1 Topic by mdrVB6

  • mdrVB6
  • Member
  • Offline
  • Registered: 2013-12-18
  • Posts: 717

Topic: Pros and Cons of various Z-axis anti-backlash upgrades

Hi everyone, in the ongoing quest to make my printer better I have decided that it is time for a Z-axis upgrade.  I’m getting some banding in general and some parts have a defect layer at even spacing, so I know its time.  I am torn between various choices so I would like to hear your input on the pros and cons.  I have an aluminum bed.
First, let’s see if I can get my terms straight.  There is a round threaded hole on the bottom of the aluminum bed that the stock 5/16” threaded rod moves through.  I think this is called the “boss nut”.  I understand the basic concept of backlash.

1.     I could keep the stock 5/16” threaded rod and put an anti-backlash nut with spring on it.  Actually, this is probably not an option for my setup, but I wanted to say that I considered it.  With the Lawsy carriages (which I love) and the stock extruder, I have only 12mm of threaded rod available when the bed is at the home position.  All designs of this style that I have seen are taller than 12 mm but if I could find one that would fit, I would try it.
Examples:
•    http://www.thingiverse.com/thing:151011
•    http://www.thingiverse.com/thing:232253

2.    I could go with a fine threaded M3 or M5 rod with an anti-backlash nut system.  Is it correct that this type of system just leaves the boss nut in place but it never gets touched since the M3 or M5 threaded rod is smaller?  I don’t see any reason to remove it and you just risk damaging something.  I can see how a two nut system can eliminate backlash.
Example: http://www.thingiverse.com/thing:101308/#comments

3.    I could go all out and attempt to do a threadless ball screw mod.  This seems like there are multiple points where I could fail but I’m willing to give it a shot.  I have several questions that I would need to resolve before attempting this.  First, 2n2r5 said he ended up using an 8mm rod.  Well, the stock rod is 5/16” = 7.9375mm, so would I have to sand the hole in the aluminum bed to get it to at least 8.00 mm?  Secondly, was the issue of people getting slipping on the up direction and/ or different steps per mm in each direction ever solved?  Adrian had the last post on the discussion here several months ago and I’m assuming that if he or any of the other highly experienced people giving input in that thread had ever solved it, it would have been mentioned.  It seems that maybe this mod is not yet ready for the masses.
Example: http://www.soliforum.com/topic/3343/thi … -2-techie/

It is most likely that I end up going for an M3 or M5 threaded rod with 2 nut anti-backlash.  So should I do M3 or M5?  I know the pitch varies but I should be able to figure that out and compute the steps per mm.  Would I be okay with a standard steel rod from McMaster or do I need to spring for a stainless or zinc coated rod?  Is there any mods out there for a M3 or M5 rod that use a spring system?  I kinda like this idea more than a double nut system because I can tell if my spring is tight easier than I can tell the nuts are threaded on top of each other exactly correctly.

Any other thoughts on the most current tech for anti-backlash Z-axis mods would be greatly appreciated.  Thanks!

SD4 w/ RUMBA, E3D Volcano, all bearings, glass bed

2 Reply by AZERATE

  • AZERATE
  • AZERATE
  • PRINTiT
  • Offline
  • From: Chicago
  • Registered: 2013-11-28
  • Posts: 2,077

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I've been using http://www.soliforum.com/topic/2727/fix … -consumer/ for about a month now and I noticed a significant improvement over the backlash brackets that use a spring.
I would recommend reading 2n2r5's thread on the slop nut.

As far as the Z rod goes, I am eyeballing this as my next mod...particularly an M5. I've read on other pages that a threadless rod can be tricky to get right, and is prone to slipping if not installed correctly. Then again, would wouldn't want to install any Z rod incorrectly. smile The possibility of slipping is the only thing keeping me from trying out the threadless.

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!

3 Reply by wardjr

  • wardjr
  • wardjr
  • Moderator
  • Offline
  • From: Minnesota
  • Registered: 2013-06-10
  • Posts: 5,624

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I have been using this well-nut for a very long time (upside down).  It probably isn't the best solution but along with using the correct math for 5/16 rod (.2963 instead of .3) for z-heights has yielded a fair amount of improvement.  It is quick, easy and available at almost any hardware store.  Just spin it on, no need to fasten to the aluminum it doesn't turn.  My plan was to use this temporarily until I could switch to m5 rod and it has worked well enough that switching over hasn't been a concern.  That doesn't mean that an m5 wouldn't be better because it would.  I would still put a well nut on an m5 rod as it is very effective at removing backlash.


http://i.imgur.com/71tbpke.jpg

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

wardjr's Website

4 Reply by LdyMox

  • LdyMox
  • LdyMox
  • Member
  • Offline
  • From: Buffalo, NY USA
  • Registered: 2013-04-27
  • Posts: 576

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I have a similiar well nut on mine also. I think I have actully 10 extras lying around because I had to buy them in a pack.

Tisk, tisk wardjr...your machine looks dirty heheeheh tongue

Tammy
Solidoodle 2
E3Dv6 Hotend, MK5 v6 version, Glass Bed, Anti Z backlash slop nut, SureStepr SD8825 1/32 Extruder Driver, makeshift breakaway plexiglass case; . L-Cheapo 3.8 Watt Laser Attachment w/Custom built enclosure
From Buffalo, NY, USA

5 Reply by mdrVB6

  • mdrVB6
  • Member
  • Offline
  • Registered: 2013-12-18
  • Posts: 717

Re: Pros and Cons of various Z-axis anti-backlash upgrades

Okay, if I can get this to work it will be by far the easiest way to deal with backlash.  It looks like this is what I want.

http://www.mcmaster.com/#93495a431/=srkh0b

It might be worth a trip to home depot to see if I can pick up a single piece out of their boxes.  However, it seems like I still might be screwed by having only 12mm of threaded rod available when the bed is homed.  This piece is 18.4 mm tall when uncompressed, so it might fit if I can sand or trim part of the bottom off while keeping it perfectly flat.  Might as well try it, I'm about to order unless anyone knows of a place that sells a shorter version?

SD4 w/ RUMBA, E3D Volcano, all bearings, glass bed

6 Reply by justsomeguy

  • justsomeguy
  • Member
  • Offline
  • Registered: 2012-10-31
  • Posts: 302

Re: Pros and Cons of various Z-axis anti-backlash upgrades

My Solidoodle 2's threaded rod barely had any clearance above the bed either.  I used a washer over a rubber o ring, instead of a spring, to allow some flex and used this to hold the nut in place. 

http://www.thingiverse.com/thing:34397

This one is also good for those with little clearance:

http://www.thingiverse.com/thing:36177

For me, all the M3 rod did was increase the frequency of the banding.  It was worse than the stock rod in every configuration I tried.

I just upgraded to the threadless ballscrew and it is the best solution I have found yet.  Why is that thread not sticky?!  Yes, the up and down movements are different rates but I just turn off all options that would result in z lift and it works great.

7 Reply by wardjr

  • wardjr
  • wardjr
  • Moderator
  • Offline
  • From: Minnesota
  • Registered: 2013-06-10
  • Posts: 5,624

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I have seen users put it the other way and have good results.  Just not as much rubber to compress but it probably doesn't need that much.

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

wardjr's Website

8 Reply by mdrVB6

  • mdrVB6
  • Member
  • Offline
  • Registered: 2013-12-18
  • Posts: 717

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I was able to pick up a well nut from Lowe's on the way home for $1.39, nice.  It is 18mm tall, so it only has about half the threads on the rod when the bed is home but it still travels with the bed all the way down and seems to stay tight to the bed.  I've got a test print going now.  I can see how it would also work upside down with less rubber to compress.

As for the threadless ball screw, I agree that it looks awesome, but like I said there are several way that I could fail so I consider it a high risk mod.   Once I take the 5/16 rod off the motor, I will probably never get it to be quite the same.  New stepper motors are not that expensive, so I could just get another one to play with various rod configurations until I find one that I like.  I'm also concerned about ever getting a strong enough print to keep everything from slipping.  There was one guy on that thread who posted about making it on a CNC mill from metal, but he seemed to vanish.  Too bad, because I bet he would have made a little chunk of money if he cranked them out in metal, set up a simple eBay store and advertised them here on the forums and elsewhere. 

For those who are using an M3 or M5 rod, did you just get a 5mm coupler and tighten it around the threaded shaft and on the motor shaft?  Does it grip okay onto the threads?

Thanks.

SD4 w/ RUMBA, E3D Volcano, all bearings, glass bed

9 Reply by justsomeguy

  • justsomeguy
  • Member
  • Offline
  • Registered: 2012-10-31
  • Posts: 302

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I ordered a metal M5 to M3 coupler from eBay when I was testing.  It was poor quality and introduced extra wobble. 

I then sized and printed a parametric coupler from thingiverse and it held well and wobbled much less.  The banding was reduced compared to stock threaded rod but was still worse than the stock threaded rod with my anti-backlash solution.  It was still unacceptably bad to me.  Others seem to have had great luck  though.

I bought an extra stepper so I could leave the stock rod attached to the original motor.

The only challenge to the threadless ballscrew with your aluminum platform is that you have to remove the nut.  If you get a second stepper for this and print a 5/16 nut holder, you can go back to the stock solution at any time.

Apparently, newer firmware releases now have an option where you can set steps per mm for upward movement to X and downward movement to Y to deal with the threadless ballscrew issue.

In answer to your last question, I found that both CNC'd aluminum and printed plastic couplers had plenty of holding power and did not slip when properly tightened.

10 Reply by grob

  • grob
  • grob
  • Member
  • Offline
  • From: Sydney, Australia
  • Registered: 2014-01-02
  • Posts: 900

Re: Pros and Cons of various Z-axis anti-backlash upgrades

I've been playing with this last couple of days.

Current config:

  • SS M5 rod, not straight grrr

  • Home-made rigid coupling that has an M5 thread in one end and a 6mm drilled w/ set screw in the other.

  • Punched a 22.5mm cutout hole into the bottom of the case, and mounted the motor from the bottom so it's outside the heat (I have a raised bottom-layer on my case for electronics)

  • Single M5 nut rigidly mounted to alu frame, removed PEM nut (I had trouble with the slop nut...)
    M99 backlash compensation for Z set to 0.6mm (!!!)

Results are:

  • With software compensation, resultant backlash acceptable for neat bottoms on prints, positioning seems fairly repeatable with dial gauge (+/- flex)

  • Quite a lot of banding at thread pitch (bent rod I think)

Possible problems:

  • Backlash of 0.6mm is too high to be left alone mechanically. I have three theories for backlash sources:

    • Thread - machined delrin nut might be in order, or a spring-arrangement - but this can't be much more than .2mm given the thread pitch is 0.8mm, so something else is likely lower-hanging fruit...

    • Motor bearings - there is a little bit of slip in there, not sure why (maybe the case isn't tightened down properly!!) - could isolate this with a bearing pack at the bottom of the threaded rod (extreme?).

    • Tilt in the bed - improve lubrication of z smooth rods so gravity always wins, or swap out for 10mm linears. This would also help with (2) I think, if I can get gravity to win all arguments.

  • Banding is due to bent rod - I can either get a straighter one, or perhaps deal with the rigidity problem directly by putting a third z-rod up the front of the case (e.g. under one of the y idler blocks), and linking this to the bed frame. Not convinced by current anti-slop nuts, but this might be an easier first-step...

Lessons learnt:

  • Make sure the rod you get is straight

  • Backlash may not come from only the threads - e.g. my dodgy motor. Also flexibility of the axis design and case can contribute somewhat. An anti-backlash nut may not remove all backlash...

  • While minimising mechanical backlash is fun, don't fear the use of software backlash compensation - it works well for these issues, and particularly as an interim solution

  • When you have the nut off, investigate how smooth the motion of the bed actually is - making this nice and slippery may successfully avoid the impact of other problems (especially rigidity).

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