1

Topic: SD4, Z dimension short

Hello,

Over the weekend, I plan to upgrade my Y-axis shaft bushings to bearings. I'd also like to put on double printed pillow blocks. The problem is, my z-axis dimensions are a bit off, 10mm cube prints at 9.6mm physically. I've got the original Z-axis leadscrew and 2268 z-steps in the firmware. And I'm printing at 0.1976 layer height.

Question to the experts, what are my options? Adjusting z-steps doesn't seem right...

2

Re: SD4, Z dimension short

Answering to myself, maybe I was too fast asking questions again.... just found a thread (http://www.soliforum.com/topic/2341/poo … -required/) and this got me thinking I need to check backlash of z-axis first and maybe the upgrade of my z-axis leadscrew to 3mm will have to be my first mod after all.

sorry for noise

3

Re: SD4, Z dimension short

you may try setting your Z axis steps/mm to 2267.72 first if you still have the stock 5/16 rod... Solidoodle just rounds everything off to the nearest whole number...
I changed mine and it is much better.

SD4 #1 & #2 - Lawsy carriages, E3D v6, Rumba controller board, mirror bed plate, X motor fan, upgraded PSU & Mica bed heater
SD4 #3 - in the works ~ Folgertech FT-5, rev 1
Printit Industries Beta Tester - Horizon H1

4 (edited by grob 2015-05-11 01:01:57)

Re: SD4, Z dimension short

Investigating backlash sounds like a plan - I highly recommend purchasing a dial indicator to be able to measure it directly, it takes a lot of guesswork out of the process.

Note that 10mm / 0.1976mm/layer = 50.607 layers, but the printer can only make an integral number of layers! Make sure you calculate the actual height from your slicer layer count (including the first layer height if it's different!), as it might not be exactly 10mm with your settings.

(If you change to an M5 or M3 rod, then the above problem goes away as you can go back to using 0.2mm layers all the way up)

heartless, I'm a little sceptical that the difference between 2267.72 and 2268 steps/mm is part of this height problem: this would be 0.012% error, and in the wrong direction... That said, what was better in your experience exactly? Was it a banding problem?

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: SD4, Z dimension short

I never really had much of a banding problem (lucky i guess) and what little I did have seems to have been resolved with a stabilizer on the front of the bed keeping it centered & stable, not wiggling around - this was done long before the Z step change. My layers do seem to be sticking together better, and it is a bit closer dimensionally than it was at the stock setting - maybe it is my imagination (a "believe in something" kind of thing) but it seems to work for me...

Before the change I was getting occasional layer cracking with ABS, after changing I haven't noticed any cracking - but that could just as easily be the items that are being printed, too...
I do most of my printing at the '0.3' layer height (0.2963 actual setting) so that may make a difference as well.

I made the change in Z steps mostly as an experiment - had seen it posted about somewhere (and several of the online calculators come up with the same number for the stock SD setup), and figured I would give it a try.

Might have to do a little experimenting with these settings... slice an object at the stock settings and again at the modified settings and see if there are any differences in the number of layers produced...

SD4 #1 & #2 - Lawsy carriages, E3D v6, Rumba controller board, mirror bed plate, X motor fan, upgraded PSU & Mica bed heater
SD4 #3 - in the works ~ Folgertech FT-5, rev 1
Printit Industries Beta Tester - Horizon H1

6

Re: SD4, Z dimension short

That first layer is usually pretty squished, so that's to account for some of the loss in Z height as well.  If you're printing with a raft then that's not an issue, but I've noticed these beds don't properly move down for each layer unless you specifically address it.

Recently on my printer I messed up some finish gcode and the bed crashed ha.  It broke from the screw, not too much damage luckily.  Upon raising it back up by hand there were plenty of spots along the guide rails it could hang up and sit in place without anything actually holding it up.  These bushings aren't worth much and I plan on replacing them soon with linear bearings.

Basically my point is, you can also have losses in individual layer heights from poor bed movement.  I hope you are able to solve your problem!

Bowden SD3, Rumba, E3D hotend, Mk5 with RtRyder changes, Direct drive Y axis and bearings, GT2 pulleys and braided fishing line, Lawsy linear bearing conversion, M3 Z screw.

7 (edited by jagowilson 2015-05-11 17:12:35)

Re: SD4, Z dimension short

Measuring prints to determine the precision of your printer's axes is kind of hairy, anyway. Slicers make compromises and approximations, the plastic shrinks some after printing, and particular geometries tend to be under-sized just because of the nature of filament printing (inner holes for example). Adjusting steps per mm based on measuring prints is also not recommended for these reasons.

Grob has the right idea. If you want to assess this with any seriousness you will need a dial indicator. But let me just point out that if you are happy with your Z axis as-is, this measurement may mean nothing. Are you having a particular issue caused by short Z height? Cracking is delamination and is almost always a thermal or flow problem.

8

Re: SD4, Z dimension short

The original thinking went, that if I print a pillow block that's off by 0.5mm in height (z-dimension), my 6mm y-axis will bend. Or possibly the backplane.

I've been setting up my printer further and the real reason for the poor dimensional z-accuracy seem to have been warping. Yesterday I got new hairspray (with polyvinyl, haven't found aquanet locally as of yet) and I no longer have sticking problems. And that seems to have corrected most of my z-dimension problems as well.

So, now I've got all the parts printed/shipped from china for
- y-axis bearings + GT2 on motor belt - mod
- z-axis slop nut + M3 lead screw - mod

heartless, what's the bed stabilizer mod you're referring to?

9

Re: SD4, Z dimension short

there are a multitude of bed stabilizers out there, but I used this one: http://www.thingiverse.com/thing:195266
because it was simple, yet very effective, and I already had the bearings it uses on hand (RC car enthusiast) as well as some brass tubing that worked quite well in place of the carbon fiber rod (I did need to ream out the holes just a tiny bit to fit the brass).
I also added a dab of silicone glue to the parts that attach to the bed frame to keep them from vibrating off periodically (during infills) - works great.

you do lose a little bit of Z height with it due to it needing to be snug against the frame, and the bottom part of the frame sits inside the vertical rails the bearings ride on, but I have yet to try printing anything that would require the full 200mm height anyway. I think I reset my Z height to 175...still a respectable height for my needs

SD4 #1 & #2 - Lawsy carriages, E3D v6, Rumba controller board, mirror bed plate, X motor fan, upgraded PSU & Mica bed heater
SD4 #3 - in the works ~ Folgertech FT-5, rev 1
Printit Industries Beta Tester - Horizon H1