get clog or extruding at lower temp than needed or extruding too fast so the filament doesn't melt and come out as fast as it goes in.

Solidoodle 4

400

Re: Odd circle test results

Jago, I sent you a PM.  Thanks for the help.

@Red, My G-Code and Printer Settings in slicer prevent the beginning of the part until both extruder and bed temperatures are met.  Does yours?