A Cheap Compact Linear Motion Slide

A "linear motion slide" allows precise positioning along precisely one axis of motion.

This Instructable will show how to build a usefully stiff and precise linear motion slide from very low cost materials, mainly 1/8" hardboard & 6 mm rod, and access to a laser cutter.

As configured in the lead photo, the slide uses a wheel on a screw to reduce tangential motion at the edge of the wheel to linear motion of the slide by about 100:1, turning small human motions into tiny slide movements. The video below shows the photo model used for most of this 'ible repeatedly returning to within ±5 μm (indicated) of a set position — while carrying a 30 lb weight (± 2/10,000 inch; 14 kg). [1]

Stackable

If one axis of motion is good, two must be better! Two copies of the example configuration stack together trivially, one across the other, for precise 2D positioning. Stick them together with your favorite double-stick tape, hot-melt glue, Blu Tack®, etc. Other configuration options include more elaborately matching surfaces with tabs for fastening a pair of slides together. The video below shows mostly the 2D action of the 3-axis application that spawned this design. (stepper motor option introduced below)

Telescopic

More often the moving part of a slide runs on bushings/bearings that have some fixed span in the direction of motion. Wider span gives greater rigidity but reduces range of motion within a given overall dimension. Narrower span allows wider range of motion but reduces rigidity. A telescoping design like this turns that range-rigidity trade-off into a run-time option: greater rigidity within limited range, or longer throw if you can accept less rigidity. I'm happy to add that this design turned out much more rigid at full extension than I expected!

This Instructable

This 'ible got huge. How much you end up reading will depend on what you want to make. Reading straight through about half of it will show how to build the example in the lead photo with the white mug. The rest describes variations on that theme.

Steps 1-38 may be all you need. They show how to build the photo example. This part includes the core sliding mechanism shared by all other variations. While mechanically simple, the core could be kind of a “bar bet” puzzle to put together without any guidance. This section uses lots of steps and pictures to walk through an assembly sequence that I hope can take you straight to a good result on the first try. The flag "※" marks notes about what might change when building a different variation.

“Variations” include swapping the hand-crank leadscrew for a stepper motor + leadscrew assembly. Actually, this design started as a powered slide and the manual variation was an idea for dividing the basic slide design from the details of putting a motor in it.

Steps 40-52 show what stepper motor fits, how to prepare it, and how to make a spring that you will need. Then that part will send you back to the first part to build a slide while showing where to fit the motor into the process.

Step 55 describes options for “accessorizing” a slide with features for attaching it to something, or something to it, or two slides together for 2D motion.

The last section introduces real live danger by showing how to assemble and use a clamp for holding a Dremel-like rotary tool.

I've tried to make this 'ible complete enough to follow from a cold start without booby traps or puzzles to solve. That probably means both over-detailed and still under-detailed. If building a second or more slides, your process may evolve. There is no warranty to void by doing something differently.

Feedback, Please?

This is the cheapest way to build a comparably capable device that I know of, and I hope it will enable/encourage someone to find other applications and improvements. Please give it a try and share how it goes.

[1] Yeah, mixed units throughout. Sorry, but metric bearings and UNC threads. No Mars landings will be attempted.

(22 May 2021 - "side" options & tool clamp added; attempt to emphasize distinguishing between variable & constant step details)

(23 April 2021 - the slide design has been slightly changed and this 'ible overhauled since first published in March to simplify the assembly process)

Skip the "long form" if this works for you:

• Cut parts from a 8" × 10" piece of 1/8" hardboard (200 × 250 × 3 mm)
  • "hardboard" ≈ "masonite" ≈ HDF
  • instructions assume single-sided hardboard with one "good" face
  • measure actual thickness
• Use one of the SVG files attached to this step
  • choose the smaller of "3.0mm" or "3.5mm" that is not less than the actual thickness of your material
  • your laser software may
    • require setting line widths to hairlines
    • be able to sequence cuts by color in a single operation
  • disregard cut sequence if it's a problem
• Place the hardboard with its good side down in the laser cutter

Skip this if the "short form" did the job:

To make the flat parts you will need material, a vector file, and some time on a laser cutter.

Material

• So far I’ve used only single-sided ⅛” hardboard for several iterations of this design, and for drafting these instructions. Also called “masonite” or HDF.
• Roughly 8” × 10” × ⅛” (180 mm × 230 mm × 3 mm). ※Take actual dimensions from the vector file before cutting.
• Measure the actual thickness, perferably with a caliper, in several places and note the high value. Less variation is easier to deal with than more variation.
• The CAD model allows material thickness from 2.5 mm to 4 mm. I’ve used material from 2.9 mm to 3.3 mm.
• If using material that is finished on both sides, choose a “good” side of the material and a way to keep track of the “good” sides of cut parts.
• Other materials between 2.5 mm and 4 mm thick “should” work, but may behave differently in details than assumed in the following instructions — especially in regard to thread forming and the screw-in-tab-in-slot joints. Without rigor, I imagine that if this design can work with single-sided hardboard then it can be made to work with “better” material. Harder material may require a tap or thread-cutting screws; one trial cut from acrylic + tapped threads worked.

Vectors

• A vector file defines the part outlines to cut.
• For a given configuration, the exact shapes of parts to cut depend on the expected thickness of material from which the parts will be cut.
• There are a couple of prepared SVGs (vector files) attached here for material up to 3.0 mm or 3.5 mm thick.
• If your material measured 3.0 mm thick or a little thinner, use the “wheel2r-3.0mm.svg” file. For material over 3.0 mm up to 3.5 mm, use the “wheel2r-3.5mm.svg” file.
• These files group the cut lines by color. Cutting in the indicated order will cut interior features of parts before cutting the parts loose from the bulk material. On the other hand, collapsing the vectors all into a single group will probably work ok.
• If your material measured over 3.5 mm thick, or if you want a closer fit for your measured thickness, you can generate a new vector file from the CAD model in Onshape. Select the “cut lines (r.click->change config)” drawing tab, right click on the 2d drawing of all the parts, select “Change configuration…”, and update the “thickness” parameter. Create a free ("public maker plan") account if needed to change the drawing configuration.
  
• To match the details shown in these instructions, verify selection of “handwheel on 2rod side” for “drive”, "stackable-simple" sides and activation of the “practice pieces” checkbox. Click the green checkmark (you may disregard the “Generate” button). Then right-click the drawing tab and select “Export…” to download the drawing. You will probably want SVG format.
• The laser cutter software may take the SVG directly, but that file will have many overlapping segments. Some post-processing to remove the overlaps will make the cut job shorter and produce cleaner part edges. I don’t know a great way to do this without expen$ive $oftware. Please comment if you do. I’ve been using Inkscape with this extension to remove overlapping and consolidate contiguous segments. In Inkscape, after installing the extension, select and “combine” everything (Ctrl-A, Crtl-K) then select Extensions->Modify Path->Remove redundant edges.
• Check the vector scale before cutting. ※If you’re cutting parts for a slide with "plain" or "stackable-simple" sides, the short dimension will be 180 mm exactly. Common scaling error factors include 25.4 (mm/inch), 1/25.4, 3/4 (pt/px) or 4/3, or a product of two of those.
• For rough cutting material, if needed, and positioning material to cut, note the actual extent of the area to be cut, which depends on the material thickness and other options, and add some allowance for imperfect alignment of material in the laser cutter. ※Some options require a larger area of material.
• If you want to arrange the parts differently, keep the end plates with simple tabs and "V" shaped hole features (you'll see) together in the same relationship so that they get cut with similar imperfection.
• A combination of material and cutting technology that leaves much of a kerf will require a different part layout without shared edges.

Cutting

• Cut single-sided material with the good side down. (Assuming a cutting beam firing down from above the material.)
• For material finished on both sides, choose a “good” side of the material and keep track of the “good” sides of cut parts.

You will need a few bits of hardware and tools in addition to the flat parts.

This design relies on "linear motion bearings" running on "linear motion rod/shaft" as a foundation for precise linear motion. Both in 6 mm diameter.

Bearings: While the bearings are the most "special" manufactured item required, this design uses a standard size available in many kinds from many sources for a wide range of cost. You could obliterate the "cheap" aspect of this design by splashing cash for marvelous parts, or you could print your own sliding bushings for nearly nothing[1]. I've been using cheap recirculating-ball bearings from AliExpress. While they are inconsistent and few feel really smooth, their competence is that they don't get worse under load. Instead they often run more smoothly under load. I won't try to say anything about lubrication beyond 1) I've applied light machine oil sparingly to each ball race & 2) here's a link.

Rods: For good fit with linear bearings, linear rods "are turned, ground, and polished to tight diameter and straightness tolerances". While familiar applications like 3d printers usually use hardened and plated steel, a hand-cranked hardboard slide will likely do less demanding duty. I've used tightly rolled paper to assemble a weak but working slide. Mostly I've used pre-cut 100 mm lengths of the usual hardened plated rod, again from AliExpress. These are the most expensive component of an un-powered slide to purchase, but old printers, copiers, and such often include 6 mm shafting available for salvage.

New bearings and rods may be coated with a waxy preservative. If so, WD-40 washes that away.

All screws & nuts have #6-32 UNC threads[2]. PC cases have used #6-32 × 1/4" screws for ages, making them a good candidate for salvage.

Check here for occasionally updated part sources (numbers there show costs for three slides).

Parts:

• flat parts - laser cut from 1/8" (3 mm) hardboard
• (×3) 6 mm × 100 mm rods — "linear motion shaft"
• (×3) LM6UU bearings
• (×25) 1/4" non-countersunk screws (#6-32 × 6 mm screws exist)

※for the example with manual leadscrew:

• (×1) leadscrew
  • either allthread cut to 3⅞" < length < 4⅝" (100 mm ≤ l. ≤ 115 mm)
  • or long screw (~3"ish - limited range of motion)
• (×5) hex nuts
• (×2) #6 flat washers (o.d. ≤ 13/32" / 10 mm)
• not pictured:
  • random weak adhesive (e.g. cheap glue stick, but whatever - not pictured)
  • random machine lubricant (e.g. light machine oil - not pictured)
    • + a way to apply very small amounts (e.g. toothpick)

※Optional revolving crank handle for handwheel:

• 3/4" or 1" non-countersunk screw
• locking nut (nyloc, cap nut, nut+CA, etc.)

• stiff tube-ish bit to fit over screw between wheel and locking nut

Tools:

• screwdriver (that fits the screws)

possibly, depending on material:

• sharp knife/scraper or other improv. - if parts are "too thick"
• coarse sandpaper & flat surface or other improv. - if parts are "too thin"

※if building the example with manual leadscrew and handwheel:

• (×2) open-end wrenches (that fit the nuts)
  • typically 5/16" for #6-32 hex nuts
  • adjustable wrench pictured because I had one 5/16" wrench
  • ideally one would be thin e.g. a tappet wrench — but I didn't find any source of 5/16" tappet wrenches (8mm might work)
• needle-nose pliers
  • very wrong tool, but more common than 5/16" tappet wrenches

[1] But casually assuming use of a 3d printer on top of the assumed laser cutter, kinda makes the whole "cheap" idea smell funny. Ditto lathe, etc. From the beginning of thinking thoughts that evolved into this, I ruled out requiring more than one fancy tool (up from zero). So, laser cut parts -> common hand tools and a little bit of cash for the rest.

[2] Screw clearance and tap diameters can be adjusted in CAD, and the model resolves for M3 and M4 dimensions, but I have only glanced at the results.

The laser-cut parts include a pair of intentionally expendable parts for learning how fit and fastening will work with your material. These pieces include lots of holes and tabs to consume for destructive learning. The smaller holes are nominally the size of a #6-32 tap drill.

Use them to discover at least three things about your material before starting with the real parts:

1. How the tabs fit into the slots
2. How to drive screws in your material
3. How tight to tighten screws

To accomplish that:

1. Check the photos to identify the two practice parts, and to make sense of the rest of this step
2. Turn one of the practice parts to aim its tabs at the slots in the good side other part and turn the good side of the "tab part" toward the tabbed edge of the "slot part" (the relevance of this orientation will emerge later)
3. Without damaging the tabs of the "tab part", try to fit them into the slots of the "slot part"
4. How did the tabs fit?:
   • just snug - sweet!
   • too thick to get the tabs in without some damage -
     1. first take care to prevent damage to the good side of the tabs
     2. if the tab fit but peeled up some fuzz from the rough side: scrape/cut away the fuzz so it doesn't continue to peel
     3. if needed, thin the tabs from the rough side just enough to allow a snug fit
     4. if you have to remove very much material from tabs that started out very much too thick, you may have other fit problems later and may need a better match between modeled and actual material thickness
   • too long so the tab ends stick out the other side of the slot part -
     1. tabs "tool long" really means the material is too thin (relative to modeled thickness) and the tabs will be thin for the slots too
     2. for a better test, slide the corner of an ATM/credit/ID-type card (0.03" ≈ 0.76 mm ≈ 1/32" nom.) in between the two parts, flat against the face of the "slot part" with the corner under the edge of the "tab part", to lift up the tab part and pull tab ends back into the slots
     3. if the tab ends still do not fall below or just even with the back face of the slot part then you will need to shorten the tabs — possibly by grinding them evenly on coarse sandpaper on a hard flat surface
     4. if the tabs are much too long then the slots are wide relative to the tab thickness and may not constrain the screw-tab-slot joint well enough
   • none of the above - ok
5. For each of the two kinds of joint:
   1. firmly drive a screw to fasten the joint
      • use firm pressure while turning for a clean start to forming thread in the hardboard
      • keep the screw straight upright - a good fit between screw & screwdriver helps
   2. when tight, continue turning slowly with care to feel how the threads progress from just beginning to yield, through failure
      • repeat at another joint or two until the progression of failure feels predictable
   3. at a fresh joint, drive a screw and tighten until just before the threads start to yield
      • remember what that feels like
   4. back the screw out and turn it in again to similar tightness, without overloading the thread, to learn the feel of tightening up a screw in previously formed thread.

Identifying parts by invariant features and brief names may help make the rest of this 'ible a little less wordy.

These photos have variable details masked off to highlight the features that identify the same core parts in all variations. While the finger-size pieces always have both ends the same, the single feature on one side of the centerline might be reflected to the other side in some cases. In other words: o-Oo = oO-o. (hmm... maybe 4 character symbols like that could be more terse and more clear... but unpronounceable... and reducing to 3 chars would abstract away asymmetry and this paragraph entirely... hmm... oOo & vVv? hmm... keeping the -/asymmetry might pay off in sequencing screws? hmm... food for next rev.)

• Set the smallest parts aside for now
• Lay out the remaining 10 rectangle-ish parts with the good side up
• Use the three images for this step to:
  1. Sort the eight ~finger-size parts
  2. Match up the two big pieces
  3. See how the terse part names work
     • In any given application, it may make sense to call the sides ‘top’ and ‘bottom’, or ‘fixed’ and ‘moving’, or ‘driving’ and ‘driven’, or some such. But the slide by itself has no orientation like that. It does have a side with one rod and a side with two rods. On each side, only one end has bearings. Of the two plates at each end, one has the “V” features that locate rods/bearings at that end, and the other “clamps” those parts into the “V” features.
     • example: 1r2bc = one-rod side / two-bearing end / clamp
     • example: 2r0bV = two-rod side / no-bearings end / Vs
     • etc.

Pre-form threads in some parts, so you can run screws through them later without force.

For the specific example configuration:

• 1r2bc for the leadscrew
• the handwheel center

The picture over-emphasizes perpendicularity of thread for the leadscrew, but this step needs a picture so until next rev...

The picture says:

• If you have a long screw, you can use that to get close to perpendicular by eye.
• If not, you can use a tap-size hole in one of the practice parts as a jig to hold a short screw straight without making it tight. (tightening up the screw in the jig makes it perpendicular to that part)

Hopefully you have a screwdriver that fits your screws and holding the driver somewhere near ⊥ to the material will suffice.

※(← the flag for details that depend on configuration):

• Thread for a leadscrew only matters when making a slide with a manual leadscrew. The leadscrew engages 2r1bc instead of 1r2bc when configured to the 1-rod side.
• This is a good time to pre-thread all tap-size holes through either side plate where present in other variations

First build the bearing end of the 1-rod side with just enough clamping pressure to keep the bearings from falling out.

Parts:

• the 1-rod side plate
• 1r2bc (clamp)
• 1r2bV
• 2 bearings
• 2 bearing supports
• 6※ screws

Tools:

• screwdriver
• 1 rod (as tool)
• any other "clamp" part (as tool)

1r2bc should be pre-threaded for the leadscrew from the previous step. ※Or not configured for a leadscrew.

Fit the clamp, V, bearing supports & bearings together.

1. Fit the flat parts together as shown
   • side plate good side up
   • clamp & V plates good sides together
   • bearing supports good sides toward clamp
2. Fit the bearings through the clamp/V pair and over the bearing supports
3. Check that you can hold the bearings by squeezing the clamp
   • you should be able to squeeze the clamp to hold the bearings so they don't fall out when you tip up the assembly
   • depending on how the parts fit, it may take some effort to actually clamp and hold the bearings even though the clamp looks closed down onto them

※While the outline shape of the side plate may vary, this step and much of the following process describe building the core common to all variations.

This step includes detail that future steps will assume without repeating.

In this step, place two screws to hold the clamp just tight enough so you can handle this assembly without the bearings falling out. The two screws shown can hold the bearings without clamping down on the rod.

1. Place the assembly upside down on a hard flat surface
   • for convenience, fit a clamp part in the other end of the side plate as a temporary tool to square up this arrangement
2. Fit a rod through the clamp as a temporary placeholder
   • don't worry about fitting the rod through both clamps
3. With a firm start as practiced earlier, drive a screw into the tab joint inboard of the bearing opposite the rod (see photo; the photo also includes the next screw that won't be there yet)
   • stop when the screw head just touches down, or nearly so
   • do not try to make it tight
   • ※The rod and the space next to it switch places in a "reflected" configuration.
4. Press down next to the screw to compress the clamp
   • see photo
   • do not rely on the screw to pull the clamp closed
5. While compressing the clamp at that point, turn the screw in to lock the joint
   • turn until the screw head makes firm contact with the side plate
   • that's all, don't try to torque it down hard yet
6. Repeat points 3-5 for the screw at the opposite end of the clamp, outboard of the other bearing
7. Check that tab ends remain below the surface of the side plate
   • see the green lines in the photo and inset detail
   • the screws can't hold the clamps closed if the tabs are too long (material too thin)
8. Check that the clamp holds the bearings in place
   • the clamp should secure the bearings against sliding out by their own weight
   • you should still be able to move/adjust the bearings
9. Align the bearings so that the groove around the exposed end just clears the V plate
   • the full diameter of the bearings must rest in the locating Vs
   • minimize extra "stick-out" beyond that
   • doing this now lets you see where the bearings should end up; they'll get bumped around until re-adjusted and tightened up later

1. Drive screws into the remaining three tab joints
   • just until the screw head touches down
   • then back out about 1/2 turn (see end-on photo - ※or its reflection)
2. While supporting the clamp from behind, drive a screw through the V into the clamp
   • back this screw out a little also, so that it does not grip the V against the clamp

The next few steps almost but not quite repeat the steps just completed.

Parts:

• the 2-rod side plate
• 2r1bc (clamp)
• 2r1bV
• 1 bearing
• 1 bearing support
• 6 screws

Tools:

• screwdriver
• 1 rod (as tool)
• any other "clamp" part (as tool)

Fit the parts together the same way as the first side:

• side plate good side up
• clamp & V plate good sides together
• bearing support good side toward clamp

Temporarily fit a rod through the end opposite from the bearing. This provides a second point to stabilize the clamp.

By squeezing the clamp you should be able to hold the bearing and the rod opposite from it.

In this step, place a screw to hold the clamp just tight enough so you can handle this assembly without the bearing falling out.

1. Place the assembly upside down on a hard flat surface
2. Fit a rod through the clamp end opposite from the bearing as a temporary placeholder
3. Drive a screw into the center tab joint, between the bearing and rod
   • stop when the screw head just touches down
4. Press down next to the screw to compress the clamp
5. While compressing the clamp at that point, turn the screw in to lock the joint
   • just firm, don't try to torque it down hard yet
6. Check that the clamp keeps the bearing in place when you tip up or handle the assembly
7. Align the bearing so that the groove around the exposed end just clears the V plate

1. Drive screws into the remaining four tab joints
   • just until the screw head touches down
   • then back out 1/2 or 1 turn
2. While supporting the clamp from behind, drive a screw through the V into the clamp
   • back this screw out a little also
3. Check that the bearing stays in place without the rod through the other end of the clamp
   • if not:
     • fit the rod though the clamp right next to the bearing and tighten the screw between the bearing and rod just enough to hold the bearing
     • check that you can still slide the rod out and fit it back through either end of the clamp again without difficulty.

Parts:

• leadscrew — one of:
  
  • measured length of threaded rod, or
  • long screw
• five nuts
• two washers
• IF using a long conventional screw for a leadscrew:
  • the first (1r2b) bearing end assembly

Tools:

• two wrenches
• two spacers
  • the two "practice parts", for example

  • or any two of the "flat parts" with screw clearance holes

※If not building with a manual leadscrew, go to variant procedure then come back to resume at Step 19.

1. IF you are using a conventional screw for a leadscrew:
   • run it through the previously threaded hole in the 1r2b clamp from the inward side, leaving the screw head inside
2. Line up the nuts, washers, and spacers on the leadscrew:
   1. two nuts
   2. everything else except one nut, in any order
   3. the last nut
3. Set the last nut exactly at the end of the screw
   • aesthetically: where it looks best is the right spot
   • functionally: use all of the nut thread with no extra screw thread sticking out to snag flesh while spinning the handwheel
4. Run everything else finger-tight against the last nut on the end
5. Use a wrench to jam the first nut against the second nut
   • with moderate torque at this point
   • unless you have a thin/tappet wrench to hold the second nut, in which case you can lock those up right now
6. Without disturbing the first two nuts, remove everything else from the screw
7. Using two wrenches, jam the two remaining nuts together hard to fix them in this position

IF you are using a conventional screw for a leadscrew

• it must be already threaded through the 1r2b clamp with the head end inward
• turn the screw to move the head close to the clamp so that it doesn't project inward beyond the ends of the bearings
  • closer than shown in the photo

IF using plain threaded rod for a leadscrew

• run the long end through the previously threaded hole in 1r2bc from the outside so that the jammed nuts remain on the outer side of that end
• continue turning it in until the jammed nuts are close to the clamp
  • closer than shown in the photo

Be careful with the long lever to avoid damaging the threads in the clamp.

Parts:

• 2r0bV
• 2 washers

Tool:

• random weak adhesive

Stick a washer to each side of 2r0bV, carefully centered around the leadscrew clearance hole (between two slightly elongated holes - see photo)

Parts:

• 2 partial sides with bearings
• 3 rods
• leadscrew (already installed in one side assembly) ※or not if not

Bring the two parts together as shown, with the bearings ends close together.

• If using an all-thread leadscrew, pass the tail of the screw through the clamp/V end of the opposite side.
• If using a conventional screw as a manual leadscrew, turn the screw to shift the head close to the end of that side to allow the bearing ends to come together as shown.

This and the next step include details that the following steps will assume without repeating.

From the last step, you should have the two slide halves arranged with the bearing ends close together.

1. Insert a rod through the single bearing on the 2-rod side
2. Carefully fit the end of the rod through its place in the 1r2b clamp
   • this should not require force but can be a little tricky
   • start the rod end through the clamp
   • angle & press up to get over the edges of the V plate
   • take care to avoid damaging the V edges
   • uniformly chamfered rods ends will help
     • but if building "cheap", the rods ends are probably not so nicely finished
   • if difficult:
     • remove the rod
     • find the least-sharp/most chamfered arc of either end
     • turn the smoothest edge of the rod "down" toward the side plate and start again
     • the smooth edge with help the rod end ride up over the V edges
3. Set the rod end about flush with the outside surface of the V plate

   • it sticks out a little in the photo just for visibility

After this step it will be tempting to start trying the slide action. There's no reason not to. But beware that if your bearings are anything like the cheap parts I've been using, it will probably feel awful. Despair not; it will get better as more parts get lined up and buttoned down.

From the earlier partial assembly of this end, two screws should be holding the bearings with light pressure and the other four should be in place but backed out a little.

First:

1. Remember the torque and feel of tightening screws before stripping threads on the practice parts
   • or review that step for a refresher

Then:

1. Shift the sides to separate the bearing ends apart
   • without separating the sides; stop before the rod comes out of the bearing
2. Place the slide on a sturdy flat surface with the 1-rod side and 1r2bc screws on top
3. Align the rod end flush with the outer face of the V plate
4. Align the bearings with the outer groove just clear of the V plate
5. Start with the same screw you installed first in this end (see photo - ※or reflection)
   
   • this time for keeps
6. Press down hard on the side plate next to the screw to compress the clamp
   • strong people: be reasonable
7. Turn the screw down to into solid contact with the side plate and about half of final torque
8. Compress and half-tighten the screw opposite the first screw, between the bearing and rod
   • this should put a fairly firm hold on both bearings
   • check/tweak bearing alignment
9. Compress and half-tighten the center screw
10. Compress and half-tighten two end screws
11. Check that the clamp has a firm hold on the bearings and at least some grip on the rod
12. Repeat the same sequence of compressing joints hard and tightening screws
    • tighten the screws firmly without stripping the threads
13. Squeeze the clamp, V and bearing supports together
    • they should be close already, but squeeze them up tight anyhow
14. While supporting the clamp from behind, firmly tighten the end screw
    • with the sides together now it's harder to support the clamp from behind
    • the photo shows an example of holding the ends of the clamp with thumb and middle finger to support the clamp while pressing into it to tighten the screw

The numbered diagram shows the screw tightening sequence described above. Note there are two 4s and two 8s. ※Remember to reflect the diagram & sequence if building a reflected slide.

tl;dr: ok to skip.

The steps here include a suggested order of tightening screws that should work.

If you wonder why, or what to consider when building a different configuration or otherwise doing something differently, read here. Otherwise carry on with the next step.

In this assembly sequence, we want to hold the bearings in place first without clamping down on the rods. Then tighten everything up later.

Each rod or bearing is a fulcrum; squeezing the clamp closed on one side of it pries the clamp open on the other side.

You know how levers work. A screw cranked down on the short arm first may get pried up really hard when the long arm gets cranked down later. Squeezing between two fulcrums is stable (applying more pressure to the closer thing). So start between clamped things rather than starting at one end and lifting other end which will have a 7:1 advantage when it get squeezed back down. As a second-order application of the same consideration, for clamps across three things I'll suggest starting with the inboard screw on the side that holds just one thing because that's the long arm of the lever across the thing in the middle. Like the first screw in the sequence just done.

Because real parts aren't geometrically perfect, the clamps that hold three things really hold only two of them. In all examples that I've built so far (>10), the clamps close down with pressure on bearings before rods. I suppose maybe because larger diameter bearings (more blunt) compress into clamp and V edges less than the smaller diameter rods (more "sharp"), but I don't know. In any case, that works out ok because each bearing is held by only one clamp while rods are held by two clamps, one of which has no bearings competing for grip. (Consequently the "second-order application" above probably doesn't matter for the 1r2b end.)

The clamp holding only one rod is free to rock across that rod. Keeping it about level adds style points but isn't really critical. To help, the end of the clamp opposite from the rod has reduced clearance for compression to keep the clamp from falling so far out of level if biased over to that side.

boring image CC0

Steps for this end assume the same details as for the first end.

1. Shift the sides to bring the bearings back together
2. Insert a rod through the bearing next to the leadscrew
3. Fit the end of the rod through its place in the 2r1b clamp
   • avoid damaging the V edges
4. Similarly install the last rod
5. Turn the slide over
6. Set the rod ends about flush with the outside surface of the V plate

From the initial assembly of this end, the center screw (and possibly its neighbor on the bearing side) should be holding the bearing with light pressure and the other five (possibly four) should be in place but backed out a little.

1. Shift the sides to separate the bearing ends apart
2. Place the slide on a sturdy surface with the 2-rod side and 2r1bc screws on top
3. Align the rod ends flush with the outer face of the V plate
4. Align the bearing with the outer groove just clear of the V plate
5. Start with inboard screw opposite the bearing
6. Press down on the side plate next to the screw to compress the clamp
7. Turn the screw down to into solid contact with the side plate and about half of final torque
8. Compress and half-tighten the opposite inboard screw, next to the bearing
   • this should put a fairly firm hold on the bearing
   • check/tweak bearing alignment
9. Compress and half-tighten the two end screws
10. Compress and half-tighten the center screw
11. Check that the clamp has a firm hold on the bearing and at least some grip on the rods
12. Repeat the same sequence of compressing joints hard and tightening screws
    • tighten the screws firmly without stripping the threads
13. Squeeze the clamp, V and bearing supports together
14. While supporting the clamp from behind, firmly tighten the end screw

1. Collect parts
   • the assembly so far
   • 1r0bc (clamp)
   • 1r0bV
   • 6 screws
2. Adjust the rod so the ends line up with the length of the side plate
3. Fit the 1r0b clamp in place
   1. Lift the rod end (prying against the bearing-end clamp)
   2. place the clamp with its good side facing outward
   3. fit the tabs into the side plate
4. Lift the rod again a little to fit 1r0bV into place with its good side facing inward (the two good sides together)

The rod should be getting harder to shift, but probably still moveable.

1. Check that the rod overlaps the Vs equally at both ends with the ends perpendicular to the side plate
2. Extend the slide so that downward pressure at the end compresses the clamp only and not the other half of the slide
3. Drive a screw into the center tab joint
4. Compress the center joint and half-tighten the screw
5. Drive, compress and half-tighten the screw on the other side of the rod
   • for maximum style points, adjust these screws to balance the clamp level across the rod and keep it level while tightening up the rest of this end
6. Continue with remaining three screws at this end end, working away from the rod
7. Repeat the same sequence and tighten each screw firmly without stripping the threads
8. Squeeze the clamp & V plates together
9. Support the clamp from behind and drive the last screw into this end through the V plate

The clamps should now hold this rod solidly in place.

1. Collect parts
   • the assembly so far
   • the last two end plates (2r0bc, 2r0bV)
   • the remaining screws (seven for this end)
2. Adjust the rod so the ends line up with the length of the side plate
3. Fit the 2r0b clamp with its good side facing outward
4. Fit 2r0bV with its good side facing inward (the two good sides together)
   • check for washers on both sides

1. Check that the rods overlap the Vs equally at both ends with the ends perpendicular to the side plate
2. Extend the slide
3. Drive a screw into the center tab joint
4. Compress the joint and half-tighten the screw
5. Continue with remaining four tab joints at this end end, working out from the center
6. Repeat the same sequence and tighten each screw firmly without stripping the threads
7. Squeeze the clamp & V plates together
8. Support the clamp from behind and drive two screws into this end through the V plate

The slide should now be constrained to one degree of linear motion only.

Check that the slide works well through its range of motion before connecting the leadscrew.

If you have a plain threaded rod leadscrew, you can easily run the slide through its full range of motion. The ends of the leadscrew may need some help to align with and pass through holes in the ends of the opposite side of the slide. If the two jammed nuts block motion, turn them in close to the clamp.

If you have a conventional screw for a leadscrew, the span between the screwhead and the jammed nuts reduces the distance that the slide can move. The screwhead prevents the slide from running off the end of the screw so I haven't suggested to cut it off. You may adjust the position of the screw to allow the slide to reach either end of the range of motion, but not conveniently.

In any case, use whatever range of free motion you have to assess how well the slide slides.

• The only points of contact between slide halves should be the bearings running on the rods, and the leadscrew sliding freely through clearance holes. Although clearances between hard parts will be under a millimeter, there should be no other contact.
• Drag should be slight and feel uniform from end to end. If drag increases or decreases progressively along the range of travel, check parallelism & concentricity of rods & bearings. Especially check that all six rod ends have the full diameter of the rod resting in the V blocks — this can be a problem with deeply chamfered rod ends (normally a value-add feature but not so helpful here)
• The “feel” probably won’t be magically buttery smooth with cheap bearings, but it should be uniform from end to end, and fairly smooth
• For a better test: the “feel” should not degrade, and perhaps improve, while shifting the slide under load. To load the slide with your hands to feel how it moves, support the 2-rod side by its ends and load the 1r2b end of the 1-rod side (see photo). I don’t lift or anything, so this isn’t saying a whole lot, but I can’t squeeze hard enough between my hands this way to make a slide feel any different than it does under lighter load.

As much as the slide should run freely in one dimension, it should move not at all in any other way.

Obviously the flexibility of 1/8" hardboard limits ultimate rigidity. Flexibility of the thin flat sides can be mitigated by applying loads close to the ends as pictured in the previous step. For example, when the four corners of the two-rod side are attached to a solid base and load applied to the two-bearing end of the moving side, the slide should strongly resist loads normal to the sliding axis. Especially for compression loads that force the two slide haves into each other rather than pulling the clamps away from the screws holding them down.

For an initial check of radial free play: firmly grip one half of the slide in each hand, move the slide to mid-travel, and twist, push, and pull in all ways that don't make the slide move along it's intended sliding axis. Nothing should go "clunk-clunk". If something does, try to isolate the "clunk-clunk" to one bearing or rod end. The photo shows one way to try to find something you can feel but not see: holding just one rod by its two ends and just one clamp by its ends lets you apply force to, and feel, just that bearing and rod. By testing both ends of the rod you can find a problem at one rod end or focus on the bearing if it feels the same at all points along the rod.

To tighten up a clamp joint -- one that's actually loose vs just twisting screws and hoping for better -- in one of the exposed ends you can simply compress the clamp as before. For the bearing ends, now captured within their opposite side, move the slide to mid-range with the clamp in the middle of its opposite side then use a credit/debit/ID-type card or something similar to fill the gap between the clamp and side plate. If compressing against a flat surface, use something thicker than the screwheads to support the slide side under the clamp.

If you find radial play between a rod & bearing -- enough to care about -- you may be able to find a different pairing of rods & bearings that works better. For example, if you have your loosest bearing on your smallest rod then re-arranging rods & bearings might give you three good fits instead of one great fit, one good fit, and one poor fit.

That might help you solve a loose fit somewhere now. If not, consider whether the actual free play interferes with your intended application — you may be feeling a very small bit of play that won't make any real difference.

※If building a variant without a manual leadscrew, return to variant instructions; you have finished this first/main part of the 'ible!

(※If building a variant without the manual leadscrew, go back to the variant procedure. You have already finished with this part of the 'ible!)

The two washers should stay with the slide structure and not turn with the screw. Adding a little tackiness between the washers and slide end helped one side of that. Adding slipperiness between the washers and leadscrew will help the other side.

Add small amounts of oil:

• between the nut already fixed on the screw and the inner washer
• to the outer washer

Consider expectations. Leadscrews have backlash. Distinct from CNC, people have been doing precision work with backlashy hand-cranked leadscrews for centuries. The trick is simply this: always approach critical positions from the same direction. So, how close to zero backlash do you really need? Do you have number in mind? Ok.

Parts:

• the slide
• two nuts

Tools:

• wrench
• needle-nose pliers (or thin/tappet wrench if available)

At this point the screw can push the two halves apart but not pull them together. The difference between the two actions is "backlash", which won't be zero. So adding the "pull" part will set some amount backlash for the finished slide.

Set backlash

1. Close up the slide to the limit set by the nuts already fixed on the screw
   • back the leadscrew away from the 1r2b end if it's not the limit to closing the slide
2. Add two more nuts to the end of the leadscrew
3. Turn the first nut finger-tight against the end of the slide
4. Turn that same nut back 1/3 turn (two faces of a hex nut — about 1/100") and hold it with the thin tips of the pliers
5. Use the wrench to lock the second nut against the first
   • without rounding the corners off the first nut!
6. Check that the leadscrew can turn without much additional friction from this end
   • if not, reset the nuts back another 1/3 turn

Now the leadscrew can do its job.

Can you live with the backlash? If yes, you're done here; move on to the next step!

Adjust backlash

At this end, anyhow. Backlash where the screw engages the threads in 1r2bc can be very small if the screw threads are consistent so the compliant hardboard material can conform closely to the screw. There is also flexure of the slide ends proportional to load, but that's a different thing.

To adjust backlash: back off the second nut, tweak the first nut, and lock it again with the second nut.

• very small rotations of the first nut can make a big difference when tweaking for small backlash
• contacting the washers with a little friction is ok — you will be driving the leadscrew at low speed and low duty-cycle with lots of torque

You can get the backlash down to something pretty small. In one imprecise attempt to measure, without the appropriate tool, a slide than I'd set up with maybe five minutes of tweaking backlash, I came up with something vaguely in the neighborhood of 50 μm.

Parts:

• handwheel
• 3/4" or 1" screw non-countersunk screw
• locking nut (nyloc, cap nut, nut+CA, etc.)
• stiff tube-ish bit to fit over screw between wheel and locking nut

A revolving crank handle isn't essential, but with ~85 turns end-to-end it may make some applications significantly less tedious. ("Revolving" means the handle revolves relative to the wheel. With some irony, the point is that the handle you're holding does not revolve while the wheel does. I didn't name the concept.)

Assemble as shown.

Required:

• the handwheel
• the last nut

Tools:

• 2 wrenches

If not already done: drive a screw through the handwheel to pre-form thread to fit the leadscrew.

1. Screw the handwheel loosely onto the end of the leadscrew
2. Use a wrench to hold only the first/innermost nut on the leadscrew
   • see photo
3. Turn the handwheel firmly into place
4. Add the last nut to the end of the leadscrew
5. With one wrench hold only the first/innermost nut on the leadscrew
6. With a second wrench tighten the last nut firmly against the handwheel
   • the last nut should end up exactly at the end of the leadscrew, with no exposed screw or nut thread

The feel of the leadscrew turning in the hardboard threads will improve after a few cycles of running the slide end-to-end. The siren of Novelty may inspire some of that cranking right now...

A small chip of hardboard, e.g. laser-cutter waste, placed as shown in the photo will limit slide travel to prevent pressing the bearings back into the slide ends.

You have completed your copy of the Cheap Compact Linear Motion Slide shown in the cover picture and "repeatability" demo!

You can stop here.

Really. You can.

The rest of the 'ible describes other stuff.

This big scary red eye-catcher is here for someone scrolling through this huge 'ible to decide whether they want to tackle it or not:

Dear you,

Please stop scrolling! That was the 'ible for the "Cheap Compact Linear Motion Slide" picture on the label. If that's what you came for, you don't need to deal with any more of what follows after here. Really.

(image CC0)

Although presented first with a simple manual leadscrew, this slide was designed around a small stepper motor + leadscrew unit. The next part of this 'ible describes how to build a powered version of the slide.

The guts of discarded CD/DVD drives — especially those that use a stepper motor and lead screw to drive the optics sled — provide great building blocks for CNC toys. But if your thing needs a little more shove from that motor — and you're already running it close to Curie temperature with a current-limiting high-voltage micro-stepping driver with the compliance shimmed out of the springy leadscrew follower — what's the least-cost step up?

Possibilities include a stepper-screw-slider unit with the same 15 mm motor as many optical drives but with the ≥3 mm pitch leadscrew swapped for M3 thread with 0.5 mm pitch. That promises much more linear force from the same motor, in trade for speed. And the self-locking shallow pitch will hold position without constant power to the motor. The CH-SM1545 motor+leadscrew unit in particular is a little shorter than similar but longer generic examples. At 88 mm overall length it's conveniently 4 × 1/8" shorter than a 100 mm rod, which works out very nicely for "cheap" and "compact".

An important limitation of that motor+screw combination in this application surprised me.

Because the screw has a self-locking thread, little axial compliance, and nearly no rotational inertia, it can chatter badly when the net load force acts in the direction of motion (§2.2). For example, when running vertically and moving downward. Gravity usually helps with that, but not this time. While this slide and motor can lift a fair load upward it may be unable let even its own empty moving part ease downward smoothly with any speed.

Conveniently, 500g counterweights are widely available, which will come up again later when this 'ible gets to the part about attaching a ~500g load.

(Photo by Ryan McGuire; Photo by Akash Rai)

From the beginning of this 'ible, follow up to Step 3 to

1. laser cut the flat parts
   • but use one of the "motor..." SVG files attached to this step in place of the "wheel..." files there.
2. collect other parts and tools

In addition to parts and tools for making the common basic slide:

Parts:

• CH-SM1545 motor+leadscrew unit (occasionally updated source/cost info may help)
• paperclip
• motor mount screws (more about these later)

Tools:

• M3 tap
• small (#0) Phillips/crosshead screwdriver to fit the motor cover screws
• Intermediate (#1 Phillips/crosshead?) driver with narrowest shaft big enough to turn the motor mount screws

The next few steps show how to modify the CH-SM1545 motor+leadscrew and the paperclip, and how they fit into the construction steps in the first mainline part of this 'ible.

Parts:

• Backlash adjuster wheel

Tools:

• M3 tap

The flat parts cut from the "motor..." flavor SVGs include a small toothed wheel. That will run on the M3 leadscrew. Because the small motor makes less torque than your arm can, and you will probably want to use the little torque it can make to move the slide, it requires a free running fit through the adjuster wheel.

• Use the tap to cut thread in the backlash adjuster.
• You may find it difficult to keep the small wheel perpendicular to the tap.
• This works for me:
  • turn a small arc with the tap (or with the wheel on the tap)
  • spin the tap and watch how the wheel wobbles
  • adjust the wobble out of the wheel
  • repeat
• Try to minimize wobble/adjustment and get to a stable straight condition as soon as possible.

Aside:

This paragraph is hidden here to avoid misdirecting anyone building a manual slide from the main body of the 'ible. The backlash adjuster should work for a handwheel slide too. Select the "spares" option for a handwheel configuration to get a backlash adjuster wheel sized for the tap diameter of the manual leadscrew (step 2 "long form"). So far the only time I've wanted to try it was with the good part of an imperfect leadscrew that made excess thread clearance for itself - but I didn't have a spring ready so I don't know how well that would have worked. It could possibly become relevant if the thread running on a leadscrew becomes worn by long use. Otherwise haven't seen any indication that it would be worth the bother. Conceivably one might put e.g. brass threaded inserts in the part threaded for the leadscrew and in a backlash adjuster wheel to get low backlash without relying on threaded hardboard; if anyone goes that far with this please let me know!

(...stashed here to avoid misdirecting anyone building a manual slide from the main body of the 'ible.)

The backlash adjuster should work for a handwheel slide too. Select the "spares" option for a handwheel configuration to get a backlash adjuster wheel sized for the tap diameter of the manual leadscrew (step 2 "long form"). It's not actually necessary with a manual screw as it is with the motor. So far backlash with manual leadscrews has been small enough to not motivate doing anything about it.

(image CC0)

Parts:

• CH-SM1545 motor+leadscrew unit (sources)
• backlash adjuster wheel (tapped)
• connector parts to mate with or convert the 1.25 mm pitch Molex Picoblade connector

Tools:

• #0 cross-head screwdriver
• sharp nipper/knife/cutty thing

The motor+leadscrew unit requires a few modifications.

• cut the extra tab off the white slider
• add the backlash adjuster to the leadscrew
• optional: modify connector if not mating with 1.25 mm Molex Picoblade compatible connector

Take care to cut only the extra tab away without taking any more material from the thin top of the slider over the internal captive nut.

Adding the backlash adjuster to the leadscrew requires taking the back cover off the motor so that the magnetic rotor attached to the leadscrew can be shifted out the back of the motor. Beware this exposes a tiny ball bearing at each end of the screw.

So far I’ve found this method fairly quick and trouble-free:

• turn the screw so that the slide is about 5-7 mm from the motor end
  • far enough for the backlash adjuster to clear under the end of the screw
  • close enough to keep the rotor just inside the back of the motor
• stand the unit upright on its nose (the end that is not the motor)
• back out but do not remove the two screws in the back (now top) of the motor
• lift the cover/spring plate off the motor, with the two screws in it
• set the cover plate aside, preferably with the screws straddling something narrow that won’t push them up out of the plate
• carefully push the slider up against the motor, without disturbing the white disk and bearing on top of the rotor
  • the rotor should not come entirely out of the motor case (for convenience)
  • the rotor magnet should hold the rotor & screw up in this position
• check whether the bearing ball at the nose/bottom end of the leadscrew is sitting in the white plastic cup or stuck to the end of the screw
  • so that you know where it is
  • when the bearing ball is on the end of the screw, I find it very difficult to place the backlash adjuster without knocking the ball off the screw; so I’ll dip the screw back down into the bearing cup and try to get the ball to stick in the cup
• bring the backlash adjuster under the end of the screw with the good side up/toward the motor
• thread the adjuster onto the screw and turn it up a few turns from the end of the screw
• with a fingertip on the white disk (& bearing) on top of the rotor, slide the screw back down so that the nose of the screw lands on its bearing and the disk fits back into the motor case without dislodging the bearing under it.
  • if the motor end bearing gets out of the cup on the end of the screw, it will likely be captured by the magnetic rotor; in which case it’s not lost but still can be tricky to get it away from the magnet and back into place
  • don’t expect the screw to turn freely; it won’t unless the bearing carrier disk on the motor end is pressed in fully to press the two bearings into the ends of the screw
• carefully lift the motor cover/spring plate and two screws and place it back on the motor
• secure the cover with the two screws
• check that the screw turns freely (with the usual “cogging” of the stepper motor)

Parts:

• four screws (see below)

Tools:

• file, maybe

The screws that attach the motor+leadscrew frame to the slide must have heads thin enough to avoid blocking the sliding leadscrew follower. You can buy or make screws that fit.

Buy: As noted in the part source list, there is a specific short #6-32 screw with a very thin head that fits this application well. In the photo they are the small screws on the left. But requiring a relatively expensive specialty part defeats the cheap common materials theme of this project.

Make: Alternatively, the same ¼” screws used throughout this assembly (e.g. right side of the photo) can be made to fit by filing down the screw heads so that the slider can pass over them. The ¼" screws will stick through the side of the finished slide. If you need the surface to be flat without screw ends sticking out, then file some length off the end of the screw. The middle screws in the photo show thinned heads and shortened threads.

Go to Step 4 and follow the main line of the 'ible up to Step 14 to start building the slide.

Parts:

• motor
• 4x motor mount screws
• ※1- or 2-rod side, as configured
  • ※2-rod side for the attached "motor2r..." example SVG
  • (photos show motor on 1-rod side of a reflected slide)

Tools:

• screwdriver
• screwdriver with thin shaft

1. Drive a (any) #6-32 screw through each of the four motor mount screw holes to form threads
   • partial interference with the motor will prevent a forceful straight drive while attaching the motor
2. Start two motor mount screw partially in the two holes along one side of the motor+screw frame
   • either side, but same side
   • leave clearance for the thickness of the motor frame under the screw head
   • photo shows one specialty screw and one modified standard screw
3. Slide the motor frame into place against the two screws under the heads. (see photo)
   • tucking the wires around under the motor and out the other side of the slide provides a degree of strain relief and immobilization to reduce fatigue where the wires join the motor
4. Start the remaining two screws
   • turn the screws down close to but not gripping the motor frame
   • the guide rods partially obscure the screw heads; requires driving these screws from an angle into the straight pre-threaded holes using a thin screwdriver
   • the photo shows how a slightly smaller screwdriver permits a much better angle for driving the screws
   • to start, lift the motor frame up to the bottom of the screw head to raise the rod as high as possible over the screw head, rather than starting with the frame flat down on the surface and the screw heads close under the rods
5. Push the frame firmly against the two screws on one side
6. Firmly tighten the screws on the side that you are pushing away from
   • especially if using slightly countersunk thin-headed screws
7. While continuing to hold pressure to the same side, tighten the two screws on the side that you are pushing toward
8. Check that the screws hold the motor firmly without allowing it to shift
9. Turn the leadscrew to move the white slider to the motor end of the screw
10. Turn the adjuster wheel to move the adjuster to the other end of the screw

Parts:

• paperclip, or >8 cm of ~0.8 mm similarly springy+bendable wire

Tools:

• at least one, maybe two pair of pliers
• wire cutter (likely integral with pliers)
• round bending mandrel, e.g. a #1 Phillips screwdriver shaft
• mm scale ruler

1. Determine the "handedness" of spring needed
   • if using the attached sample "motor2r.." SVG, the spring will lay under the rod to the right of the leadscrew clearance cut when viewed in the orientation shown: call it "green-handed"
   • if building a different configuration, determine whether to make a "green-handed" or "magenta-handed" spring from the location of a rod next to the clearance cut
2. Straighten the paperclip
3. Bend it as shown in the drawn diagram
   • to make a "green-handed" spring follow the green/second/middle row of steps
   • to make a "magenta-handed" spring following the magenta/third/last row of steps
4. Tweak it as shown in the photo diagram
   • tweaks for a "green" spring shown
   • reflect for a "magenta" spring

1. Follow the photo sequence to install the spring
   • there will be a screw head in the way that does not appear in these photos; you may be able to get the spring over the screw head without deforming the "spring" out of it; if not remove & replace the screw head; beware to avoid capturing the end of the spring under the screw head when replacing the screw
2. Check that, when compressed, the spring returns by its own force
   • tweak if needed so that it does

1. Return to the main body of the 'ible and continue from step 19 to join the two sides
   • the cutout in the end of the driven half of the slide should fit closely over the leadscrew between the slider and backlash adjuster.
   • when securing the clamp that crosses the leadscrew, ※1r2bc at step 22 if using one of the attached SVGs, place one or (more likely) two credit/debit/ID-type cards or other uniformly flat incompressible shims under the clamp on each side of the motor so that compressing the clamp does not put pressure on the leadscrew. (see photo for shim placement, although you will have the slide less extended [bearing ends further apart] for that step)
2. Build up through Step 31 to complete and inspect the free-moving linear slide
   • the slide will have full range of motion if the leadscrew follower and backlash adjuster wheel are pre-positioned to opposite ends of the screw as indicated in preceding steps for preparing the motor
   • if attaching the motor with modified 1/4" screws, check that the driven clamp clears the screwheads (see photo)
3. Continue with the following step to set and adjust backlash

1. Bring the leadscrew slider and backlash adjuster together, at some point away from the non-motor end of the screw, so that some length of screw is exposed
   • turn the leadscrew to move the slider a short distance away from the motor end
   • while holding the screw, turn the adjuster wheel to move it toward the slider
2. When the backlash adjuster reaches the locking spring on the driven slide half, lift the spring away from the leadscrew and turn the adjuster wheel to move it under the spring
   • when released, the spring should engage the teeth of the adjuster wheel to prevent it from rotating relative to the slide
   • the integrated leadscrew slider and the backlash adjuster wheel will then move in tandem, with the driven half of the slide held between them, as the leadscrew turns
3. To minimize backlash, turn the adjuster toward the slider until it clamps the slide end between the two; then turn it back away from the slider by small increments until the motor is able to turn the screw
   • it should be possible to reduce backlash so that there is no feel of free play
   • there is some axial compliance built into the motor and leadscrew; to check backlash just between the screw and slide, hold the screw against one end of its range of compliance

The slide includes the dubious “feature” of a place for a screw adjustment to constrain the axial compliance of the leadscrew. My first trial of that function didn’t go well. So far I haven't met need to try it again. If you try it, please let me know how it goes!

If you need something to drive a stepper motor, here is another 'ible that might help:

Scrappy Integrated Grbl CNC Controller & Power

You have completed a Cheap Compact Linear Motion Slide with precisely controllable stepper motor drive.

Each full step of the motor nominally moves the slide 1/20 of 1/2 mm, or 25 μm ≈ 0.001 inch.

Step 2, the "long form" for cutting flat parts, links to the CAD model where you can configure your own slide.

For each side of the slide you can choose a variation of side plate for different functions:

• Plain: plain
• Mounting Tabs: extends a side to include a square pattern of screw holes on 96 mm centers for attaching the slide to something bigger; includes spacer feet to clear screw heads on that side so mounting screws may be tightened down on solid material
• Detachable Plate: adds a simple detachable plate and screw hole pattern for attaching smaller things to the slide or to provide a consumable surface; the plate and hole pattern have simple fixed dimensions for easy reproduction
• Stacking: when selected as mating surfaces of two slides (e.g. "top" of one and "bottom" of the other) provides for attaching them together with perpendicular axes as a 2D unit; dimensions vary with material thickness so requires both slides to be cut for same thickness of material
• Stacking (simple): simply provides clearance for the screwheads of another slide so the two can stack together for ad hoc attachment
• Tool Clamp: holds a generic Dremel®-like rotary tool or may be adapted for other objects; further described in following steps

Apart from adding more tap-size screw holes to pre-thread before assembly, slides with differently shaped side plates still go together the same ways as described in the bulk of this 'ible.

This has all been mostly harmless so far, apart from general hazards of getting out of bed and touching stuff.

The "side" options include a clamp to hold a Dremel-like rotary tool. Rotary tools are actively dangerous with typically irreversible effects. Half of the ways a slide can move an attached rotary tool are in the direction of dangerous effect. A motorized slide can be commanded, intentionally or unintentionally, to move a dangerous tool up to about a finger-length in the direction of irreversible effect with ABSOLUTELY ZERO empathy.

Not at all harmless.

Whether or not to attach a rotary tool or any other cause of effects to a slide and what harm or benefit may follow if you do are entirely up to you.

1. If you've already built a slide to carry the tool clamp, the two clamp sides and three saddles should go together easily.
   • the good sides of the side parts face inward -- they are not identical/interchangeable
   • turn the good side of the saddle at the spindle to face inward -- to better hold the end/edge of the cylindrical part of the tool
   • Beware the clamp is very fragile until attached to the slide; when driving screws make sure to directly hold the part that you will press into because the slide will fold up easily if you press on one side while simply holding the clamp in your hand
2. Attach the clamp to the slide side
   • fit the hooked tabs of the clamp into the long slots in the slde
   • rotate the clamp close to parallel with the slide
   • push the clamp in the direction of the hooked tabs until the split tabs can fit into their slots
   • push the split tabs into their slots to rotate the clamp into place parallel with the slide
   • secure the clamp with four screws

The clamp attaches to the slide with screws at only one end so you can add/remove or swap one clamp for another without disassembling the slide. As an experiment to see if this can work without a sloppy fit at the hooked end, the clamp hooks and slots have intentional interference. Assuming the material yields enough for the parts to fit together -- hardboard will -- this should pull the clamp sides together and tight against the slide. At least until the material yields further or wears away. So far it's worked for me.

While many dimensions derive from the material thickness, the relative positions (obviously not the thicknesses) of the clamp tabs should hold constant for different material thicknesses. While tabs may require thinning or shimming, a slide should be able to use clamps cut from different thicknesses of material.

Expect a small gap under the middle saddle. It has a little clearance to keep it from bearing against the bendy middle of the slide side plate. It serves more as a reference and imprecise constraint for axial positioning of the tool than for strength.

Use low stretch light line ~2mm or less in diameter. In the photos I've used very light synthetic line because that was what came to hand the first time I tried this idea. It worked. Since then I've continued to use the same stuff to see how long it lasts and how it fails. So far it has lasted and not failed. But if choosing more deliberately I'd look for something a little less light.

Follow the photo example:

1. Find the two small holes in each of the two end saddles and one small hole in each clamp side
2. Thread a line through the saddle and one of the windlass wheels at each end of the clamp and tie off the ends with stopper knots that can be untied after loading; i.e. not overhand knots!; I've used figure-eight knots
3. For each side tie a loop (I've used bowlines) and secure through one side with a stopper knot

The "horns" on the clamp saddles help to hold the loops open as a convenience when passing the rotary tool through the loops.

Easy to do: Adjust the clamp loops to "close enough" and use the "Spanish windlass" wheels to wind them up tight like a tourniquet. You should be able to wind on all the tension you want until something breaks.

Easy to use: Adjust the clamp loops to a point that's just "too tight" so that you can't freely turn the windlass wheel to "close" the loop from a "Z" shape to a straight line, but with carefully applied firm torque it will "snap" through the peak resistance and then hold the loop tight with much less torque than it took to get there. In my experience, a loop adjusted "just right" will "snap over" without any awkward excess of torque and hold the tool effectively without winding any further ("tight" photo). When adjusted to "snap" closed, but maybe not quite as tight as you would like, winding another 45° or 90° will add significantly more tension. Assuming low-stretch cord.

Either way: Adjust the retaining loops to centerline on top of the tool so the points where the loops hold a hook of the windlass wheels line up close to the centerline and pull tangent to the wheels.

For more about the origin and ongoing application of this cheap compact slide mechanism:

• Minamil: a minimal CNC mill
• A Cheap Compact Linear Slide
• "Desk Accessory" CNC mill

And a little further back in the causal chain:

• CDCNC

This is the cheapest way to build a comparably capable device that I know of, it's working pretty well, and I hope it will enable/encourage someone to find other applications and improvements.

Please give it a try and share how it goes.