Printrbot Talk Forum

by **JohnSL** » 2013-Jun-Sat-11-Jun

Like others, I've been plagued with the "sagging" Y on my Printrbot. The following photo shows the end result of not being able to have the bed flat when it's leveled relative to the head:

P6011625_Cropped.jpg
As you can see, the vertical is tilted to one side at the same angle as my bed.

I've tried a number of different solutions to this problem, all without luck. After some use I notice that the Y sag was worse, so I replace all the zip ties and used a zip tie gun to get them really tight. That got the sag back to my original sag. I think my sag is a result of warping of the base board, but I'm not really sure. In any event, getting this fixed with the current design seems really hard.

So, I have an idea that I thought I'd throw out there for discussion before I have a design that's ready to print. Here is my idea in a nutshell. Replace the 5mm laser-cut plywood Y/Z plate with two printed plates that can be slightly rotated to adjust the Y axis angle. I've started to create the 3D models, which will just barely fit onto my print bed. Here is what I have so far:

Y Z Plates_400.jpg
The idea is that i'll use an M5 screw on the right side as a pivot point. The stepper is attached to the Y plate (the front one that you see). What I haven't drawn yet, and I'm still thinking about, is how I'll adjust the angle between the two plates. I'm thinking of some mechanism along the left side.

What do you guys thing?

-- John

**Sponsor**

by **twistedmedium** » 2013-Jun-Sat-12-Jun

I think you are on the right track. I had a similar idea but mine may be simpler. My time has been super short lately so I will explain it to you and if you decide it will work feel free to create it. Instead of a whole plate like you have I think the easier option (and printable on the simple) is to just make individual bearing holders and out of the back of the bearing holder extend four tabs that fit through or clip to the existing holes where the zip ties would go and then make it so that the bearing can be zip tied to the holder. Depending on how you make the tabs that go through the Simple's existing holes you could then have a few options for attachment. I can scan my initial drawings if my explanation doesn't help. If you don't like the idea, then I think you are perfectly fine on your path!

by **JohnSL** » 2013-Jun-Sun-19-Jun

That's a very good idea. It would certainly eliminate any droop resulting from the zip ties not being super tight.

As far as I can tell, the drooping of my Y axis isn't a result of loose zip ties. Instead, it appears to be warpage of the laser-cut wood base, and perhaps a slight missalignment of the rod holes. I haven't been able to figure out the exact cause. The design I'm working on will have the positiving holding of the bearings, as well as a way to adjust the angle a few degrees with a screw so you can get it very close to level with the base. Did you have a way in mind to make that kind of adjustment with separate bearing blocks?

by **RetireeJay** » 2013-Jun-Sun-19-Jun

I'd say be cautious about how your linear bearings are mounted. Your manufacturing process has to be exactly spot on or else you will have bearings that don't share the same axis (slightly offset) or aren't perfectly parallel. These bearings are very high precision, so we're talking microns of alignment. Or else (and my preference) is to find a way to allow the bearings to turn just a little so they are co-axial.

If you have ever seen very high-precision machine work where, for example, two plates need to be perfectly aligned with one another, there will be two pins on one of the plates and two holes on the other plate. One hole will exactly match the diameter of the pin, and the other hole will be slightly elongated to allow for manufacturing tolerance and thermal expansion.

by **twistedmedium** » 2013-Jun-Sun-21-Jun

If you search on Thingiverse or even just Google for "lm8uu bearing holder" you will find a multitude of designs. My goal was/is to slot the bearing into the holder and then be able to clamp it in either with 1-2 zip ties or with 1-2 screws/nuts on the front edge, and then what I wanted was some sort of system that fit through the existing zip tie holes in the laser cut wood that either clipped securely or offered some sort of method of connecting the upper and lower tab that sticks through to each other for bracing. That last bit was actually my hangup. I was last toying with a design that had four recessed holes in the bearing holder for small screws and nuts to go through the zip tie holes and then use nuts and then they wouldn't interfere with the bearing sliding into the holder and then the holder still can clamp down at the front.

My laser cut wood is straight and I actually don't have much more than 1 degree of Y sag on mine with just zip ties, so I wasn't trying to correct for any sag in the wood, just make a more secure mounting for the bearings to eliminate the sag I had. Another part that can be utilized is the slot that the bearing currently sits in since it isn't needed with the holder design, I just was using a bit of plastic that fit snugly in there to add more rigidity but there is probably a design that would utilize it in some way that would be better.

by **rsilvers** » 2013-Jun-Sun-22-Jun

The solution, and I listed this already, is for Printrbot to laser cut the bearing slots at a 1.5 degree angle, then send us all new wood as part of the upgrade package.

by **DumpRun** » 2013-Jun-Mon-11-Jun

by **twistedmedium** » 2013-Jun-Mon-11-Jun

I'll toss in my 2 pence on the issue... I think trying to compensate int he cutting of the wood assembly is probably not the solution. The sag will change and be different for everyone. I doubt mine is even a full degree off currently, so if it were cut 1.5 I'd be off about the same as I am now. Also, I do believe that with time the zip ties will flex a bit more, since they are all cut off flush this will be hard to fix without cutting and re-tying the bearing and again will keep it from being a static 1.5 degrees off. Obviously the best solution is something completely rigid for the bearings/rods, metal being the top choice with something like aluminum rapid rails. That is unlikely. Within the constraints of what we have stock I think the problem is very difficult, maybe impossible to solve. There may be an ingenious fix floating out there though. Maybe a different wooden bearing slot/holder is an option. I don't see a place for a bracing change (aside from the zip tie trick) which doesn't really help Y. Offloading the weight helps but still won't eliminate it. I think this is the one aspect that some other similarly priced 3D printers will have over the Simple no matter what, and it is kind of a big deal. We all can still get decent prints and careful object placement can minimize it on certain objects (like the filament stand) and that really may be the best it gets. I'm hoping Brook or someone comes up with something cool though as a fix, and I keep mulling ideas over but none seem like universal fixes.

by **JohnSL** » 2013-Jun-Mon-21-Jun

I agree that this is a difficult problem. I thought my sagging might be a result of the weight of the head. So I attached a string and a counter weight to see if that would change the bed leveling. It really didn't have much effect. Likewise, I've redone all my zip ties, without fixing the issue. My conclusion is that the 1.25 degree droop I'm seeing is a result of warping, or missalignment of some sort.

The design I started with is a prety involved design, and may be a good solution. The print time is likely to be quite a few hours. Before I spent too much time on that solution, I thought I'd trie twistedmedium's suggestion. I've made several versions, and it's tricky getting the tolerances just right so they don't twist in the existing hole on the laser-cut Y-Z plate.

Right now I'm printing version 3 of separate bearing holders. The idea I'm working with is to have the bearing holder at an agle to the slot, but securing them in the slot is where I'm having problems. Version 3 includes a wedge that I'm hoping will hold it firmly in place via the zip ties. I'll share photos once I have them printed. I have two wedges printing at the moment, and they should be done in about 20 minutes. Then I need to print a second bearing holder, which will take another 40 minutes. So I should have photos in a couple of hours.

by **twistedmedium** » 2013-Jun-Tue-06-Jun

Great news, That was also the same issue I was running up against and I'm glad you have taken it and run with it. I recently started a new job (the same week my printer kit came, of course) and responsible for a global network/systems so my free time has been at a premium with the addition of a small child too. I have tons of updates and pics I need to get up here but just haven't had the time. Your entire plate system is a good idea, and I've been thinking about it and came up with a possible solution: marry the two ideas! Instead of trying to print a large flat plate, simply start with acrylic/metal and drill some holes in it for the motor and individual bearing holders. That way you are still only printing bearing holders and anyone can get a piece of acrylic cut to size and then knock out a few drill holes with a paper template even. Just another thought. At this point I have more time to think than to design, but I'm getting there.

by **JohnSL** » 2013-Jun-Tue-10-Jun

Well, no photos yet. I made a stupid mistake in my design, so I'm making the parts again.

That's a great idea twistedmedium. I have access to a laser cutter, so I should be able to make some plates. I think I'll try that after finishing my current design. Right now I need to get ready to leave for work.

-- John

by **JohnSL** » 2013-Jun-Wed-00-Jun

Success!!!

P6041633_Cropped.jpg
When I measure the difference between the carriage and bed from front to back along the Y axis, the difference is 0.14mm. I'd say that's good enough, especially when you consider it was 2.75mm difference before. I used rsilvers' angle of 1.5 degrees, which turned out to be perfect on my machine.

Here are some more pictures of the parts. There is a front and back part. The back part is the wedge on the left, which has a slight angle so you can push it in until it firmly holds the front part in position:

P6041629_Cropped.jpg
Here is a view showing the front piece in place. I only have two zip ties on at the moment, which turns out to be really strong because of the positive seating of the bearings in the curved holder:

P6041630_Cropped.jpg
And finally, here is a view from the back, so you can see the wedge in place:

P6041631_Cropped.jpg
The front pieces each take about 40 minutes to print, and the wedge a lot less time (I don't remember how long). I have a few more changes I want to make to the wedge as it's a little too tight right now. But at least this solves my Y sag, and may solve it for others as well.

You can find it here: http://www.thingiverse.com/thing:99364

-- John

by **twistedmedium** » 2013-Jun-Wed-06-Jun

Great job! Very close to my idea in reality and I'm super glad that it works! I had only about an hour last night to play and I was printing 6 fairly detailed pieces and the ones that were positioned back on the Y axis printed fine, the ones up front had tons of small problems and it was all directly attributed to the Y sag. I will definitely start here and see if I can use them as-is or possibly improve anything. Really glad you were able to make sense of my text and actually execute!

I really think something along these lines should be doable by Brook and hopefully will help him fix this issue!

by **JohnSL** » 2013-Jun-Wed-10-Jun

Thanks twistedmedium. Your idea was a whole lot simpler, and this is one of the great things about community.

I plan to make a few changes to the wedges tonight, and then publish the files to Thingiverse. I'll be particularly interested to see how well this works for others. Perhaps we don't really need any adjustment.

If we do need some adjustment, I could simply upload angled blocks that have different angles. Alternatively, I could pursue your idea of laser-cut plates. If we don't need any adjustment, on the other hand, I'm almost done

by **twistedmedium** » 2013-Jun-Wed-11-Jun

It is really cool! It just kills me that my time is so short right now, I have so much I want to do and make and I just can't get to it. Things should get back to normal soon though once I get acclimated at the new job and put out some of the bigger fires that eat up my time.

As for the wedges, I think targeting the 1.5 degree angle is fine. If anyone is less than that they can just file it a bit or if they are over they probably need to tighten their ties, etc. but could always use a small shim to get them spot on. I have been trying to pinpoint the source since my zip ties are rock solid and I still have a tiny sag (1-1.5 degree at the extreme front) from my investigation it seems to be play in the bearings that causes it not the ties (at least for me). If that is indeed the case, then everyone would be off by just about the same amount since the smooth rod/bearings are all the same and the 1.5 degree fix is the only one needed. If that is indeed the case, then rsilvers is absolutely correct in his initial statement that the fix is for Brook to simply laser cut the area at a set 1.5 degree angle and then the printer is level.

My only question to you and has been my concern from the beginning when I was thinking about the issue is that if this 1.5 degree angle solves the issue for the extreme of Y, does it then introduce a problem at the Y min causing it then to be off? If so, I guess it can be fixed by bed leveling but is still an inherent issue.

by **JohnSL** » 2013-Jun-Wed-12-Jun

I'm also stumped about the source of the sag. The bearings introducing the sag is an interesting theory. I wonder, if that is the source, would there be vertical/horizontal play of +- 1.5 degrees? Before I installed the angled bearing blocks, I didn't have any play at all. Like you, my zip ties were really tight. Very odd...

With the angled bearing blocks, my print head now follows the bed perfectly from front to back in the Y axis with the print bed flat as shown in the photo above. So no, there isn't a front-to-back issue. The angled blocks fix the problem completely, whatever the source is...

Another idea just occurred to me. I wonder if it's possible that the holes for the smooth rods are slightly off from top to bottom. If that's the source, it would explain why rsilvers' suggestion of 1.5 degrees perfectly fixed the problem for me. When I get home, I'll have to do some measurements of various parts to see if I can pinpoint the source.

Hope your life evens out. Designing and building upgrades for the Simple is a lot of fun. I'm looking forward to seeing what other ideas you come up with.

by **rsilvers** » 2013-Jun-Wed-13-Jun

My measurements show that 1.6 would be perfect.

by **DumpRun** » 2013-Jun-Wed-14-Jun

by **ahaer** » 2013-Jun-Wed-14-Jun

Is there a way to compensate for this in software?

I did a quick look at g-codes (http://reprap.org/wiki/G-code) and their doesn't seem to be anything but maybe there is something in slicr config???

What we want is some sort of vertical y-axis offset (ie y = y + z*ofs)

by **ahaer** » 2013-Jun-Wed-14-Jun

by **twistedmedium** » 2013-Jun-Wed-14-Jun

by **ahaer** » 2013-Jun-Wed-18-Jun

Look at the way things move when you lift up and down on the extender I think the root of the problem is that the curved z-axis nut support piece is way too flexible. There is no way that either the x and y bearings can things perpendicular a force (the weight of the head) is applied soo far away from the support (the z-nut).

You can see this sag if you put a second nut on the z-rod and spin it from the bottom of the motor to the top
nut_top.jpg
nut-bot.jpg

I jammed in a small shim and was able to live the extruder when it was fully extended.
nut-shim1.jpg
nut-shim2.jpg
However as I attempted more lift the soft coupling started bending from the force. I'm going to print a solid coupling now and try again. Theoretically, if the motor being supported at two points along the z-rod and the rod doesn't wander then the problem is solved. My initial experiments with moderate shim/lifting didn't indicate any binding problems, and there really shouldn't be as the sagging force is distributed over a large distance

BTW If this does work I expect millions of reputation points

by **twistedmedium** » 2013-Jun-Wed-18-Jun

That is a great idea! Not only that but you made me think of the Y axis issue differently and maybe a solution! If there was a rail extending out that the Y motor rested on and slid freely on in the inside of the printer it would support the motor on its journey out and eliminate any sag! Think of an L shaped piece that could act as both a shim and a rest! Sadly my Z motor pooped the bed tonight so I am out of commission.

by **ahaer** » 2013-Jun-Wed-20-Jun

getting a solid concentric coupling isn't as easy I hoped. I printed a coupling (http://www.thingiverse.com/thing:94815) but the shaft hole was a bit to tight so I had to drill it out. Being a compete idiot I didn't it by HAND and made the hole out of non-concetric. Will try again tomorrow.

I think I'm going ask someone at work to machine me a piece on the lathe with my 3/16 shaft on on end and a 5/16 tap drill on the other. That HAS to work properly...

by **rsilvers** » 2013-Jun-Wed-21-Jun

I am ready to print the JohnSL solution when it is on Thingiverse.

by **JohnSL** » 2013-Jun-Wed-22-Jun

I've just uploaded the parts to Thingiverse: http://www.thingiverse.com/thing:99364

Please let me know how it works for you. One of the outstanding questions is if 1.5 degrees is the right amount of fix for everyone.

Thanks,
-- John

by **ahaer** » 2013-Jun-Thu-05-Jun

by **JasperJ** » 2013-Jun-Thu-06-Jun

by **DumpRun** » 2013-Jun-Thu-09-Jun

I disassembled my Simple (Less than a week after getting it in the mail) to take measurements of the Z-carriage arm so I can design and print my own. I just received my All-in-one Stepper motor/Lead screw assembly in the mail yesterday, but I think I will first try to use JohnSL's solution in order to get my printer to print accurately before I proceed with replacing the Z-screw assembly and support arm. I figure if I can get it to print the parts better I will have an increased chance of success, and be able to reduce the number of times I will have to make revisions. I will also be adding lithium grease to the 5/16 screw as I have noticed that there is a significant amount of carbon dust sitting on top of the Z-motor plate (indicating wear/friction).

by **JohnSL** » 2013-Jun-Thu-10-Jun

I forgot to mention anything about printing. The bearing block requires support material, and I used Kisslicer to make the parts. Kisslicer seems to do a good job of adding support material that isn't that hard to remove. Here are a couple photos that show the before and after:

P6041626_Cropped.jpg

I used a small flathead screwdriver to carefully pry out the support material:
P6041627_Cropped.jpg

by **rsilvers** » 2013-Jun-Thu-12-Jun

The parts won't slice with Slic3r. I used Netfabb to fix them, and then they worked. You should repair them and re-upload them.

by **rsilvers** » 2013-Jun-Thu-20-Jun

I made it. Then I made some changes and uploaded them:

http://www.thingiverse.com/thing:99845?

by **JohnSL** » 2013-Jun-Thu-23-Jun

Ah, now I know I need to check that Slic3r can slice the files before I upload them to Thingiverse. Thanks for letting me know rsilvers. I've uploaded a fixed version to Thingiverse.

So, how did it turn out on your machine. Is your print bed now flat?

by **rsilvers** » 2013-Jun-Fri-00-Jun

Yes. My bed is flat now. I think this is something people should do.

My calibration print:

by **ScrappyMagoo** » 2013-Jun-Tue-20-Jun

Sweet, I think this will help, since I've been leveling my bed according to sag of the "y". Another option I was considering was putting a counter weight on the "y", but wasn't sure if the "z" motor had enough torque to do the job.

by **UndCon** » 2013-Jun-Thu-11-Jun

Cool - I downloaded your fixes for the Y-axis.

Now I need a mount for a Dial indicator...

by **errehache** » 2013-Jul-Mon-07-Jul

I hope this is the right place to ask this question... My setup does not print any support material. I am sure that I checked the required options in repetier-slic3r. Any ideas? (I am using a mac setup osx 10.8.4)

by **JasperJ** » 2013-Jul-Mon-15-Jul

by **elkniwt** » 2013-Jul-Wed-08-Jul

by **ahaer** » 2013-Jul-Wed-09-Jul

That is GREAT that you figured out where to hook in a fix!

I'm dying to get home and try it out!

Here's the link to Thingiverse http://www.thingiverse.com/thing:119204 . The link to the original model on tinker cad seems to be broken but the model and the source file are in the "thing" itself

To use this level the bed as normal (ie make the hotend "almost touch" in all four corners) and the software will "push" the model forward as it goes up so that the model winds up perpendicular to the sloping bed.

One day I'd like to figure out the firmware to add this functionality internally

by **elkniwt** » 2013-Jul-Wed-09-Jul

by **ahaer** » 2013-Jul-Wed-09-Jul

by **elkniwt** » 2013-Jul-Thu-14-Jul

by **ahaer** » 2013-Jul-Thu-15-Jul

by **elkniwt** » 2013-Jul-Tue-12-Jul

It looks like I can post links now. I'll go ahead and do this.

by **cmcook** » 2013-Sep-Mon-06-Sep

Thanks for creating these parts guys. I've just got a simple and have printed out a few parts, including some useful bits. I've managed to print out the leaning tower or printrbot and applied the software fix but running slic3r via the perl code in windows doesn't seem to slice properly. So I would like to print these parts out for a hardware solution.

Now my simple is the latest version and appears to have a little less sag (1.2degrees) than the 1.5 degrees these are drawn at so thought I would modify them using the solidworks files. Unfortunately they are in a newer version than the SW2010 I have. Is it possible to save them in an older, compatible version please?

Thanks very much,
Chris

by **JohnSL** » 2013-Sep-Mon-10-Sep

Unfortunately, there is no way to export to an older version of Solidworks.

I can certainly provide a new version with a different slant. My suggestion would to first print out and try the 1.5 degree part. When I measured the slant on my Simple V1, it was 1.25 degrees, but rsilvers reported it as 1.5, so I made the part for 1.5 degrees. When I installed the part on my Simple, I discovered that it was perfect. And the same may happen for you. If it doesn't, I'll modify the part and upload a new version.

-- John

by **plastiqmanb** » 2014-Jan-Sat-22-Jan

by **plastiqmanb** » 2014-Jan-Sat-22-Jan

ok maybe i spoke too soon... was there something programmed in my Simple V2 that counteracted the sag? I mean i was getting level prints with the bed on a 1.5 degree angle but i didnt want to keep it that way. Now look at the print.... this should be a straight vertical wall.

Quick Edit - i printed a 50cm square test print and it came out perfect, but it was only 5mm in height. Going to run some more tests here to find out whats going on.

by **thawkins** » 2014-Jan-Sun-01-Jan

by **trobb13** » 2015-Jan-Mon-09-Jan

I just shoved some toothpicks between the zipties, works great!

by **pdkline** » 2015-Oct-Sat-22-Oct

I was experiencing serious Y sag. I ordered a zip tie gun from Amazon (would you believe Home Depot didn't have one?) but didn't want to wait the two days for arrival. Did the toothpick thing...brilliant. Worked like a charm. Huge difference in sag. I'm going to do the same on my x axis bearings to reduce the play in those as well.

by **thawkins** » 2015-Oct-Sun-11-Oct

Printrbot Talk Forum

Fixing Sagging Y

Who is online